OPINION

In this case we decide whether a claim for infliction of emotional distress can be brought in a divorce proceeding. Because the judgment of the court of appeals is based on negligent infliction of emotional distress, and cannot be affirmed on that or any other basis, we reverse the judgment of that court and remand this cause for a new trial in the interest of justice. Tex. R.App.P. 180. We deem a new trial appropriate because of our recent decision that no cause of action for negligent infliction of emotional distress exists in Texas. Today, however, we expressly adopt the tort of intentional infliction of emotional distress, and hold that such a claim can be brought in a divorce proceeding.

I.

Sheila and William Twyman married in 1969. Sheila filed for divorce in 1985. She later amended her divorce petition to add a general claim for emotional harm without specifying whether the claim was based on negligent or intentional infliction of emotional distress. In her amended petition, Sheila alleged that William “intentionally and cruelly” attempted to engage her in “deviate sexual acts.” 1 Following a bench trial, the court rendered judgment dissolving the marriage, dividing the marital estate, awarding conservatorship of the children to Sheila, ordering William to pay child support, and awarding Sheila $15,000 plus interest for her claim for emotional distress. William appealed that portion of the judgment based on emotional distress, contending that interspousal tort immunity precluded Sheila’s recovery for negligent infliction of emotional distress. The court of appeals affirmed the judgment, holding that Sheila could recover for William’s neg*621ligent infliction of emotional distress. 790 S.W.2d 819.

While this case has been pending, we have refused to adopt the tort of negligent infliction of emotional distress. See Boyles v. Kerr, 855 S.W.2d 593 (Tex. 1993). Thus the judgment of the court of appeals cannot be affirmed. We consider, therefore, whether the court of appeals’ judgment may be affirmed on alternative grounds. Because Sheila’s pleadings alleging a general claim for emotional harm are broad enough to encompass a claim for intentional infliction of emotional distress, we consider whether the trial court’s judgment may be sustained on that legal theory.

While this court has never expressly recognized the tort of intentional infliction of emotional distress, we found no reversible error in the court of appeals’ opinion in Tidelands Automobile Club v. Walters, which did so. 699 S.W.2d 939 (Tex.App.-Beaumont 1985, writ ref’d n.r.e.). There, the court of appeals adopted the elements of the tort as expressed in the Restatement (Second) of Torts § 46 (1965). The Restatement elements of intentional infliction of emotional distress are: 1) the defendant acted intentionally or recklessly, 2) the conduct was extreme and outrageous, 3) the actions of the defendant caused the plaintiff emotional distress, and 4) the emotional distress suffered by the plaintiff was severe. Id. According to the Restatement, liability for outrageous conduct should be found “only where the conduct has been so outrageous in character, and so extreme in degree, as to go beyond all possible bounds of decency, and to be regarded as atrocious, and utterly intolerable in a civilized community.” Id. cmt. d. Of the forty-six states that have recognized this tort, forty-three have adopted this Restatement formulation.2 The other three states, although not adopting the Restatement definition, require the equivalent of “outrageous” conduct.3 Today we become *622the forty-seventh state to adopt the tort of intentional infliction of emotional distress as set out in § 46(1) of the Restatement (Second) of ToRts.

We do not, however, adopt this tort only because of its broad acceptance in jurisdictions throughout the United States. As distinguished from the tort of negligent infliction of emotional distress, we believe the rigorous legal standards of the Restatement formulation of intentional infliction of emotional distress help to assure a meaningful delineation between inadvertence and intentionally or recklessly outrageous misconduct. The requirements of intent, extreme and outrageous conduct, and severe emotional distress before liability can be established will, we think, strike a proper balance between diverse interests in a free society. That balance, at minimum, must allow freedom of individual action while providing reasonable opportunity for redress for victims of conduct that is determined to be utterly intolerable in a civilized community.

This holding represents a middle ground between the polar positions adopted by various members of the court.4 JUSTICE HECHT, joined by JUSTICE ENOCH, in arguing against our express adoption of the tort of intentional infliction of emotional distress, maintains that judges and juries are guided by insufficient standards, that liability may be imposed arbitrarily, that reported cases either supporting or refusing to support an award of damages disclose no uniform patterns, and that the sensitivities of aggrieved people are entirely too subjective and unpredictable. We disagree, and believe that such objections could just as easily be made to well-established causes of action in Texas. For example, one might also contend that the legal standards for ordinary negligence are vague, and that juries must necessarily rely on their own notions of fault. Because jurors’ ideas about what is “ordinary” and “reasonable” may vary, the same arguments about lack of uniformity, unpredictability, and personal sensitivities could be made. Yet just as we trust juries to decide questions of negligence, proximate cause, and damages, when guided by appropriate legal standards we think them equally capable of resolving factual disputes giving rise to the tort of intentional infliction of emotional distress.

JUSTICE SPECTOR, joined by JUSTICE DOGGETT, on the other hand, agrees with us that this tort should be adopted, but uses this case as another opportunity to question the wisdom of our decision in Boyles, in which we refused to adopt the tort of negligent infliction of emotional distress. They join some amici curiae5 in implying that the court has disregarded the tort’s unique role in addressing women’s psychic injuries. One need only identify those who have brought claims for negligent infliction of emotional distress, however, to dispel the suggestion that women will be disproportionately affected. • Of the thirty-four Texas appellate cases in which a claim for negligent infliction of emotional distress was alleged,6 thirteen were brought by women,7 twelve were brought *623by men,8 seven by husbands and wives jointly,9 one by an executrix on behalf of an estate,10 and one by a corporation.11 These cases demonstrate that the tort has been alleged by litigants in a wide variety of circumstances. There is simply no factual or legal basis for the suggestion that by choosing not to recognize this particular tort, the court demonstrates insensitivity to female claimants.

JUSTICE SPECTOR also argues that because of our refusal to recognize the tort of negligent infliction of emotional distress some wrongs will go uncompensated because of the difficulty in proving the actor’s intent when the actor intends nothing more than to satisfy his own desires. Infra, 855 S.W.2d at 644. But in Sheila Twy-man’s case, and in countless other cases involving both men and women, we believe that our adoption of the Restatement formulation of the tort of intentional infliction of emotional distress provides a reasonable opportunity for redress for outrageous conduct.

Of course, rarely will a defendant admit knowing of a substantial certainty that emotional harm would befall the victim. Juries, however, are free to discredit the defendant's protestations that no harm was intended and to draw necessary inferences to establish intent. See McGalliard v. Kuhlmann, 722 S.W.2d 694, 697 (Tex. 1986); Fichtner v. Richardson, 708 S.W.2d 479, 483 (Tex.App.—Dallas 1986, writ ref’d n.r.e.) (“The jury may believe all, part, or none of the testimony in arriving at the finding it concludes is the most reasonable”); see also Walters v. American States Ins. Co., 654 S.W.2d 423, 426 (Tex. 1983) (holding that jury is free to make a *624reasonable inferential leap based on evidence). Ironically, for JUSTICE SPECTOR to argue in favor of applying a negligence standard to this case is to argue that as a matter of law the emotional harm William caused was foreseeable but not substantially certain to occur.12 We disagree with that characterization and believe that on retrial the factfinder should be permitted to consider whether William knew with substantial certainty that his actions would probably cause Sheila emotional harm.

Moreover, Section 46 of the Restatement definition of the tort expressly includes situations in which the actor recklessly inflicts emotional distress. An actor is reckless when he “knows or has reason to know ... of facts which create a high degree of risk of ... harm to another, and deliberately proceeds to act, or fails to act, in conscious disregard of, or indifference to that risk.” Restatement (Second) § 500, cmt. a. Again, on retrial, the jury may consider whether William acted recklessly toward Sheila.

II.

We now consider whether the cause of action for intentional infliction of emotional distress may be brought in a divorce proceeding.13 In Bounds v. Caudle, this court unanimously .abolished the doctrine of interspousal immunity for intentional torts. 560 S.W.2d 925 (Tex. 1977). Ten years later, we abrogated interspousal immunity “completely as to any cause of action,” 14 including negligence actions for personal injuries. Price v. Price, 732 S.W.2d 316, 319 (Tex. 1987). Under the rules established in Caudle and Price, there appears to be no legal impediment to bringing a tort claim in a divorce action based on either negligence or an intentional act such as assault or battery.15

The more difficult issue is when the tort claim must be brought and how the tort award should be considered when making a “just and right” division of the marital estate. See Tex.Fam.Code § 3.63(b). Of the states that have answered this question, several have held that the tort case and the divorce case must be litigated separately. See e.g. Walther v. Walther, 709 P.2d 387, 388 (Utah 1985); Windauer v. O’Connor, 107 Ariz. 267, 485 P.2d 1157 (1971); Simmons v. Simmons, 773 P.2d 602, 605 (Colo.Ct.App. 1988). Other states require joinder of the two actions. See, e.g. Tevis v. Tevis, 79 N.J. 422, 400 A.2d 1189, 1196 (1979); Weil v. Lammon, 503 So.2d 830, 832 (Ala. 1987).

