I.A Basic background

This brief stage-setting section rehearses a few elements of basic tort doctrine, relying, for convenience, on the Restatement (Third) of Torts: Liability for Physical and Emotional Harm (hereafter, ‘the Restatement’). The Restatement, a partly prescriptive summary of tort law written by legal experts representing a non-governmental organization, is not legally binding or undisputed.Footnote ¹⁰ That said, a short overview of a few of its provisions efficiently indicates how tort distinguishes between physical and emotional harm, and how it generally makes it more difficult for plaintiffs to recover for the latter.

The Restatement distinguishes between ‘physical harm’ and ‘emotional harm’ (its very subtitle is “Liability for Physical and Emotional Harm”). It defines ‘physical harm’ as “the physical impairment of the human body (‘bodily harm’) or of real property or tangible personal property (‘property damage’).”Footnote ¹¹ It defines ‘emotional harm’ as “impairment or injury to a person’s emotional tranquility.”Footnote ¹² For the Restatement, then, bodily harm constitutes only a sub-set of physical harm. This is noteworthy because it suggests that, in distinguishing between physical and emotional harm, the Restatement’s intent is less to single out emotional harm and more to draw lines between harms that are more and less easily verifiable.

Like many other areas of law, tort generally makes it more difficult for plaintiffs to recover for emotional harm.Footnote ¹³ Although tort has been moving in the direction of liberalizing recovery for emotional harm since at least the late nineteenth century, it continues to impose special limits on recovery for this type of harm.Footnote ¹⁴ Section I.B below considers some of these limits in detail. For now, however, consider three of the limits as the Restatement describes them.

First, Section 5 of the Restatement (“Liability for Intentional Physical Harm”) explains that, as one moves from physical to non-physical harm, a showing of intent to cause harm becomes less likely to suffice to state a claim:

Interestingly, that the first of this section’s two examples involves loss of prospective economic advantage suggests that the distinction in which the Restatement is interested here is ‘physical-non-physical,’ not ‘physical-emotional.’ A second way in which tort imposes special limits on recovery for emotional harm appears in the Restatement’s discussion of the cause of action of intentional infliction of emotional distress:

This section limits liability to “severe” harm that results from “extreme and outrageous conduct.” A third way in which tort limits recovery for emotional harm is visible in the Restatement’s discussion of NIED:

This section limits liability by requiring that the harm be “serious” and that the defendant’s behavior either (a) create a risk of physical harm, or (b) occur in an emotionally delicate context. Taken together, the preceding Restatement sections provide a snapshot of some of the ways in which the physical-emotional distinction shapes contemporary tort doctrine.

I.B Using tort’s physical-emotional distinction to draw lines between legally cognizable and non-cognizable harm

Several legal scholars have recently criticized tort’s physical-emotional distinction on the alleged basis that it commits tort to the objectionable position of mind-body dualism.Footnote ¹⁸ This section challenges this criticism. It argues that tort’s physical-emotional distinction, far from being a relic of dualist or otherwise unscientific thinking, is a heuristic that judges have used to structure and simplify the difficult but unavoidable task of drawing lines between legally cognizable and non-cognizable harm.

Embedded in the work of recent critics of tort’s physical-emotional distinction are several accurate observations, such as the following: (1) an emotional harm (e.g., chronic paranoia) can reduce a person’s welfare more than does a physical harm (e.g., a bruise); (2) people often disagree about whether a given harm is better described as physical or as emotional; and (3) the physical-emotional distinction can be collapsed on the basis that emotional harms have neural or other physical substrates. It bears noting that these observations are basically uncontroversially correct. For example, if some people in the past believed that emotional harm was necessarily insignificant, very few or no informed people do so today. Moreover, contemporary courts as a whole understand that “[e]motional injury can be as severe and debilitating as physical harm,”Footnote ¹⁹ that “mental and emotional distress is just as ‘real’ as physical pain,”Footnote ²⁰ and that “genuine mental illness constitutes real harm.”Footnote ²¹

At the core of the criticism that tort’s physical-emotional distinction commits tort to dualism is the thought that, in distinguishing between the physical and the emotional, tort denies or under-states the fact that emotional harms have neural or other physical substrates and often affect the physical body, as when the experience of disgust corresponds to activity in the anterior insula, or when the experience of fear increases heart rate. The criticism that tort’s physical-emotional distinction commits tort to dualism is incorrect. Given that both courts ^{Footnote 22} and legal scholars ^{Footnote 23} have acknowledged the inter-relatedness between the mind and the body for decades, it is very unlikely that contemporary tort doctrine can be plausibly said to hew to dualism. Moreover, dualism often or usually co-occurs with rejection of a naturalistic view of the mind,Footnote ²⁴ and there is no reason to suppose that tort practitioners are particularly likely to reject naturalism. A better reading of contemporary tort doctrine is that, although it recognizes that dualism is false, it preserves the physical-emotional distinction for reasons that are unrelated to any type of dualist motivation.