We believe that the best approach lies between these two extremes. As in other civil actions, joinder of the tort cause of action should be permitted,16 but subject to the principles of res judicata.17 See Barr *625v. The Resolution Trust Corp., 837 S.W.2d 627, 631 (Tex. 1992) (reaffirming the transactional approach to res judicata analysis). See also Tex.R.Civ.P. 51; Henriksen v. Cameron, 622 A.2d 1135 (Me. 1993). Of course, how such claims are ultimately tried is within the sound discretion of the trial court. See Tex.R.Civ.P. 174. But join-der of tort claims with the divorce, when feasible, is encouraged. Resolving both the tort and divorce actions in the same proceeding avoids two trials based at least in part on the same facts, and settles in one suit “all matters existing between the parties.” 18 Mogford, 616 S.W.2d at 940 (citing Parkhill Produce Co. v. Pecos Valley S. Ry., 348 S.W.2d 208, 209 (Tex.Civ.App.—San Antonio 1961) writ ref'd n.r.e. per curiam, 163 Tex. 88, 352 S.W.2d 723 (1961)).

When a tort action is tried with the divorce, however, it is imperative that the court avoid awarding a double recovery.19 When dividing the marital estate, the court may take into account several factors, including the fault of the parties if pleaded. See Murff v. Murff, 615 S.W.2d 696, 699 (Tex. 1981). The trial court may also consider “such factors as the spouses’ capacities and abilities, benefits which the party not at fault would have derived from continuation of the marriage, business opportunities, education, relative physical conditions, relative financial condition and obligations, disparity of ages, size of separate estates, and the nature to the property.” Id. See also Young v. Young, 609 S.W.2d 758, 761 (Tex. 1980); Bell v. Bell, 513 S.W.2d 20, 22 (Tex. 1974). However, a spouse should not be allowed to recover tort damages and a disproportionate division of the community estate based on the same conduct. Therefore, when a factfin-der awards tort damages to a divorcing spouse, the court may not consider the same tortious acts when dividing the marital estate. Contrary to CHIEF JUSTICE PHILLIPS’ contention, an award for tor-tious conduct does not replace an analysis of the remaining factors to be considered when the trial court divides the marital estate. 855 S.W.2d at 626 (Phillips, C.J., dissenting). The court may still award a disproportionate division of property for reasons other than the tortious conduct. To avoid the potential problem of double recovery, the factfinder should consider the damages awarded in the tort action when dividing the parties’ property. If a jury is used to render an advisory division of the parties’ estate, the judge should limit, by appropriate instruction, the jury’s consideration of the alleged tortious acts and later consider the award of damages in determining a just and right division of the marital estate.20

Sheila Twyman cannot recover based on the findings of fact made by the trial court in this case.21 It is likely, however, that *626this case proceeded on a theory of negligent infliction of emotional distress in reliance on this court’s holding in St. Elizabeth Hospital v. Garrard, 730 S.W.2d 649 (Tex. 1987), which we recently overruled. See Boyles v. Kerr, 855 S.W.2d 593 (Tex. 1992). As we noted in Boyles, this court has broad discretion to remand for a new trial in the interest of justice when it appears that a case proceeded under the wrong legal theory, and when it appears that the facts when developed on retrial may support recovery on an alternative theory. Id. at 603. See also American Title Ins. Co. v. Byrd, 384 S.W.2d 683 (Tex. 1964); Dahlberg v. Holden, 150 Tex. 179, 238 S.W.2d 699 (1951). When, as here, a party presents her case in reliance on precedent that has been recently overruled, remand is appropriate. See Scott v. Liebman, 404 S.W.2d 288, 294 (Tex. 1966). Therefore, in the interest of justice, we reverse the judgment of the court of appeals and remand this cause to the trial court for a new trial.

concurring and dissenting.

I join in the Court’s recognition of the tort of intentional infliction of emotional distress. Unlike negligent infliction of emotional, distress, which we rejected in Boyles v. Kerr, 855 S.W.2d 593 (Tex. 1993), recovery for intentional infliction of emotional distress is permitted in almost every other state. While not free from conceptu*627al difficulties, this cause of action has proved useful because it “provides a convenient mechanism for achieving a just result in a given case without either dissembling about the facts, manipulating doctrine, or creating new rules regulating the underlying relationship.” Daniel Givelber, The Right to Minimum Social Decency and the Limits of Evenhandedness: Intentional Infliction of Emotional Distress by Outrageous Conduct, 82 Colum.L.Rev. 42, 74 (1982). See also Towards a Jurisprudence of Injury: The Continuing Creation of a System of Substantive Justice in American Law 5-11, 5-17 (Report to the. American Bar Association from the Special Committee on the Tort Liability System, 1984) (the doctrine “represents a synthesis of ideas developed by many courts reacting to numerous factual situations,” and is “an excellent modern example of the common law tradition”). It is not, as one concurring and dissenting opinion suggests, merely a tort which “makes outrageous conduct tortious.” 855 S.W.2d at 629 (Hecht, J., concurring and dissenting). A successful plaintiff must prove not merely outrageous conduct, but outrageous conduct committed by the defendant with the intention of, or reckless disregard to the probability of, inflicting severe emotional distress on the plaintiff. Thus, unlike claims for negligent infliction of emotional distress, actions for intentional infliction can and have been successfully pursued in Texas for a number of years without any demonstrated harm to the orderly administration of justice.

In recognizing this tort, however, I would not extend it to actions between spouses or former spouses for conduct occurring during their marriage. Although this Court has abolished interspousal immunity, Price v. Price, 732 S.W.2d 316 (Tex. 1987); Bounds v. Caudle, 560 S.W.2d 925 (Tex. 1977), it does not necessarily follow that all conduct actionable between strangers is automatically actionable between spouses. Several courts in other jurisdictions have expressly made such a reservation when abolishing interspousal immunity. For example, the Utah Supreme Court wrote in abolishing inter-spousal immunity:

The marriage relation is created by the consent of both of the parties; inherently within such relationship is the consent of both parties to physical contacts with the other, personal dealings and ways of living which would be unpermitted and in some cases unlawful as between other persons.

Stoker v. Stoker, 616 P.2d 590, 592 (Utah 1980). See also Brown v. Brown, 381 Mass. 231, 409 N.E.2d 717, 718-19 (1980); Imig v. March, 203 Neb. 537, 279 N.W.2d 382, 386 (1979); Beaudette v. Frana, 285 Minn. 366, 173 N.W.2d 416, 420 (1969). In accordance with these sentiments, I believe that a tort which is grounded solely on a duty not to inflict emotional distress should not be cognizable in the context of marriage.

Married couples share an intensely personal and intimate relationship. When discord arises, it is inevitable that the parties will suffer emotional distress, often severe. In the present case, for example, Ms. Twy-man testified that she suffered “utter despair” and “fell apart” upon learning that her husband was seeing another woman. She further testified that “[t]he mental anguish was unbelievable to realize, hoping every time, when he went off to Houston, that he was just going to fly and not be with her.” Yet Ms. Twyman seeks no recovery for this distress, and apparently cannot do so under Texas law. Cf Tex. Fam.Code §§ 4.05; 4.06. In such circumstances, the fact finder is left to draw a virtually impossible distinction between recoverable and disallowed injuries.

Furthermore, recognition of this tort in the context of a divorce unnecessarily restricts the trial court’s discretion in dividing the marital estate. Prior to today’s opinion, the trial court could, but was not required to, consider fault in dividing the community property. See Murff v. Murff, 615 S.W.2d 696, 698 (Tex. 1981). The court had broad discretion to weigh any fault along with other appropriate factors, such as relative financial condition, disparity of ages, and the needs of the children. Id. at *628698-699. Now, however, where fault takes the form of “outrageous” conduct intentionally or recklessly inflicted, it becomes a dominant factor that must be considered at the expense of the other factors. Unlike battery, fraud, or other torts resting on more objective conduct, a colorable allegation of intentional infliction of emotional distress could arguably be raised by one or both parties in most intimate relationships.1 As the court noted in Chiles v. Chiles, 779 S.W.2d 127, 131 (Tex.App.—Houston [14th Dist.] 1989, writ denied):

While we recognize the trial court may consider fault in the distribution of community property, we believe permitting such separate damages [for intentional infliction of emotional distress] in divorce actions would result in evils similar to those avoided by the legislature’s abrogation of fault as a ground for divorce.

See also Henriksen v. Cameron, 622 A.2d 1135, 1151 (Me. 1993) (Glassman, J., dissenting) (recognizing the common-law tort of intentional infliction of emotional distress in the marriage context skews “a carefully constructed scheme of legislation governing the marriage relationship”).

Perhaps because of these difficulties, the tort of intentional infliction of emotional distress has not been generally recognized in the marital context. Although most states, like Texas, have abolished inter-spousal immunity, it appears that, until today, only two state supreme courts have expressly held that intentional infliction of emotional distress may be applied to marital conduct. See Henriksen, supra; Davis v. Bostick, 282 Or. 667, 580 P.2d 544 (1978). Moreover, these two decisions do not appear to represent typical actions for the recovery of emotional distress damages. In Henriksen, the husband inflicted on his wife not only verbal abuse but also physical attacks, including multiple assaults and rapes. 622 A.2d at 1139. Similarly in Bos-tick, the husband broke his wife’s nose, choked her, and threatened her with a loaded pistol. 580 P.2d at 545-46. To the extent that emotional distress results from a physical attack or threat of attack, it is already compensable under tort theories previously recognized in Texas. The court in Henriksen apparently recognized the risk of spurious claims in the divorce context, noting that “to protect defendants from the possibility of long and intrusive trials on meritless claims, motions for summary judgment should ... be viewed sympathetically in interspousal cases.” 622 A.2d at 1139. By contrast, it is far from clear that Texas’ strict summary judgment standard will allow our trial courts to use this procedure in weeding out meritless or trivial claims.

I recognize that a few intermediate appellate courts have also applied the tort to spousal disputes. Simmons v. Simmons, 773 P.2d 602 (Colo.App.1988, appeal denied); McCoy v. Cooke, 165 Mich.App. 662, 419 N.W.2d 44 (1988, appeal denied); Koepke v. Koepke, 52 Ohio App.3d 47, 556 N.E.2d 1198 (1989). See also Hakkila v. Hakkila, 112 N.M. 172, 812 P.2d 1320 (App. 1991), appeal denied, 112 N.M. 77, 811 P.2d 575 (recognizing the tort in extremely limited situations); Ruprecht v. Ruprecht, 252 N.J.Super. 230, 599 A.2d 604 (1991) (trial court opinion). On the other hand, at least two high courts have expressly rejected the cause of action as between spouses. Weicker v. Weicker, 22 N.Y.2d 8, 290 N.Y.S.2d 732, 733, 237 N.E.2d 876, 877 (1968); Pickering v. Pickering, 434 N.W.2d 758 (S.D. 1988). See also Browning v. Browning, 584 S.W.2d 406 (Ky.App. 1979).

Just as I join the Court’s decision to recognize a tort now available in nearly every American jurisdiction, I depart from the Court’s decision to extend that tort to a type of dispute where it is not generally applied in other states. I fail to understand how, lexigraphically or logically, it can be “medieval” or “archaic” to decline to adopt a position which has been expressly embraced by only two other state supreme courts. 855 S.W.2d at 643 (Spector, *629J., dissenting). I therefore would reverse the judgment of the court of appeals and render judgment that Sheila Twyman take nothing on her tort claim.

concurring and dissenting.

The wrongful conduct for which the common law offers redress by an award of damages should be defined by standards sufficiently objective and particular to allow a reasonable assessment of the likelihood that certain behavior may be found to be culpable, and to adjudicate liability with some consistency in the various cases that arise. The tort of intentional or reckless infliction of emotional distress which the Court adopts for the first time today does not meet these standards. As proof, even the Court itself is unable to say whether the conduct complained of in this case either is, might be, or is not tortious. The fault lies in the principal element of the tort, the requirement that a defendant’s conduct be outrageous. Outrageousness, like obscenity, is a very subjective, value-laden concept; what is outrageous to one may be entirely acceptable to another. To award damages on an I-know-it-when-I-see-it basis is neither principled nor practical.1 The diverse perspectives present in a free society make decisions about what is outrageous controversial and uncertain in all but the clearest cases; even then, the law must guard against the danger that a consensus against certain conduct is forged of prejudice and passion rather than indignation against intolerable behavior. A law which made outrageous conduct criminal would almost certainly be unconstitutionally vague. A law which makes outrageous conduct tortious, while it may not offend the constitution, is equally injudicious.

For this reason and others, a general, independent tort of intentional or reckless infliction of emotional distress either should not be recognized or should at least be carefully restricted. The plurality opinion gives little indication that it has considered, or why it has rejected, the arguments against adopting this tort. It bases its decision solely on the fact that the high courts of almost all the other states have at one time or another recognized a tort of intentional infliction of emotional distress in some context, and on the conclusion that this tort is more manageable than negligent infliction of emotional distress. While a consensus of our sister courts on a proposition may be some indication of its merit, that circumstance alone has not always been, and should never be, reason enough to justify our concurrence. Besides, our sister courts have agreed only that some intentional infliction of emotional distress should be actionable, a proposition which is virtually irrefutable but which does not support the conclusion that all such conduct should be actionable. Of all the courts which have endorsed the general proposition, only two have ever applied it to allow an action for intentional or reckless infliction of emotional distress between spouses, and two have refused to do so. To the extent we should be guided by the decisions of other courts, they counsel against the Court’s holding in this case. As for the argument that intentional infliction of emotional distress is a more manageable tort than negligent infliction of emotional distress, while it is no doubt true, it proves far too little. A new cause of action, especially one as significant as intentional infliction of emotional distress, should not be adopted simply because it is not as ill advised as other actions which can be imagined.

Slight though the foundation is for today’s decision, its effects are far-reaching. There is little doubt that the new tort will be asserted in many if not most contested divorce cases. The Family Law Section of the State Bar of Texas has filed a brief as amicus curiae assessing the impact of spouses’ suing one another for intentional infliction of emotional distress, discussing *630the arguments for and against allowing such an action without taking a position on the issue, and urging us to exercise caution in considering these arguments. The plurality opinion makes no mention of the family bar’s arguments. But the impact of this case is not limited to divorce litigation; this new general tort may be, and certainly will be, asserted in other contexts. The effect of this change on relationships like employee-employer, insured-insurer, debt- or-creditor, patient-physician, client-lawyer, and buyer-seller, as well as the public generally has not been assessed. Nor has the Court considered the burden of additional jury trials in divorce cases on the judicial system. This Court, as steward of the common law, possesses the power to recognize new causes of action, but the mere existence of that power cannot justify its exercise. There must be well-considered, even compelling grounds for changing the law so significantly. Where, as here, no such grounds are given, the decision is more an exercise of will than of reason.

In my view, intentional or reckless infliction of emotional distress is too broad a rubric to describe actionable conduct, as this case illustrates. Accordingly, I dissent from the Court’s decision to remand this case for trial on such a cause of action. I concur only in the reversal of the court of appeals’ judgment allowing recovery for negligent infliction of emotional distress.

I

The Court accepts section 46 of the Restatement (Second) of Torts [“the Restatement”] as the definitive description of the tort it calls intentional infliction of emotional distress. Section 46(1) states: “One who by extreme and outrageous conduct intentionally or recklessly causes severe emotional distress to another is subject to liability for such emotional distress, and if bodily harm to the other results from it, for such bodily harm.” The tort has four elements: (1) extreme and outrageous conduct (2) done intentionally or recklessly (3) causing (4) severe emotional distress.2 Although the tort is referred to as the intentional infliction of emotional distress, it includes reckless conduct. A person acts with intent if he “ ‘desires to cause the consequences of his act, or ... believes that the consequences are substantially certain to result from it.’ ” Reed Tool Co. v. Copelin, 689 S.W.2d 404, 406 (Tex. 1985), (citing Restatement § 8A (1965)). Recklessness is a lesser standard. A person acts recklessly if he knows or should know that there is a strong probability that harm may result. Restatement § 500, pamph. 2, at 590. The tort is thus broader than its name suggests.

The central element of the tort is extreme and outrageous conduct. Comment d to section 46 of the Restatement describes this conduct as follows:

The cases thus far decided have found liability only where the defendant’s conduct has been extreme and outrageous. It has not been enough that the defendant has acted with an intent which is tortious or even criminal, or that he has intended to inflict emotional distress, or even that his conduct has been characterized by “malice,” or a degree of aggravation which would entitle the plaintiff to punitive damages for another tort. Liability has been found only where the conduct has been so outrageous in character, and so extreme in degree, as to go beyond all possible bounds of decency, and to be regarded as atrocious, and utterly intolerable in a civilized community. Generally, the case is one in which the recitation of the facts to an average member of the community would arouse his resentment against the actor, and lead him to exclaim, “Outrageous!”

The liability clearly does not extend to mere insults, indignities, threats, annoyances, petty oppressions, or other trivialities. The rough edges of our society are still in need of a good deal of filing down, and in the meantime plaintiffs must necessarily be expected and required to be hardened to a certain amount of rough *631language, and to occasional acts that are definitely inconsiderate and unkind. There is no occasion for the law to intervene in every case where some one’s feelings are hurt. There must still be freedom to express an unflattering opinion, and some safety valve must be left through which irascible tempers may blow off relatively harmless steam.

This language describes a very narrow category of behavior which, to summarize, does not include insults, indignities, threats, annoyances, petty oppression, rough language, inconsiderate or unkind acts, or conduct which is only tortious, criminal, intended to inflict emotional distress, malicious, or w;orthy of punitive damages. Outrageous conduct is “extreme”, “beyond all possible bounds of decency”, “atrocious”, and “utterly intolerable in a civilized community.”