Consider, again, the Restatement. It explains that its definition of ‘bodily harm’ “is meant to preserve the ordinary distinction between bodily harm and emotional harm.”Footnote ²⁵ As the Restatement’s use of ‘ordinary’ suggests, its physical-emotional distinction is not intended to be scientific or metaphysical. Moreover, the Restatement recognizes that emotional harms have neural substrates, as in its remark that, “if the defendant’s negligent conduct (for example, negligent driving) frightens the plaintiff (for example, a pedestrian crossing the street), the harm to the plaintiff’s nerve centers caused by this fear does not constitute bodily harm” as the Restatement defines the latter.Footnote ²⁶ More generally, that tort distinguishes between the physical and the emotional does not warrant the conclusion that it denies monism, if only because an area of law can use an idea for instrumental purposes without thereby deeming the idea correct in a metaphysical sense.

Moreover, the criticism that tort hews to dualism is, in a well-defined sense, too easy. It is easy to find in any substantial body of legal doctrine ideas that do not withstand scrutiny from one or another technical perspective, to criticize the doctrine for including the ideas, and to do so without providing an account of why the doctrine should adhere to the relevant technical perspective. For example, some researchers in philosophy hold that ‘emotion’ is not a meaningfully unified category, and that in the future more sophisticated models of the mind will probably dispense with it.Footnote ²⁷ On the basis of this technical perspective, one can criticize tort not only for drawing the physical-emotional distinction but also for, at a more basic level, using the concept of emotion at all. Like the criticism that tort draws the physical-emotional distinction, the criticism that it uses the concept of emotion would face the following problem: it is not clear why the fact that some specialists in fields outside of law find the concept of emotion unsupportable warrants the conclusion that tort should abandon the concept.Footnote ²⁸

* * *

The remainder of this section argues that a better explanation of tort’s physical-emotional distinction is that it helps judges draw lines between legally cognizable and non-cognizable harm.

Because judges must regularly adjudicate disputes between interested and often strategic parties—and because they must do this in the presence of attention, resource, and time constraints—they must draw lines between legally cognizable and non-cognizable harm. The physical-emotional distinction has at least three features that make it a useful tool for judges who face this task. First, it is easily understandable and widely understood. Although people often disagree about whether a given harm (e.g., a headache) is better described as physical or as emotional, nearly every adult can and does encode a meaningful and salient difference between, say, impairment of a person’s arm and impairment of their self-esteem.Footnote ²⁹ As one California court put it, “it cannot be denied that [from the perspective of ordinary language] there is a difference between a pinched nerve and one which has been a conduit for an emotionally traumatic experience based on fright.”Footnote ³⁰ One way to appreciate the understandability of the physical-emotional distinction is in terms of what some researchers in psychology call the ‘prototype theory of categorization.’Footnote ³¹ A corollary of this theory is that, even in the presence of a large number of border-line cases, as long as people can reliably differentiate between the mentally represented prototypes of x and of y, they can easily understand both (1) that there is a distinction between x and y, and (2) the features on the basis of which the distinction rests. For example, people can easily understand that there is a distinction between ‘furniture’ and ‘decoration’ and what the distinction is about, and they can do this notwithstanding the existence of a large number of border-line cases, such as ‘lamp,’ ‘rug,’ and ‘vase.’ In this regard, what is true for the furniture-decoration distinction is true for the physical-emotional distinction as well. At any rate, that the physical-emotional distinction is easily and widely understood means that it is accessible to people both with and without legal training. The physical-emotional distinction is ecumenical and secular, not sectarian or mysterious. This in turn means that judges in a democratic polity can appeal to it in explaining or justifying their decisions without thereby running afoul of what some researchers in political philosophy call the ‘public reason’ constraint.Footnote ³²

A second and closely related feature of the physical-emotional distinction that makes it a useful tool for judges who must draw lines between legally cognizable and non-cognizable harm is that the distinction may be cognitively intuitive. Consider one reported finding that supports this hypothesis. At least two research teams in developmental psychology have reported that the capacity to perceive emotional harm as morally wrong requires more ontogenetic development than does the capacity to perceive physical harm as morally wrong. ^{Footnote 33} This reported finding means that young children develop the ability to encode physical harm (e.g., ‘Mary hit William’) as morally wrong before they develop the ability to encode emotional harm (e.g., ‘Mary humiliated William’) in this way. A corollary of this finding is that it is cognitively easier for people—including adults—to perceive physical harm (relative to emotional harm) as requiring redress, even where the difference in cognitive difficulty is small. Put differently, people perceive physical harms (relative to emotional harms) as more prototypical members of the category ‘moral wrong’; physical harms are moral wrongs par excellence.Footnote ³⁴ The greater cognitive ease of perceiving physical harm (relative to emotional harm) as morally wrong suggests that the physical-emotional distinction is cognitively intuitive.

Thirdly, the physical-emotional distinction happens to correlate with another and, for judges, more important distinction: that between easier- and harder-to-verify harm (here, ‘verification’ means something such as ‘objective corroboration’). To be sure, some harms that both ordinary language and tort tend to classify as physical (e.g., chronic back pain) are often hard to verify, whereas some harms that they tend to classify as emotional (e.g., catatonia) are often easy to verify. However, on average, traditional physical harm is easier to verify than is traditional emotional harm. As the Restatement puts it, “emotional harm is less objectively verifiable than physical harm and therefore easier for an individual to feign, to exaggerate, or to imagine.”Footnote ³⁵ This means that judges can use the physical-emotional distinction as a proxy for the distinction between easier- and harder-to-verify harm. Many courts have described the physical-emotional distinction as a proxy in this sense, explaining that what determines judicial classification of a harm as physical versus emotional is not where the harm falls along a mind-body continuum but the degree to which it is verifiable.Footnote ³⁶ Consider, for instance, the Maryland Supreme Court’s statement that, in the context of limitations on recovery for NIED, “the term ‘physical’ is not used in its ordinary dictionary sense. Instead, it is used to represent that the injury for which recovery is sought is capable of objective determination.”Footnote ³⁷

In summary, at least three features of the physical-emotional distinction make it a useful tool for judges who must draw lines between legally cognizable and non-cognizable harm: (1) it is easily understandable and widely understood; (2) it may be cognitively intuitive; and (3) it is a proxy for the distinction between easier- and harder-to-verify harm.