How much worse than criminal or malicious must conduct be to be beyond all possible bounds of decency,. and not merely offensive, deplorable or even unbearable, but utterly intolerable in a civilized community? Only the most extremely egregious conduct would seem to qualify. Applying comment d, is it extreme and outrageous conduct — for one person to tell another, as a practical joke, that the other’s spouse has been badly injured in an accident and is in the hospital with both legs broken? —for a man, knowing of another person’s pathological fear of men in women’s clothing, to dress as a woman to surprise and startle the other person? —for the president of a rubbish collectors association to summon a member to a meeting, accuse him of working in another member’s exclusive territory, and threaten that if he does not pay over what he has earned he will be physically beaten and his business destroyed? —for a man to give a woman a bathing suit which dissolves in water, leaving her naked in front of strangers? —for a person to “hex” the farm of a superstitious landowner to force him to sell it? Many would agree that one or more of these situations involves truly outrageous conduct, but I doubt whether there would be much consensus among judges, let alone juries, that the conduct in every instance is “beyond all possible bounds of decency” and “utterly intolerable in a civilized community.” Those who considered one such situation to fall into this narrow category might even object to equating it with others of these examples. In some instances, conduct is more or less reprehensible depending upon the setting in which it occurred and the personalities involved. If comment d were taken literally, truly outrageous conduct would be rare indeed, yet the illustrations in the comments to section 46, taken from actual cases, reflect a less exacting view of outrageousness: in all of these examples a court has determined the conduct to be outrageous, giving rise to liability. The Restatement’s illustrations of outrageous conduct suggest that comment d, though rigid in word, is much more flexible in application.

The vice in such indeterminacy is not that the tort sweeps too broadly, resulting in liability more often than it should. A review of the cases in which intentional or reckless infliction of emotional distress is alleged indicates that while the claim is routinely asserted, it is seldom successful. See Annotation, Modern Status of Intentional Infliction of Mental Distress as Independent Tort; “Outrage”, 38 A.L.R.4th 998 (1985). The vice in the nebulous standard of outrageousness is rather that it results in erratic decisions which appear to have no unifying principle. The cases reveal no clear patterns of application of the standard, nor should they be expected to. What is outrageous unavoidably depends upon the sensitivities of the person asked to decide and to some extent the community in which the conduct occurs. “The term ‘outrageous’ is neither value-free nor exacting.” Daniel Givelber, The Right to Minimum Social Decency and the Limits of Evenhandedness: Intentional Infliction of Emotional Distress by Outrageous Conduct, 82 Colum.L.Rev. 42, 51 (1982). Because outrageousness is a subjective, almost personal, notion, its application is as much a matter of who decides as of what happened.

Proving that conduct is or is not outrageous is virtually impossible. What evi*632dence is there which tends to prove that certain behavior was or was not outrageous? The issue is certainly not a factual one, like whether the traffic light was red or green, on which a witness can be called to testify. It is a matter of opinion, but it involves no “scientific, technical, or other specialized knowledge” as to which a witness could be qualified as an expert “by knowledge, skill, experience, training, or education”. Tex.R.Civ.Evid. 702. The only possible evidence which could be offered on the issue would be the opinion of a lay witness, but it is difficult to conceive of an instance in which a lay opinion would be helpful to a determination of outrageousness and thus admissible. See Tex.R.Civ. Evid. 701. The rules would certainly not allow witness after witness to be called to testify one by one that, “Yes, in my opinion defendant’s conduct was outrageous,” or “No, in my opinion it was not.” Quite simply, the principal element of the tort-of intentional or reckless infliction of emotional distress is something on which testimony cannot be offered and which can neither be proved nor disproved.

This is not a flaw in the tort’s design; it is the design itself. The Restatement contemplates that outrageousness is not to be determined by evidence but according to the views of judges and jurors. Comment h to section 46 states:

It is for the court to determine, in the first instance, whether the defendant’s conduct may reasonably be regarded as so extreme and outrageous as to permit recovery, or whether it is necessarily so. Where reasonable men may differ, it is for the jury, subject to the control of the court, to determine whether, in the particular case, the conduct has been sufficiently extreme and outrageous to result in liability.

Comment h, which the plurality opinion attempts to follow in this ease without actually adopting it, leaves the respective roles of judges and juries ill-defined. Judges are to make the determination whether conduct is outrageous unless “reasonable men may differ”. But how does a judge know whether reasonable people would differ in their view of specific conduct? Are the parties to present affidavits of people who actually do differ? If so, then I suspect it will not be difficult to do so in almost every case, and a threshold judicial determination of outrageousness will be rare. May a judge take judicial notice that reasonable differences of opinion exist or do not exist? If so, then again, it would be a rare situation in which the issue should not be given to the jury to decide. May a judge rely simply on personal knowledge and experience? This is what comment h seems to intend, but to allow a judge to rule on the basis of personal views alone is not only broad discretion indeed but a violation of the rule of law. May special exceptions to pleadings be sustained or summary judgment granted solely because the judge believes that no reasonable person would consider the alleged conduct outrageous? How are such rulings to be reviewed on appeal? As in the trial court, the only standard available on appeal is the personal view of the judge.

Whether a judge or a jury decides the issue of outrageousness, neither has a standard for doing so. Without evidence or rules to guide a decision, both can resort only to their own views, and their own prejudices. Opinions about outrageous conduct may vary widely in various contexts. An employee may accuse his employer of behaving outrageously in terminating him, an insured may accuse his insurer of denying a claim outrageously, or a debtor may accuse a creditor of outrageous conduct in attempting to collect the debt. That class of persons to which each plaintiff or defendant belongs will almost certainly not share the same view of the alleged conduct as that class which includes the other party. To be sure, there will be a few instances when almost everyone is in accord, but in the vast majority of cases, creditors, for example, are less likely to be outraged by aggressive debt collection than debtors. What rule or standard can a judge or jury use to resolve such disputes? There is none. The composition of the jury, whether of bankers or borrowers, is likely to play heavily in the verdict delivered.

*633Ordinarily the law forbids factfinders from relying upon personal prejudices in deciding a case, but with this tort there is no alternative. “To suggest, as the Restatement does, that civil liability should turn on the resentments of the average member of the community appears to turn the passions of the moment into law.” Givelber, supra, at 52. Contrary to the assertions of some of my colleagues, I am certainly not critical of the ability of either judges or juries to make reasoned decisions, based upon the evidence and the law. My point is that we should not make judges or juries Rumpelstiltskins, spinning legal decisions out of personal attitudes. I am not aware of any other instance when a judge or juror is asked to decide the determinative issue in a case solely upon his personal views. How can such a decision be reviewed on appeal? There is no level of evidence which is legally or factually sufficient to establish the appropriate standard, for there is no evidence at all. The only issue for appeal is whether another set of judges has a different view of outra-geousness.

Outrageousness is thus different from other amorphous legal standards, such as ordinary care, the core concept of negligence.

The issue [in negligence cases] will frequently be whether there were alternative means of achieving the same end that would have eliminated the particular risk from which plaintiff’s injury resulted, and safety statutes and customs within the industry may provide explicit guidance on this point. Even where they do not (and certainly where they do) expert testimony may serve the function of identifying alternatives and placing a cost upon them. Even when the question cannot be intelligently viewed in terms of alternatives — e.g., where the defendant did not come to a full stop at the stop sign — there may be statutes that inform the decision. The issue of negligence is seldom decided without guidance from some external source: custom, relevant statutes and regulations, eviden-tiary doctrines such as res ipsa loquitur, or expert testimony on alternatives. In other words, a jury is rarely presented a story and asked to decide whether the defendant behaved reasonably simply by referring to its own sense of appropriate behavior. Yet this is what juries are asked to do with regard to the intentional infliction of emotional distress: there are no external standards for outrageousness comparable to those available in the negligence area.

Givelber, supra, at 56. Evidence is always available on the standard of care to which a reasonable person can be expected to adhere in a given situation; in many cases, such as professional malpractice, such' evidence is required. The jury is obliged to confine its deliberations to that evidence and cannot create its own standard of care. Thus, jurors who regularly drive in excess of the speed limit cannot equate ordinary care with their own conduct. But evidence of what is outrageous in a given situation is never available. Jurors are not only permitted to judge a defendant’s conduct by their own personal experiences; they are required to do so.

There are also moral overtones to outra-geousness that are absent in ordinary care. Conduct which negligently causes an accident is not “beyond all possible bounds of decency” and “utterly intolerable in a civilized community.” In a negligence action a jury must decide whether the defendant was careless; in an action for intentional or reckless infliction of emotional distress, the jury must decide whether the defendant was bad. A finding that conduct was outrageous is qualitatively different from a finding that it was negligent.

The Court’s sole response to all these criticisms is this: we trust juries to decide. This truism simply does not address the problems raised by the standardless tort the Court recognizes. The issue is not whether juries are qualified to decide fact issues — certainly they are — but what standards they are to employ in doing so. In a society ruled by law and not by men, we do not allow judges or juries to base adjudications on personal predilections. This is not a matter of trust, but of power properly exercised. We trust juries to judge accord*634ing to the law; we trust no one to judge according to his own views.

Because outrageousness depends almost entirely upon the views of the judge or jury in each case, it is difficult to know whether any particular conduct will give rise to liability. “[Ajssuming that there were a class of defendants who wanted to alter their behavior in order to conform to the dictates of this tort, the decided eases would not provide much clear instruction with respect to what activities are forbidden.” Givelber, supra, at 75. Plaintiffs also are given little firm indication whether particular conduct is outrageous. The result, as the experience of other jurisdictions indicates, is that claims for intentional or reckless infliction of emotional distress are asserted in a large number of cases in the sincere but usually unrealized anticipation that a judge or jury may view them favorably. See Annotation, supra. The burden of so many meritless claims on the judicial system is significant; the Court simply assumes that this burden is outweighed by the benefits of such claims.