Showing that the physical-emotional distinction is a useful tool for judges who must draw lines between legally cognizable and non-cognizable harm is not the same as showing that judges have in fact used the distinction for this purpose. To see that judges have in fact used the physical-emotional distinction as a heuristic in this sense, it is helpful to consider case law. The following discussion focuses on case law pertaining to the ‘direct-victim’ variant of NIED.Footnote ³⁸

Consider four of the limits on recovery for emotional harm that courts either imposed in the past or continue to impose today.Footnote ³⁹ It can be helpful to understand these limits as exceptions to a default rule prohibiting recovery for emotional harm.

• • the ‘physical injury’ rule: recovery for emotional harm may be allowed if defendant negligently causes plaintiff physical harm and the emotional harm accompanies the physical harm.

• ○ Example: In Homans v Boston Elevated Railway,Footnote ⁴⁰ plaintiff was riding in defendant’s street-car when, due to an accident caused by defendant’s negligence, plaintiff was thrown against a seat. Plaintiff also experienced ‘nervous shock’ due to the accident. The court allowed recovery for the nervous shock on account of plaintiff’s collision with the seat.

• • the ‘physical impact’ rule: recovery for emotional harm may be allowed if, in negligently causing plaintiff the emotional harm, defendant makes physical contact with plaintiff.

• ○ Example: In Porter v Delaware, Lackawanna, and Western Railroad Co,Footnote ⁴¹ plaintiff was walking in the street when, due to an accident caused by defendant’s negligence, defendant’s railway bridge collapsed. Plaintiff experienced nervous shock due to the accident. The accident also scattered dust, some of which entered plaintiff’s eyes. The court allowed recovery for the nervous shock on account of plaintiff’s physical contact with the dust.

• • the ‘physical manifestation’ rule: recovery for emotional harm may be allowed if defendant negligently causes plaintiff the emotional harm and the emotional harm has a physical manifestation.

• ○ Example: In Brown v Matthews Mortuary, Inc,Footnote ⁴² defendant-mortuary lost the ashes of plaintiff-widow’s late husband and gave her the ashes of a stranger instead. Plaintiff later learned of the mistake. The court noted that plaintiff’s resulting headaches, sleep problems, and stomach pains would qualify as “physical manifestations” sufficient to state a claim of NIED.Footnote ⁴³

• • the ‘zone-of-danger’ rule: recovery for emotional harm may be allowed if, in negligently causing plaintiff the emotional harm, defendant places plaintiff in immediate danger of physical harm and causes plaintiff to fear for their safety.

• ○ Example: In Norfolk Southern Railway Co v Everett,Footnote ⁴⁴ a train that plaintiff-engineer was driving derailed due to defendant-railway company’s negligence. The court held that plaintiff’s claim of NIED survived summary judgment because, although the accident had not caused plaintiff physical harm, it had placed him in immediate danger of physical harm and caused him to fear for his safety.

For exposition, I will call these four rules the ‘traditional’ rules. This appellation is appropriate because, although each of these rules remains in force in at least one U.S. jurisdiction today, U.S. courts have been gradually moving away from them for decades.Footnote ⁴⁵ That said, the traditional rules are far from extinct. For example, the Restatement endorsed the ‘zone-of-danger’ rule in 2012,Footnote ⁴⁶ and, in 2015, five states used the ‘physical impact’ rule. ^{Footnote 47} As many courts have explained, the primary purpose of these rules is to provide judges with confidence that the relevant claim of emotional harm is genuine and serious, as opposed to feigned or trivial.Footnote ⁴⁸ As the Tennessee Supreme Court put it with respect to the ‘physical injury’ rule in particular, the rule “served to objectify the inquiry; it assured that the plaintiff’s allegations of emotional injury were grounded in an independently verifiable event.” ^{Footnote 49}

This section does not seek to add to the literature about the traditional rules’ history,Footnote ⁵⁰ justifiability,Footnote ⁵¹ or fit with general negligence doctrine.Footnote ⁵² Instead, it considers the rules’ conceptual structure, as well as what this structure can reveal about how and why judges created them.

Examining the traditional rules together helps illuminate at least three of their features. First, from a conceptual-structural perspective, all of the traditional rules operate in a similar way: they in effect define the concept of legally cognizable emotional harm by reference to the concept of physical harm. Moreover, in each of the traditional rules, the physical-emotional distinction operates recursively: it operates once to delineate a class of harms that require special limitation (namely, emotional harms), and it then operates a second time to set the limitation.

Secondly, despite the traditional rules’ obvious susceptibility to generating both false positives and false negatives, they are a rational response to the problem of drawing lines between legally cognizable and non-cognizable harm. The traditional rules are susceptible to generating false positives because not every emotional harm that satisfies the relevant rule is worthy of legal recognition, and they are susceptible to generating false negatives because not every emotional harm that fails to satisfy the relevant rule is unworthy thereof. However, despite the traditional rules’ susceptibility to generating wrong results, they are not irrational. As long as one accepts the reasonable assumption that, on average, an emotional harm that is indexed to a physical harm is more compelling than one that is not, the rules do, on balance, serve the function of differentiating between emotional harm that is and is not compelling. Consider, for instance, the ‘zone-of-danger’ rule, according to which recovery for emotional harm may be allowed if, in negligently causing the plaintiff the emotional harm, the defendant places the plaintiff in immediate danger of physical harm and causes the plaintiff to fear for their safety. As long as one accepts the reasonable assumption that, on average, an emotional harm that is indexed to the fear of physical harm is more compelling than an emotional harm that does not have this feature, the zone-of-danger rule does, on balance, serve the function of differentiating between emotional harm that is and is not compelling. Many post-1970 courts ^{Footnote 53} —as well as some earlier courtsFootnote ⁵⁴ —have described the traditional rules as ‘arbitrary’ or ‘artificial.’ The truth-value of this description is complex: the traditional rules are arbitrary and artificial—but not entirely so.Footnote ⁵⁵ They are arbitrary and artificial insofar as the presence of a connection to physicality—a physical nexus—does not guarantee that the relevant emotional harm is compelling (and, conversely, insofar as the lack of such a connection does not guarantee that the emotional harm is not compelling). However, they are not arbitrary or artificial insofar as, on balance, they are non-random and rational strategies for advancing a highly intelligible goal.Footnote ⁵⁶

Thirdly, examining the traditional rules together helps generate hypotheses about the cognitive-psychological processes by which judges created them. Consider two such hypotheses, which I call ‘association’ and ‘cognitive miser,’ respectively. According to the association hypothesis, in creating the traditional rules, judges were using simple associative thinking. More precisely, judges were accustomed to treating physical harm as legally cognizable, so they adopted the strategy of treating as legally cognizable that emotional harm that they could associate with physical harm. One implication of the association hypothesis is that there is at least one well-defined sense in which the traditional rules are common-sensical. According to the cognitive miser hypothesis, judges creating the traditional rules were economizing on cognitive resources. More precisely, instead of determining whether a given emotional harm is legally cognizable by asking the more difficult question of (e.g.) whether the harm ought to qualify for judicial redress, judges did so by asking the easier question of whether the harm stands in a certain type of relation to a physical harm. One implication of the cognitive miser hypothesis is that there is at least one well-defined sense in which the traditional rules are a default solution to the problem of drawing lines between legally cognizable and non-cognizable emotional harm. Although these two hypotheses are testable, they are obviously speculative, and I will not do more than outline them here. My goal here is only to suggest that treating the traditional rules as a group helps identify interesting and experimentally tractable questions about judicial cognition and decision-making.

The foregoing analysis argued, inter alia, that the traditional rules are an unsurprising response to the problem of drawing lines between legally cognizable and non-cognizable harm. One way to support this argument is by showing that decision-makers other than judges have used the traditional rules for similar purposes. An episode from British military history seems to provide evidence in this regard. During the First World War, as soldiers began to report ‘shell shock,’ the British military faced the problem of drawing lines between claims that it would deem compelling (which it sometimes called ‘wounds’ or ‘shell shock W’) and those that it would deem non-compelling (‘sickness’ or ‘shell shock S’).Footnote ⁵⁷ One of the ways in which the British military responded to this problem was by adopting the following rule: a soldier reporting shell shock might be deemed ‘wounded’—and thus eligible for a military pension—if his shell shock was connected to a discrete, identifiable shell explosion.Footnote ⁵⁸ Note that this rule seems to have been very similar to tort’s ‘physical impact’ and ‘zone-of-danger’ rules: a soldier whose shell shock is linked to a discrete shell explosion is more likely to have been physically touched (satisfying ‘physical impact’) and physically endangered (satisfying ‘zone-of-danger’) by the explosion. Like U.S. courts applying the traditional rules, the British military seems to have used the presence of a physical nexus as a heuristic for drawing lines between cognizable and non-cognizable emotional harm. Presumably, the decision-makers who adopted this rule were not primarily lawyers.Footnote ⁵⁹ This episode suggests that the traditional rules, rather than being an idiosyncratic creature of the judiciary, are tools that various types of decision-makers may use to structure and simplify a difficult line-drawing task.

In summary, the traditional rules represent an effort by judges to use the physical-emotional distinction as a heuristic for drawing lines between legally cognizable and non-cognizable harm.

* * *

Beginning around the 1970s, courts in many U.S. states began to break away from the traditional rules.Footnote ⁶⁰ More precisely, courts began to draw lines between legally cognizable and non-cognizable emotional harm without defining cognizable emotional harm in terms of physical harm. It is difficult to generalize about how courts did this, because the relevant doctrinal developments differed from state to state. Moreover, in some states, the high court was not always clear or consistent about its position, creating confusion at the bar.Footnote ⁶¹ As the North Carolina Supreme Court remarked in 1990, “[w]e perceive no single clear doctrine to which it can be said that a majority of states adhere” in the area of NIED.Footnote ⁶²

Just as courts had adopted more than one traditional rule, so they adopted more than one approach for breaking away from the traditional rules. A brief discussion of four post-1970 high court decisions—in California, Montana, Massachusetts, and Kentucky, respectively—conveys the diversity of approaches that courts took in this regard.