In his separate opinion, Chief Justice Phillips calls the tort of intentional infliction of emotional distress “a convenient mechanism for achieving a just result” without the bother of having to apply rules to facts. Ante at 627. This candid assessment of the tort exposes in a few words its faults. The rules governing the tort are so broad they hardly matter. The entire process consists of presenting a factual situation to a jury and asking their personal views. This is not justice.

It is a mistake, in my view, to adopt a tort as vaguely defined as intentional infliction of emotional distress. Tort law develops most effectively by specifying discrete categories of conduct which are actionable, not by adopting broad prohibitions. If this were not so, the law could be much simplified by replacing all the various torts that have been recognized and with a single cause of action for wrongful infliction of recoverable damages, or WIRD for short. All tort actions could simply be WIRD. We could dispense with rules and their application; we could simply ask a jury whether, based upon what they had heard, the defendant’s conduct was WIRD, and what damages should be awarded. This ought to be as ridiculous as it sounds, but the tort the Court adopts today is not too far removed from WIRD. If tort law could be reduced to its single basic jurisprudential foundation — a sort of unified field theory of torts — it would not be very useful in deciding particular cases. Workable rules require more detail.

II

Recognition of a general tort of intentional or reckless infliction of emotional distress does not require that it be allowed in all contexts. As the plurality opinion notes, the high courts in a great majority of states have recognized this general tort. Ante at 621. Indeed, the plurality opinion bases its holding on this fact. The plurality may believe that because interspousal tort immunity has been abolished in Texas, Bounds v. Caudle, 560 S.W.2d 925 (Tex. 1977), Price v. Price, 732 S.W.2d 316 (Tex. 1987), recognition of the tort generally entails that it may be asserted between spouses. However, interspousal tort immunity has been abolished in all the states which recognize the tort of intentional infliction of emotional distress, yet only two have allowed spouses to assert the action against each other. Henriksen v. Cameron, 622 A.2d 1135 (Me. 1993); Davis v. Bostick, 282 Or. 667, 580 P.2d 544 (1978). Two other state high courts have refused to permit the tort between spouses. Weicker v. Weicker, 22 N.Y.2d 8, 290 N.Y.S.2d 732, 237 N.E.2d 876 (1968) (per curiam); Pickering v. Pickering, 434 N.W.2d 758 (S.D. 1989). This fact, which the plurality opinion does not deny, casts its rationale in self-conflict. On the one hand, the principal reason the plurality opinion recognizes intentional infliction of emotional distress is its adoption in a vast number of states. Yet the plurality opinion allows the tort in the marital context despite the fact that only two other states would do so. The plurality opinion states that it knows of no legal impediment to such a tort. Actually, there are several.

*635Texas law does not permit a divorce court to divest a spouse of separate property in dividing the marital estate. Eggemeyer v. Eggemeyer, 554 S.W.2d 137 (Tex. 1977). Allowing a tort action between spouses in connection with the dissolution of the marriage circumvents this bar. Although the Court does not address this issue, the damages assessed against a spouse for intentional infliction of emotional distress should be paid out of that spouse’s share of the community or his or her separate property. Damages could not be paid out of the balance of the community without punishing the aggrieved spouse. I assume the Court intends that both compensatory and punitive damages may be awarded for the tort, as for other intentional torts. Damages paid from one spouse’s separate property awards the other spouse more than would otherwise be allowed in a divorce action. While it is already possible for one spouse to recover damages from the other’s separate property for an intentional tort, such as battery, the adoption of intentional or reckless infliction of emotional distress greatly expands the circumstances in which such recovery may be had, further distorting the basic community property system.

Awarding one spouse tort damages against the other in the context of divorce also creates a problem of double recovery. The fault of a spouse contributing to the breakup of a marriage may be considered by the divorce court in dividing the marital estate. Murff v. Murff, 615 S.W.2d 696 (Tex. 1981). But the aggrieved spouse should not be awarded both tort damages and a disproportionate share of the community for the same conduct of the other spouse. It is difficult to imagine how wrongful conduct which contributes to the breakup of the marriage can be segregated from that which is tortious to prevent double recovery. If the tort claim is tried to the court, the judge may attempt to separate the property division from the damages award. But if the tort claim is tried to a jury, who does not have authority to divide the property, it is impossible for either the judge or the jury to take into account the other’s actions. The judge cannot instruct the jury that he has decided to make an uneven division of the property on account of fault without commenting on the weight of the evidence in the tort trial; nor can the judge determine from a jury verdict what conduct should not be considered in dividing the property. The judge may request an advisory verdict on the division of property and instruct the jury to consider tort damages separately, but even when this is feasible, the trial court must follow the advisory verdict or the effect of the instruction will be lost. In sum, allowance of a tort of intentional or reckless infliction of emotional distress in a divorce proceeding creates serious trial management problems which affect the substantive rights of the parties.

There can be little doubt that both spouses in a great many divorce cases will allege intentional or reckless infliction of emotional distress against each other. The Family Law Section of the State Bar of Texas, in an amicus brief, endorses this assessment. Whenever the tort is alleged, the parties will be entitled to a jury trial. At present, the right to a jury trial in divorce cases is very limited. The likely increase in the number of jury trials in such cases can only delay the resolution of divorce cases in many courts. The Court does not consider this potential impact of its ruling.

Allowing a tort action between spouses in divorce also presents the problem of whether attorney’s fees may be charged on a contingent basis. Ordinarily, an attorney may not charge a contingent fee in a divorce proceeding, but may do so in a tort action. When the actions are combined, the alleged fault giving rise to the tort claim will also affect other issues in the divorce proceeding, including property division, custody, and support. Segregation of the time an attorney spends on these matters is virtually impossible. As a practical matter, contingent fees must either become the principal compensation for attorneys in divorce eases involving tort claims, or else they must be disallowed in such situations altogether.

The standard of outrageousness is certainly no easier to apply in the marital *636context than in other contexts, as the facts of this case illustrate. Sheila Twyman’s claim of intentional infliction of emotional distress is based upon the following testimony at trial, which was mostly undisputed. William, a Navy pilot, and Sheila, a college graduate with a degree in nursing, were married in 1969. In 1975, on two or three occasions at William’s suggestion, the couple engaged in what they referred to as “light bondage” — tying each other to the bed with neckties during their sexual relations. Sheila testified that William did not force her to participate in these activities. After the last occasion Sheila told William she did not like this activity and did not want to participate in it further. She revealed to him that she associated the activities with the horrible experience of having been raped at knifepoint earlier in her life. William never again suggested that she engage in the activities, nor was the subject discussed again for ten years. In 1985 Sheila inadvertently discovered that William was consulting with a psychologist. When she asked him why, he told her that he was involved with another woman. William told Sheila that if she could only have done bondage, nothing else would have mattered. For the remainder of the year the couple sought counseling. At times during this period William made derogatory remarks to Sheila about her sexual ability, comparing her to his girl friend. On their counselor’s advice, William and Sheila discussed William’s bondage fantasies, and Sheila again tried to participate in bondage activities with William. But she found the activity so painful and humiliating that she could not continue it. Their last encounter, which did not include bondage activities, was so rough that she was injured to the point of bleeding. At one point Sheila was distressed to discover that their ten-year-old son had found magazines William kept hidden, which portrayed sadomasochistic activities. Eleven months after she first learned of William’s affair, Sheila separated from him and filed for divorce. Throughout that period, Sheila testified, she experienced utter despair, devastation, pain, humiliation and weight loss because of William’s affair and her feelings that the marriage could have survived if only she had engaged in bondage activities.

To recover damages Sheila must prove that William’s conduct was outrageous— that is, “extreme”, “beyond all possible bounds of decency”, “atrocious”, and “utterly intolerable in a civilized community.” Although outrageousness is, according to the plurality opinion and the Restatement, a question for the court in the first instance, this Court refuses to say whether William’s conduct was or was not outrageous. If it was not, as a matter of law, then there is no need to remand this case for further proceedings. If William’s conduct was outrageous, or if that issue must be decided by a jury, then it is unclear what components of the conflict between Sheila and William were actionable. There is no question from the record that Sheila claims to have suffered bitterly, but there appear to have been three causes: William’s affair, his interest in bondage, and the breakup of the marriage. If the first or last causes constitute outrageous behavior, then there a tort claim may be urged successfully in most divorces. Allowing recovery based upon the first cause of Sheila’s emotional distress is simply to revive the old action for alienation of affections abolished by the Legislature. Tex. Fam.Code § 4.06. I doubt whether the Court intends this result. If William’s outrageous conduct was attempting to interest Sheila in sexual conduct which he considered enjoyable but she, in her words, “did not like”, then again, this tort may be very broad indeed.