California. In Molien v Kaiser Foundation Hospitals, a doctor mistakenly diagnosed a wife with syphilis and told her to inform her husband of the diagnosis so that the husband could test himself for the disease.Footnote ⁶³ Although neither the wife nor the husband had in fact contracted syphilis, in what appears to have been an Ibsen-like misfortune,Footnote ⁶⁴ the diagnosis in effect ruined the marriage. The husband sued the doctor for NIED. The California Supreme Court held that a plaintiff may be able to recover for NIED absent a showing of a physical nexus if the emotional harm was (1) “foreseeable,” (2) “serious,” and (3) “objectively verifiable.”Footnote ⁶⁵ In Molien, in other words, instead of defining legally cognizable emotional harm in terms of physical harm (as the traditional rules do), the Court defined it as emotional harm that is foreseeable, serious, and objectively verifiable.

Montana. In Johnson v Supersave Markets, a man inadvertently issued a bad check to a store.Footnote ⁶⁶ Although he subsequently re-paid the debt, the store failed to notify the local prosecutor that he had done so, causing the man to be detained for about two hours. The man sued the store for NIED. The Montana Supreme Court held that a plaintiff may be able to recover for NIED absent a showing of a physical nexus if the emotional harm (1) invaded “a legally protected interest” of the plaintiff and (2) caused “a significant [but not necessarily physical] impact upon the person of plaintiff.”Footnote ⁶⁷ In Johnson, in other words, instead of defining legally cognizable emotional harm in terms of physical harm, the Court defined it as emotional harm that invades a legal interest of the plaintiff and significantly affects the plaintiff’s welfare.

Massachusetts. Until 1993, Massachusetts used the ‘physical manifestation’ rule:

Element (4) is a version of the ‘physical manifestation’ rule. In 1993, in Sullivan v Boston Gas,Footnote ⁶⁹ the Massachusetts Supreme Judicial Court re-interpreted element (4) to rid it of reference to the concept of physicality:

Thus, after 1993, Massachusetts plaintiffs were potentially able to recover for NIED absent a showing of a physical nexus if the emotional harm was (1) ‘objectively corroborated’ and (2) ‘reasonable.’Footnote ⁷¹

Kentucky. In Osborne v Keeney,Footnote ⁷² the plaintiff was inside of her house when a pilot flying a small plane crash-landed into her roof. The collision did not produce a physical contact with the plaintiff. The plaintiff sued for NIED, and the defendant moved for a directed verdict on the alleged basis that plaintiff had failed to satisfy Kentucky’s ‘physical impact’ rule. The trial court denied the motion, determining that the noise of the collision had constituted a ‘physical impact.’ Noting that the trial court had found it necessary to finesse the ‘physical impact’ rule, the Kentucky Supreme Court rejected the rule altogether. It determined that a plaintiff may be able to recover for NIED absent a showing of a physical nexus if the emotional harm was (1) “severe” and (2) supported by “expert medical or scientific proof.”Footnote ⁷³

In summary, the four preceding high court decisions provide a sense of the diversity of ways in which, beginning around the 1970s, courts in many states began to replace the traditional rules with other tools for drawing lines between legally cognizable and non-cognizable emotional harm.

For exposition, I will call rules such as those of Molien, Johnson, Sullivan, and Osborne ‘non-traditional’ rules. From a legal-doctrinal perspective, many of the non-traditional rules are what Dan Dobbs has called “ordinary negligence-foreseeability” rules, in that they treat NIED claims as more-or-less ordinary negligence claims.Footnote ⁷⁴ For present purposes, however, a different aspect of these rules is more relevant: they do not refer to the concept of physical harm.

Consider three features of the non-traditional rules. First, although they do not index legally cognizable emotional harm to physical harm, they continue to use the physical-emotional distinction as a line-drawing heuristic. They continue, in other words, to limit recovery for emotional harm in ways that they do not limit recovery for physical harm. This suggests that judges as a whole differentiate between (1) indexing legally cognizable emotional harm to physical harm, and (2) applying special limits to recovery for emotional harm.

Secondly, and relatedly, the non-traditional rules demonstrate a simple but perhaps non-obvious conceptual point: a judge can use the physical-emotional distinction as a heuristic for drawing lines between legally cognizable and non-cognizable harm without thereby defining cognizable emotional harm in terms of physical harm. This point, which may seem pedantic, matters, if only for rhetorical reasons. Where judges define legally cognizable emotional harm in terms of physical harm, as in the traditional rules, they risk creating the impression that they deem emotional harm second-class. This is because the traditional rules risk creating the impression that emotional harm is legally cognizable only insofar as it resembles physical harm. In this regard, the non-traditional rules are different. Because they define legally cognizable emotional harm in terms of ‘agnostic’ concepts such as ‘corroboration,’ ‘foreseeability,’ and ‘severity’—concepts that are largely orthogonal to the physical-emotional distinction—they are less likely to create the perception that tort depreciates emotional harm.

Thirdly, the process by which judges moved from the traditional to the non-traditional rules exemplifies judicial learning. More precisely, judges made this transition partly because they had learned—on the basis of first-, second-, and third-hand experience—that the traditional rules were flawed in their ability to draw satisfactory lines between compelling and non-compelling emotional harm.Footnote ⁷⁵ It can be helpful to schematize the transition from the traditional to the non-traditional rules as a one-step move. At time one, judges created the traditional rules partly because they hoped that these rules would allow for bright-line determination of whether a given emotional harm is legally cognizable. Consider, for instance, the ‘physical impact’ rule. Upon a casual reading, it is easy to suppose that this rule reliably allows for bright-line determination—that an emotional harm either does or does not result from an event in the course of which the defendant made physical contact with the plaintiff. As one Oregon court put it, “[t]he impact rule seems to us to reflect the best policy option. The line it provides is clear.”Footnote ⁷⁶ However, over time, as litigation brought an ever-larger store of fact patterns to judges’ attention, they gradually learned that the ‘physical impact’ rule would not reliably allow for bright-line treatment. Where defendant-hospital erroneously diagnoses plaintiff-patient with HIV, is the jab by which the hospital drew blood from the patient an ‘impact’ for purposes of the ‘physical impact’ test?Footnote ⁷⁷ Where plaintiff-restaurant patron bites into a chicken salad that contains an unwrapped condom, is the contact between plaintiff’s mouth and the salad an ‘impact’?Footnote ⁷⁸ Where a robber enters a suite in defendant-hotel and nudges plaintiff-hotel guest into the bathroom so that he is better able to search the living room for valuables, is the nudge an ‘impact’?Footnote ⁷⁹ The answers to questions such as these are not immediately obvious. Thus, at time two, judges began to move away from the traditional rules and toward the non-traditional ones. This transition can be usefully understood as an instance of a generic human learning process. Human learning often takes the form of creating an action rule, subsequently encountering previously unforeseen situations in which the rule’s performance falls short of the agent’s aspiration level, and then modifying the rule in light of the new situations. Judges encounter many opportunities to perceive an action rule’s inadequacy, because their work brings them into regular contact with what the Massachusetts Supreme Judicial Court has called “the infinite variety of new patterns of human activity.”Footnote ⁸⁰

For one source of evidence that judges’ move from the traditional to the non-traditional rules exemplifies learning, consider how judges themselves have explained why they made the transition.

As these statements suggest, judges who broke away from the traditional rules did so partly because they had learned the rules’ flaws. Consider, in this regard, Dillon v Legg, which is perhaps the best-known twentieth-century NIED decision.Footnote ⁸⁶ Dillon involved the following tragedy. A mother and two of her daughters, Cheryl and Erin, were crossing a street when a driver hit and killed Erin. Both Cheryl and the mother sued the driver in the bystander variant of NIED, on the basis of the emotional harm that they suffered when they saw what happened to Erin. The trial court applied the ‘zone-of-danger’ rule and sustained Cheryl’s claim but dismissed the mother’s, because, unlike Cheryl, the mother had been standing too far from the car to have been physically endangered by it. The California Supreme Court used Dillon to reject the ‘zone-of-danger’ rule, noting the obvious awkwardness of allowing only Cheryl, and not also the mother, the opportunity to recover. The California Supreme Court’s decision-making in Dillon suggests that, sometimes, judges reject a rule when and because they encounter a previously unimagined fact pattern that dramatically increases the salience of the rule’s flaws. As the California Supreme Court put it in Dillon, “[t]he instant case exposes the hopeless artificiality of the zone-of-danger rule.”Footnote ⁸⁷ Dillon dramatically shifted judges’ attention to this rule’s flaws by presenting a fact pattern that is unusual and thus difficult to imagine before the fact, and to which the rule gives an obviously awkward answer. Judges working before Dillon had not been exposed to the suit’s dramatic demonstration of the rule’s awkwardness. That a previously unimagined fact pattern can dramatically increase the salience of a rule’s flaws is obvious, but it raises at least two non-trivial points. First, present-day legal commentators may need to exert effort to perceive that historically earlier judges did not have access to the store of fact patterns with which today’s commentators are acquainted. More generally, in evaluating the performance of historically earlier decision-makers, it can be easy to over-look that they did not know all that we know today and thus to over-attribute their decision-making to dispositional factors (this over-attribution may be an instance of what some psychologists call the “fundamental attribution error”).Footnote ⁸⁸ Secondly, judges may need to exert effort to avoid being unduly influenced by recency effects, as when a particularly salient fact pattern causes a judge to exchange one rule for another that is equally or more flawed but whose flaws are not presently salient.

* * *

The foregoing analysis provides an account of why judges adopted the traditional rules, why many judges later became dissatisfied with them, and why many of the later judges subsequently replaced the traditional rules with the non-traditional ones. This account is schematic and abstracts from much doctrinal, historical, and sociological detail. However, it suffices to sustain the contention that, both before and after the 1970s, far from being invested in dualism, judges used the physical-emotional distinction as a common-sense tool to help them draw lines between legally cognizable and non-cognizable harm. In using the traditional rules, judges identified legally cognizable emotional harm with emotional harm that stands in a sufficiently close relationship to physical harm. In using the non-traditional rules, judges identified it with emotional harm that satisfies criteria that do not refer to the concept of physicality. In both cases, judges used the physical-emotional distinction as a heuristic to structure and simplify a difficult line-drawing task.