The sexual relationship is among the most intimate aspects of marriage. People’s concepts of a beneficial sexual relationship vary widely, and spouses may expect that some accommodation of each other’s feelings will be necessary for their mutual good. Any breach of such an intimate and essential part of marriage may be regarded as outrageous by the aggrieved spouse and will often be the cause of great distress. There are many other aspects of marriage which are likewise sensitive. How money is to be spent, how children are *637to be raised, and how time is to be allocated are only a few of the many areas of conflict in a marriage. Not infrequently disagreements over these matters are deep and contribute to the breakup of the marriage. If all are actionable, then tort claims will be commonplace in divorce cases, and judges and juries with their own deeply felt beliefs about what is proper in a marital relationship will face the hard task of deciding whether one spouse or another behaved outrageously with no standards but their own to guide.

The inquiry which must be made to determine whether a spouse’s conduct is outrageous entails too great an intrusion into the marital relationship. Although courts are already called upon to consider fault in divorce actions, allowance of tort claims requires a more pervasive inspection of spouses’ private lives than should be permissible. In this case the parties were called to testify in detail and at length about the most private moments of their marriage. If the court’s only concern were the degree to which a spouse’s fault had contributed to the demise of the marriage, the inquiry into each spouse’s conduct need not have been so detailed. To recover damages, however, Sheila was required to testify at length before a jury, and to rebut her claim, William was obliged to answer in equal detail. The prospect of such testimony in many divorces is too great an invasion of spouses’ interests in privacy, and promises to make divorce more acrimonious and injurious than it already is.

The plurality opinion’s justification for allowing the tort of intentional infliction of emotional distress between spouses is that this represents a middle ground among the various positions taken by Members of this Court. But being in the middle does not equate to being right. Certainly the Court is not in the middle of the views of other state courts; rather, it is to one extreme.

Ill

. I must add a word about the Court’s lack of restraint in deciding a very important issue without requiring that it be raised in the lower courts and argued by the parties. As the plurality opinion acknowledges, Sheila Twyman did not specifically plead intentional infliction of emotional distress. I am willing to concede that Sheila’s pleadings may be read broadly to state such a claim. However, as the plurality opinion also acknowledges, the trial court was not requested to, and did not, make findings to support recovery on any cause of action for intentional infliction of emotional distress. Sheila did not argue in the trial court and has not argued on appeal that such a cause of action exists or that the judgment in her favor can be supported on that theory; she has argued only that she is entitled to recover on her cause of action for negligent infliction of emotional distress. The court of appeals agreed with her argument and did hot consider the subject of intentional infliction of emotional distress. 790 S.W.2d 819. The Court reverses the judgment of the court of appeals, reaffirming its decision of a few weeks ago that no cause of action for negligent infliction of emotional distress exists in Texas. Boyles v. Kerr, 855 S.W.2d 593 (1993). But the Court goes much further and holds that a cause of action exists for intentional infliction of emotional distress generally, and that it may be asserted without restriction between spouses.

The Court does not conclude that the evidence in this case demonstrates an intentional infliction of emotional distress; the plurality opinion suggests only that “it appears that the facts when developed on retrial may support recovery”. Ante at 626 (emphasis added). It is unusual for a court to recognize a cause of action for the first time in a case in which no party argues for the action and the evidence at trial does not support it. Today’s decision might as well issue in a law review article based upon hypothetical situations or decisions in other jurisdictions as in an opinion in this case; its holding is more an abstract statement of law than a determination of the case before it. The Court’s failure to apply its new legal principle to the evidence in this case leaves the parties and the trial court with little guidance in future proceedings and, if the case is not settled, the likelihood of more appeals raising the same *638points. Although the parties testified fully at trial, and the plurality opinion holds that whether conduct is sufficiently outrageous to be actionable is, at least at the threshold, a legal question, the Court gives no hint whether William’s conduct in this case meets the threshold legal test. The Court simply advises Sheila that she may have a cause of action and affords her an opportunity to assert it.

Many others equally affected by the Court’s decision today have even greater cause for concern. The Family Law Section of the State Bar of Texas has filed a brief in this Court as amicus curiae because of the “extremely significant policy questions involved in this case” and the “potential impact of the Court’s decision”. Brief of the Section at 1-2. Unlike Sheila’s brief, the Section’s brief does discuss whether a cause of action for intentional infliction of emotional distress should be recognized “because it involves similar policy issues” which inform the decision whether a cause of action for negligent infliction exists. The Section’s brief provides us a thorough and authoritative analysis of the arguments for and against causes of action for intentional and negligent infliction of emotional distress between spouses, the brief urges “caution in making the determination of what claims spouses can assert against each other in a divorce.” Id. at 49. It is precisely this counsel from the segment of the bar most knowledgeable about the issue the Court decides and most affected by it that the Court wholly ignores.

Neither the Family Law Section nor anyone else in this case has attempted to assess the impact of recognizing a general tort of intentional infliction of emotional distress in other contexts. I believe it is most unwise to render so sweeping a decision in the circumstances of this case.

IV

Finally, I must say a word in response to Justice SpeCTOR’s dissenting opinion, the principal thesis of which is that the Court has denied recovery for negligent infliction of emotional distress in this case because of an institutional bias against women. “It is no coincidence”, JUSTICE Spector contends, that both this case and Boyles v. Kerr, 855 S.W.2d 593 (Tex. 1993), involve claims for emotional distress by women against men. Post at 642. Actually, it is just that: a coincidence.

It is a further coincidence that another case we decide today, Valenzuela v. Aquino, 853 S.W.2d 512 (Tex. 1993), in which we also deny recovery for negligent infliction of emotional distress, does not involve claims solely by women against men. That case involves a claim by Dr. Aquino and his family for damages caused by anti-abortion picketers demonstrating in front of his home. Our refusal to allow recovery for negligent infliction of emotional distress in Valenzuela contradicts Justice Spector’s assertions of bias. Accordingly, Justice Spector does not attempt to argue that our decision in Valenzuela is motivated by prejudice in favor of abortion or against doctors, or by any other bias. There is no more basis for the assertion that the views against recognition of the tort in the present context are born of a latent antagonism to women.

Further, Justice Spector’s assertions are somewhat antagonistic toward one another. She states: “In the judicial system dominated by men, emotional distress claims have historically been marginalized.” Post at 642. She also claims, however: “From the beginning, tort recovery for infliction of emotional distress has developed primarily as a means of compensating women for injuries inflicted by men insensitive to the harm caused by their conduct.” Post at 642. It is not entirely clear how the justice system can have developed a tort primarily to compensate claims by women at the same time that the system was dominated by men intent upon marginalizing women’s claims. Justice Spector herself states that “[njeither the Fort Worth Court of Appeals, nor any of the other courts at the time [in 1918] were primarily concerned with protecting women’s rights.” Post at 643. If, as I will concede, the justice system has been historically dominated by men, and if, as I am willing to assume, those men were not always sympathetic to *639women’s claims, then development of theories for recovering damages for emotional distress cannot have been due primarily to a desire to compensate women. There must be other factors which better explain the development of the law.

There is evidence of such factors. It cannot be denied that differentiations have been drawn between the emotional distress suffered by men and that suffered by women. Some early cases and commentaries explicitly recognized the distinction between a female plaintiff and a male plaintiff suffering a similar injury, and commentators indicated that the gender of the plaintiff is one of the relevant factors in determining liability. See, e.g., Fort Worth & Rio Grande Ry. Co. v. Bryant, 210 S.W. 556 (Tex.Civ.App.-Fort Worth 1918) (in which a daughter recovered for exposure to coarse language in a train depot, but her father did not);3 Calvert Magruder, Mental and Emotional Disturbance in the Law of Torts, 49 Harv.L.Rev. 1083, 1046 (1936) (suggesting that the plaintiff’s gender should play a part in the determination of defendant’s liability). Prosser stated the idea clearly: “There is a difference between violent and vile profanity addressed to a lady, and the same language to a Butte miner and a United States marine.” William L. Prosser, Intentional Infliction of Mental Suffering: A New Tort, 37 Mich.L.Rev. 874, 887 (1939) (footnote omitted). This language was modified to reflect the “eggshell plaintiff” concept in the comment c to section 48 of the Restatement, which states that “language addressed to a pregnant or sick woman may be actionable where the same words would not be if they were addressed to a United States Marine.”

One may argue that these attempted differentiations of emotional distress reflect a patronizing view of women.4 Or one may also argue that the differentiations were drawn without particular regard for gender. But there is no evidence that ensuring recovery for uniquely female claims, because of any uniquely female characteristics, has been a primary factor in recognizing recovery for such injuries. The earliest cases to hold defendants liable for the infliction of mental suffering involved common carriers and innkeepers who were held to have breached an implied contract to be polite. W. Page Keeton, et al., Prosser and Keeton on the Law of Torts, § 12, at 57 (5th ed. 1984). Neither the cases nor the commentaries gives any indication that women more frequently pursued this cause of action. See, Annotation, Right to Recover for Mental Pain and Anguish Alone, Apart from other Damages, 23 A.L.R. 361 (1923), and cases cited therein. (In one of the cases awarding punitive damages to a male plaintiff, the facts showed that the conductor of a railroad had permitted the train to become overcrowded without informing the passengers that some would have to stand. The conductor intentionally humiliated a passenger who refused to give up his ticket before being provided an alternate seat by saying that a lady would be asked to give up her seat to him. Cave v. Seaboard Air Line Ry., 94 S.C. 282, 77 S.E. 1017 (1913).)