Before turning to some of the normative implications of the foregoing analysis, I would like briefly to consider the relationship between neuro-imaging (e.g., CT, fMRI, PET) evidence and claims of emotional harm. If and when neuro-imaging technology reaches the point at which emotional harm is just as verifiable as is physical harm, then one of tort’s important reasons for distinguishing between the two will dissolve. At the same time, pointing to the mere possibility that neuro-imaging technology may someday reach this point is not likely to undermine tort’s physical-emotional distinction. Unless and until neuro-imaging technology is able to close the verifiability gap between physical and emotional harm—something that it cannot do at present—most courts will continue to find the physical-emotional distinction useful in drawing lines between legally cognizable and non-cognizable harm. ^{Footnote 89}

II.A Users of tort’s physical-emotional distinction should clarify that they neither endorse dualism nor depreciate emotional harm

Users of tort’s physical-emotional distinction—such as judges, lawyers, and Restatement authors—may find it useful to clarify that, in distinguishing between the physical and the emotional, they neither endorse dualism nor depreciate emotional harm. The Restatement, for instance, makes some statements that suggest that it understands the physical-emotional distinction as a heuristic, but it is not as explicit as it can be in this regard.Footnote ⁹⁰ Restatement authors may consider making more explicit that they do not endorse dualism and that they know well that the boundary between the physical and the emotional is fuzzy and collapsible. They may also clarify that, in drawing the physical-emotional distinction, they hold that physical harm and emotional harm differ not in every case but only on average. Restatement authors may also explain that the document’s use of the physical-emotional distinction is wholly compatible with the recognition that emotional harm can severely reduce a person’s welfare.

At issue here are, inter alia, perceptions of the Restatement’s sophistication. Critics of the Restatement have argued that its use of the physical-emotional distinction impugns its consilience with science, and it is in the interest of Restatement authors to void this argument.Footnote ⁹¹ One way in which they may be able to do this is by emphasizing that the Restatement’s physical-emotional distinction amounts to an intentionally idiosyncratic use of an ordinary-language distinction. More generally, critics have argued that, in using the physical-emotional distinction, tort disregards “the importance of mental health to our lives.”Footnote ⁹² The many legal professionals who use the physical-emotional distinction while being highly cognizant of the importance of mental health need not, and should not, leave this criticism unrebutted.

II.B Because judicial expertise may not extend to the task of drawing lines between legally cognizable and non-cognizable harm, judicial performance in this area may be more adequate than critics suggest

The problem of drawing lines between legally cognizable and non-cognizable harm raises a question about the proper normative standard by which to evaluate judges’ performance in the area of NIED. Consider two perspectives that one may adopt in this regard. From one perspective, judges have performed poorly, producing limitations on the legal cognizability of emotional harm that are arbitrary, incoherent, or unprincipled. From the other perspective, judges have performed satisfactorily, producing limitations that, given the nature of the task, do an adequate job of distinguishing between compelling and non-compelling emotional harm. Which of such perspectives one adopts depends partly on what normative standard one applies.

In evaluating judges’ performance in general (i.e., beyond the area of NIED), it can be helpful to distinguish between problems in which judicial expertise—the particular expertise that judges command in their capacity as judges—is more and less likely to provide judges with an advantage over non-specialists or lay-persons. Because expertise is domain-specific (one cannot be an expert across the board), it is always relevant to ask whether a specialist’s expertise extends to a given problem. Judges’ expertise is probably more likely to provide them with an advantage in problem solving that requires, for instance, knowledge of the law, exposure to currents in academic research, or computational ability. For example, a judge should be better able than is a non-specialist to quickly understand the meaning of ‘a high court’s reversal of an appellate court’s reversal of a trial court’s denial of a motion to dismiss a complaint’—a task in which performance benefits from both knowledge of the law and computational ability. By contrast, judges’ expertise may be less likely to advantage them in problem solving that requires, for instance, over-riding the influence of strong occurrent emotions or of morally irrelevant numerical anchors.Footnote ⁹³

I suggest that the problem of drawing lines between legally cognizable and non-cognizable harm is one in which judges have less—rather than more—of an advantage over non-specialists. Put differently, this is a problem-domain in which judicial expertise makes only a limited difference to performance. This type of line-drawing is difficult largely because people—be they judges or ordinary citizens—do not tend to store in long-term memory well-formed prototypes of the concept ‘legally cognizable harm.’ Thus, judges who must draw lines between harm that the judiciary should and should not recognize face at least two challenges. The first challenge is to verbally define ‘legally cognizable harm’ (e.g., ‘a harm is legally cognizable if it impairs one or more of a person’s capabilities’). The second challenge is to produce a definition that other judges can understand and apply. To consider the preceding example, the benefit of defining ‘legally cognizable harm’ in terms of ‘capabilities’ is minimal if other judges are not clear what the first judge means by a ‘capability.’ It may be helpful to understand the first challenge as involving the creation of a standard, and the second as involving the creation of a justiciable standard.

The foregoing helps us understand why, where the task is to draw lines between legally cognizable and non-cognizable harm, judges have relied on a tool that is as simple as is the physical-emotional distinction. This tool is not uninteresting on account of its simplicity; on the contrary, its simplicity is among the reasons why it is notable. Where the task is to draw lines between legally cognizable and non-cognizable harm, judges have adopted simple solutions partly because complex ones are hard to find or not available.