In the first successful cases against defendants other than common carriers, the defendants’ actions were found to have caused physical injuries as a result of emotional distress in the absence of physical impact. See John E. Hallen, Hill v. Kimball-A Milepost in the Law, 12 Tex. L.Rev. 1, 7-8 (1933). The cases frequently involved female plaintiffs, and Prosser described the pattern that quickly emerged as follows:

*640Nearly all of the plaintiffs have been women, usually in that delicate condition whose standardized consequences have typified mental suffering cases with the “customary miscarriage.”

William L. Prosser, Intentional Infliction of Mental Suffering: A New Tort, 37 Mich.L.Rev. 874, 888 (1939); see also Hubert Winston Smith, Relation of Emotions to Injury and Disease: Legal Liability for Psychic Stimuli, 30 Va.L.Rev. 193, at Appendix B (1944) (listing a large number of cases involving miscarriage). No reason is suggested for the large number of cases involving miscarriages either by Prosser or other commentators that identify the pattern. Prosser, Intentional Infliction of Emotional Distress, at 888 & n. 81; Green, ‘Fright Cases, ’ 27 Ill.L.Rev. 761, (1933). However, it is possible that the reason for the frequency of such cases was the requirement of a physical manifestation of injury to recover for emotional distress. Thus the fact that the majority of the cases involved women plaintiffs does not reveal anything except the fact that men do not miscarry.

Although miscarriage was the most common type of injury for which recovery was awarded in the early cases, other physical manifestations of emotional harm gave rise to recovery. See Hubert Winston Smith, Relation of Emotions to Injury and Disease: Legal Liability for Psychic Stimuli, 30 Va.L.Rev. 193, at Appendix B (1944) (categorizing over 300 cases by type of injury alleged). Contrary to the dissent’s suggestion, however, the reasoning supporting recovery in these cases did not develop “[fjrom the beginning” to provide “a means of compensating women for injuries inflicted by men insensitive to the harm caused by their conduct.” Post at 642. Rather, the cases permitted recovery for physical injury resulting from the negligent actions of the defendants, regardless of the fact that the injury resulted from emotional trauma. The duty imposed in these cases was a duty not to inflict foreseeable physical injury by whatever means; the early cases did not impose a duty not to inflict the emotional trauma itself. See Boyles v. Kerr, 855 S.W.2d 593, 599 & n. 4 (Tex. 1993).

Even from this very abbreviated overview, to the extent that any pattern can be discerned in this history, it is not one of a struggle for recognition of women’s rights, but of either a condescending and patronizing view of women, or a development of the law without particular regard for gender. Arguments for allowance of claims for emotional distress owe their support to far more factors than the relationships between men and women.5 The tort the Court recognizes today will have marked impact on marital relationships, but it will have far broader impact on the many other relationships it will affect. To assert, as Justice SpectoR does, that the principal effect of today’s decision falls upon women, overlooks the broader reality which, as I have explained, is the basis for my concern.

[[Image here]]

For all the foregoing reasons, I dissent from the opinion and judgment of the Court.

Justice ENOCH joins in this concurring and dissenting opinion.

dissenting.

Over five years ago, a trial court issued a divorce decree that included an award to Sheila Twyman of $15,000 for the years of abuse she had suffered at the hands of her husband. At the time, the award was consistent with prevailing Texas law. Today, the plurality sets aside the trial court’s award and sends Sheila Twyman back to start the process over in a new trial. Because justice for Sheila Twyman has been both delayed and denied, I dissent.

*641I.

At trial, Sheila testified that her husband, William Twyman, introduced bondage activities into their relationship after their marriage. Sheila told William that she could not endure these activities because of the trauma of having been raped several years earlier. She also informed William that she had been cut with a knife during the rape, and had been placed in fear for her life. Although William understood that Sheila equated bondage with her prior experience of being raped, he told Sheila that if she would not satisfy his desires by engaging in bondage, there would be no future to their marriage.

As a result, Sheila experienced “utter despair” and “devastation,” as well as physical problems — weight loss and, after one encounter, prolonged bleeding that necessitated gynecological treatment. The pain and humiliation of the bondage activity caused her to seek help from three professional counselors.

The trial court found that William “engaged in a continuing course of conduct of attempting to coerce [Sheila] to join in his practices of ‘bondage’ by continually asserting that [their] marriage could be saved only by [Sheila] participating with [William] in his practices of ‘bondage.’ ” The trial court also determined that Sheila’s suffering was certainly foreseeable from William’s continuing course of conduct, “in light of his existing knowledge of her long-existing emotional state, which was caused by her having been forcibly raped prior to their marriage.” Finally, the trial court found that Sheila’s mental anguish was a direct proximate result of William’s sexual practices.

Based on the pleadings, evidence, and arguments, the trial court concluded that the facts and the law supported Sheila’s recovery of $15,000 for William’s negligent infliction of emotional distress. The court of appeals, in an opinion by Justice Gam-mage, affirmed the trial court’s judgment under prevailing tort law and noted that this court had expressly approved the recovery of damages on a negligence claim in a divorce action. 790 S.W.2d 819, 823 (citing Price v. Price, 732 S.W.2d 316 (Tex. 1987)).

This court, however, has now rejected Texas law established to provide redress for injuries of the kind inflicted by William Twyman. While allowing some tort claims to be brought in a divorce action, the plurality forbids recovery for negligent infliction of emotional distress, and insists that Sheila Twyman proceed on a theory of intentional infliction of emotional distress.

II.

Today’s decision is handed down contemporaneously with the overruling of the motion for rehearing in Boyles v. Kerr, 855 S.W.2d 593 (Tex. 1993), in which this court reversed a judgment in favor of a woman who was surreptitiously videotaped during intercourse, then subjected to humiliation and ridicule when the tape was displayed to others. In Boyles, as in this case, a majority of this court has determined that severe, negligently-inflicted emotional distress does not warrant judicial relief — no matter how intolerable the injurious conduct. The reasoning originally articulated in Boyles, and now implied in this case, is that “[t]ort law cannot and should not attempt to provide redress for every instance of rude, insensitive, or distasteful behavior”; providing such relief, the Boyles majority explained, “would dignify most disputes far beyond their social importance.” 36 Tex. S.CtJ. 231, 233-234 (Dec. 2, 1992).1

Neither of these cases involves “rude, insensitive, or distasteful behavior”; they involve grossly offensive conduct that was appropriately found to warrant judicial relief. The decision in Boyles overturns well-reasoned case law, and I strongly agree with the dissenting opinion in that case. For the same reasons, I strongly disagree with the plurality here; the rule embodied *642in Boyles is no less objectionable when applied to the facts of this case. Sheila Twyman is entitled to recover the amount awarded by the trial court for the injuries inflicted by her husband.

III.

It is no coincidence that both this cause and Boyles involve serious emotional distress claims asserted by women against men. From the beginning, tort recovery for infliction of emotional distress has developed primarily as a means of compensating women for injuries inflicted by men insensitive to the harm caused by their conduct. In “[t]he leading case which broke through the shackles,”2 a man amused himself by falsely informing a woman that her husband had been gravely injured, causing a serious and permanent shock to her nervous system. Wilkinson v. Downton, 2 Q.B.D. 57 (1897). Similarly, in the watershed Texas case, a man severely beat two others in the presence of a pregnant woman, who suffered a miscarriage as a result of her emotional distress. Hill v. Kimball, 76 Tex. 210, 13 S.W. 59 (1890). By World War II, the pattern was well-established: one survey of psychic injury claims found that the ratio of female to male plaintiffs was five to one. Hubert Winston Smith, Relation of Emotions to Injury and Disease: Legal Liability for Psychic Stimuli, 30 Va.L.Rev. 193 (1944).

Even today, when emotional distress claims by both sexes have become more widely accepted, women’s claims against men predominate. Of the thirty-four Texas cases cited by the plurality — all decided since 1987 — women’s claims outnumbered men’s by a ratio of five to four; 3 and only four of the thirty-four involved any female defendants.4 Of those cases involving relations between two individuals — with no corporations involved — five involved a woman’s claim against a man;5 none involved a man’s claim against a woman.

I do not argue that women alone have an interest in recovery for emotional distress. However, since the overwhelming majority of emotional distress claims have arisen from harmful conduct by men, rather than women, I do argue that men have had a disproportionate interest in downplaying such claims.

Like the struggle for women’s rights, the movement toward recovery for emotional distress has been long and tortuous. See Peter A. Bell, The Bell Tolls: Toward Full Tort Recovery for Psychic Injury, 36 U.Fla.L.Rev. 333, 336-40 (1984). In the judicial system dominated by men, emotional distress claims have historically been marginalized:

The law of torts values physical security and property more highly than emotional security and human relationships. This apparently gender-neutral hierarchy of values has privileged men, as the traditional owners and managers of property, and has burdened women, to whom the emotional work of maintaining human relationships has commonly been assigned. The law has often failed to compensate women for recurring harms— serious though they may be in the lives *643of women — for which there is no precise masculine analogue.

Martha Chamabas and Linda K. Kerber, Women, Mothers, and the Law of Fright: A History, 88 Mich.L.Rev. 814 (1990). Even Prosser recognizes the role of gender in the historical treatment of claims like that involved in Hill v. Kimball:

It is not difficult to discover in the earlier opinions a distinctly masculine astonishment that any woman should ever allow herself to be frightened or shocked into a miscarriage.

W. Page Keeton et al., Prosser and Keeton on the Law of Torts § 12, at 55-56 (5th ed. 1984).

Displaying a comparable “masculine astonishment,” the dissenting opinion by Justice Hecht insists that, with a few possible exceptions, women have played no distinct part in the development of tort recovery for emotional distress. As a general matter, Justice Hecht questions how a legal system dominated by men could develop a tort to compensate women even while marginalizing women’s claims. The answer is amply illustrated by the present case: to provide some appearance of relief for Sheila Twyman, the court recognizes the tort of intentional infliction of emotional distress; but in doing so, it restricts her to a theory which, as Justice Hecht observes, is “seldom successful.” 855 S.W.2d at 631.

Justice Hecht acknowledges that in the early cases, recovery for emotional distress “frequently involved female plaintiffs.” 855 S.W.2d at 631. However, rather than viewing this phenomenon as an indication of actual, serious injuries, Justice Hecht suggests that it may have been due to a patronizing attitude on the part of the courts.

There is little doubt that some of the case law in this area, as in any other, reflects a patronizing view of women. More often, though, the case law reflects the logical application of existing law to a wide range of claims. For example, in the only case cited by Justice Hecht to illustrate an arguably patronizing view of women, there was evidence that men employed by a railroad had humiliated a man’s ten-year-old daughter by subjecting her to obscene language; but there was no evidence that the language had humiliated the father. Fort Worth & Rio Grande Ry. Co. v. Bryant, 210 S.W. 556 (Tex.Civ.App.—Fort Worth 1918, writ ref’d). There is nothing patronizing about holding a railroad company responsible for the harm caused by its employees’ conduct.

I would group Bryant with the many other common carrier cases that were decided, in Justice Hecht’s terms, “without particular regard for gender.” 855 S.W.2d at 639. Neither the Fort Worth Court of Appeals, nor any of the other courts at the time were primarily concerned with protecting women’s rights. But in Bryant, as in so many of the other cases, the evolution of the law regarding emotional distress claims did enable a female to recover for emotional harm inflicted by men. This fact does not reflect a charitable desire to help women; it reflects the fact that the serious emotional distress claims usually involved injuries inflicted by men upon women.

Given this history, the plurality’s emphatic rejection of infliction of emotional distress claims based on negligence is especially troubling. Today, when the widespread mistreatment of women is being documented throughout the country — for instance, in the areas of sexual harassment 6 and domestic violence7 — a majority of this court takes a step backward and abolishes one way of righting this grievous wrong.

*644IV.

Rather than dismissing Sheila’s claim outright, the plurality remands this cause to the trial court to allow Sheila to seek recovery under an alternative theory. I agree that Sheila is entitled to pursue a claim based upon intentional infliction of emotional distress, as set out in section 46 of the Restatement (2d) of Torts.

However, in restricting recovery for emotional distress to claims based upon intent, the plurality, joined by Justice Gonzalez’s concurring opinion, demonstrates a basic misunderstanding of claims like those presented by Susan Kerr and Sheila Twyman. While recognizing that recovery should be allowed for conduct intended to inflict injury, the plurality fails to acknowledge the severe emotional distress often caused unintentionally.

This court has previously made clear that the distinguishing feature of an intentional tort is “the specific intent to inflict injury.” Reed Tool Co. v. Copelin, 689 S.W.2d 404, 406 (Tex. 1985) (citing Restatement (2d) of Torts § 8A (1965)); see also Rodriguez v. Naylor Indus., 763 S.W.2d 411, 412 (Tex. 1988). This definition of an “intentional” injury is echoed in the portion of the Restatement governing intentional infliction of emotional distress:

The rule stated in this Section applies where the actor desires to inflict severe emotional distress, and also where he knows that such distress is certain, or substantially certain, to result from his conduct.

Restatement (2d) of Torts § 46 cmt. i (1965).

Unfortunately, in many cases, severe emotional distress is caused by an actor who does not actually desire to inflict severe emotional distress, and who is even oblivious to the fact that such distress is certain, or substantially certain, to result from his conduct. It may well be the case, for example, that William Twyman never actually intended to inflict emotional distress upon Sheila, and never expected the injury that his conduct caused. Rather, he may have insisted on bondage activities solely for the purpose of satisfying his own desires. Similarly, Dan Boyles may have videotaped his activities with Susan Kerr not for the purpose of injuring her, but rather for the purpose of amusing himself and his friends.

I do not argue, as the plurality asserts, that “the emotional harm William caused was foreseeable but not substantially certain to occur.” 855 S.W.2d at 624. I do argue, though, that Sheila’s recovery for William’s conduct should not depend upon proof of William’s sensitivity. To apply a standard based on intent is to excuse William’s conduct so long as he believed his actions were harmless.

Brutish behavior that causes severe injury, even though unintentionally, should not be trivialized. Foreclosing recovery for such behavior may prevent litigation of frivolous claims; but it also denies redress in exactly those instances where it is most needed.

V.

While the plurality would allow some possibility of recovery for injuries like Sheila Twyman’s, the dissenting opinions by Chief Justice Phillips and Justice Hecht would allow none at all. Adopting the medieval view of marital relations, Chief Justice Phillips argues that spouses should be shielded from liability for even the most outrageous acts against one another. This view echoes William Twyman’s assertion at trial that, by consenting to marriage, Sheila Twyman assumed the risk of physical injury and emotional harm. Fortunately, in Texas, this archaic view has been soundly rejected; interspousal immunity has been abolished “completely as to any cause of action.” Price v. Price, 732 S.W.2d 316, 320 (Tex. 1987). Insulating spouses from liability, we have noted, “would amount to a repudiation of the constitutional guarantee of equal protection of the laws.” Id. Thus, recovery for intentional infliction of emotional distress should be available to spouses and non-spouses alike, as other states have recognized.8

*645Justice Hecht not only agrees that spouses should have special protection from liability, but further argues that recovery for intentional infliction of emotional distress should never be allowed in any case. In Justice Hecht’s view, the tort set out in section 46 of the Restatement is “too broad a rubric to describe actionable conduct, as this case illustrates.” 855 S.W.2d at 680. Cases from the forty-six jurisdictions that recognize this tort comprise a unified body of law that would suggest otherwise. As Justice Hecht acknowledges, claims under section 46 are “seldom successful”; defendants have been held subject to liability only in those instances in which the defendant’s conduct was clearly “beyond the bounds of decency.”9 Unlike Justice Hecht, I believe the judicial system is fully capable of distinguishing trivial acts from those acts that are sufficiently outrageous to warrant relief.

VI.

The claim asserted by Sheila Twyman was based on a simple premise: her husband should be held accountable for the foreseeable consequences of his conduct. The courts below, in applying the law, understood the nature and extent of such conduct; the plurality does not. Tragically, the lack of understanding shown today will only lead to more delay and more injustice.

Justice DOGGETT joins in this dissenting opinion.

concurring.

What happened to Sheila Twyman in this case involves grossly offensive conduct which warrants judicial relief and the length of time it took this Court to decide this case is regrettable. However, the actions of William Twyman in engaging in bondage activities with Sheila Twyman, under the rationale that such activities were necessary to the future of their marriage, were all intentional in nature. None of William Twyman’s actions could be in any way described as negligent, or careless, or accidental.

After Twyman was granted there was great debate on and off this Court as to the scope of St. Elizabeth Hospital v. Garrard, 730 S.W.2d 649 (Tex. 1987). The debate centered on whether the Court in this 5-4 opinion created an all purpose, free standing tort. Also, divorce was only one factual scenario in which we were reviewing this tort. Almost contemporaneously with Twyman, the Court was reviewing this tort in a child abduction context, Weirich v. Weirich, 833 S.W.2d 942 (Tex. 1992); as an independent tort, Boyles v. Kerr, 855 S.W.2d 593, 599 n. 3 (Tex. 1993); and in the free speech context, Valenzuela v. Aquino, 853 S.W.2d 512, 514 (Tex. 1993). Harmon v. Grande Tire Co., 821 F.2d 252, 258 (5th Cir. 1987), decided after Garrard, stated “[wjhile the majority opinion in Garrard is broadly written, we read it as being directed only to the matter of proof.” See also In re Air Crash at Dallas/Ft. Worth Airport, 856 F.2d 28 (5th Cir. 1988); Fiorenza v. First City Bank-Central, 710 F.Supp. 1104, 1105 (E.D.Tex. 1988).

Therefore, for the reasons stated in my concurring opinion on motion for rehearing in Boyles v. Kerr, 855 S.W.2d 593, 603 (Tex. 1993), I agree with the judgment of the Court.