The possibility that complex solutions to this problem are not available bears on the question of what normative standard one should adopt in evaluating judges’ performance in the area of NIED. If simple solutions are the only ones available in this problem-domain, then the better conclusion may be that judges have performed satisfactorily there. It may be unrealistic to expect judges to generate sophisticated solutions in a problem-domain in which such solutions may not exist.

II.C Although it may not be possible to determine the optimal way of drawing lines between legally cognizable and non-cognizable emotional harm, moral-philosophical tools such as Rawlsian and Scanlonian contractualism may be able to identify partial or pro tanto considerations for choosing among different ways of doing so

In the background of this essay’s analysis is a relatively basic question in legal design: what limits should tort impose on the legal cognizability of emotional harm? Consider, again, some of the relevant options. In principle, at least, tort can use a traditional or a non-traditional rule. Figure 1 lists the eight rules that this essay has discussed in this regard. As noted above, this list is not comprehensive.Footnote ⁹⁴ That said, taken together, the eight rules listed in Figure 1 provide a useful sense of the range of options that are available to courts imposing limits on the legal cognizability of emotional harm. How can one choose among rules such as those listed in Figure 1?

Figure 1.

Eight Competing Rules for Limiting the Legal Cognizability of Emotional Harm

The Introduction suggested that it may not be possible to use conceptual analysis to demarcate a category of ‘legally cognizable emotional harm’ that is both clear and uncontroversial. This philosophical point has a legal-doctrinal corollary: on the basis of the language or text of a rule, it may not be possible reliably to predict which emotional harms the rule will and will not recognize. Consider, for instance, the Sullivan rule, which uses the concepts of objective corroboration and reasonableness. Suppose that a court applies the Sullivan rule to a plaintiff-office worker’s claim that they suffered emotional harm due to being trapped in defendant-management company’s elevator for 30 minutes. Suppose also that the plaintiff introduces the testimony of a treating clinical psychologist that, as a result of this incident, the plaintiff experienced a debilitating anxiety attack. The language of the Sullivan rule does not reveal whether this claim is legally cognizable: it is not obvious whether it is reasonable to experience a debilitating anxiety attack due to being trapped in an elevator for 30 minutes; nor is it obvious whether the testimony of a treating clinical psychologist constitutes objective corroboration (among other things, whether a treating mental health professional can be relied on to provide objective testimony is a contested question in clinical psychology).Footnote ⁹⁵ To know which emotional harms the Sullivan rule categorizes as legally cognizable, one must, it seems, await judges’ interpretations and applications of the rule—interpretations and applications that are difficult or impossible to predict in advance. This example involving the Sullivan rule generalizes to other rules for imposing limits on the legal cognizability of emotional harm. The example suggests that it may not be possible meaningfully to choose among the rules on conceptual or a priori grounds.

In principle, at least, one option that is available to researchers who doubt the ability of conceptual analysis to demarcate a category of ‘legally cognizable emotional harm’ is to turn to the ‘social values’ approach that some or many contemporary policy-makers favor. ^{Footnote 96} Theoretically, one can use a social values approach such as the following to determine what limits tort should impose on the legal cognizability of emotional harm. First, create a list of all possible emotional harms. Next, elicit the public’s judgment about which of the listed emotional harms are compelling and which are non-compelling. Then, for every possible rule, determine how the rule would classify each of the emotional harms. Add a value of ‘1’ for every compelling emotional harm that the rule would recognize, and subtract a value of ‘1’ for every non-compelling emotional harm that it would recognize. Finally, choose the rule with the largest total value; this rule is the optimal rule—the rule that best draws the line between compelling and non-compelling emotional harm.

I take it as obvious (and I assume that the reader does as well) that this approach is infeasible on account of administrative, epistemic, and methodological difficulties. The infeasibility of the social values approach suggests that we do not know, in any rigorous sense of ‘know,’ which rule is best. This means that disagreements among legal scholars about how tort should limit recovery for emotional harm must, by necessity, involve guesses, hunches, or intuitions. More generally, nobody is in a position confidently to determine what is the optimal solution to the problem of drawing lines between legally cognizable and non-cognizable emotional harm.

Although it may not be possible to determine the optimal way of drawing lines between legally cognizable and non-cognizable emotional harm, certain moral-philosophical tools may be able to identify partial or pro tanto considerations for choosing among different ways of doing so. Consider, briefly, two such tools, both of which I draw from normative moral and political philosophy: Rawlsian and Scanlonian contractualism.Footnote ⁹⁷

One way to apply Rawlsian contractualism to the emotional harm question is as follows. Suppose that the members of a polity are deprived of information about variables such as their personality traits, the types of social environments in which they are likely to operate, and the probability that they will be plaintiffs or defendants in lawsuits involving claims of emotional harm. Deprived of such information about themselves, what rule for drawing lines between legally cognizable and non-cognizable emotional harm will these people choose? Answering this question is one way to apply the Rawlsian approach to the emotional harm question. I am skeptical of the Rawlsian approach on, inter alia, the ground that it is usually impossible to determine what policies, rules, or states of affairs hypothetical contracting parties would choose. That said, researchers who are more convinced of the determinacy of the Rawlsian approach may find it valuable to apply the approach to NIED in particular and to legal design more generally.Footnote ⁹⁸

Consider, now, one way to apply Scanlonian contractualism to the emotional harm question. A judicial regime in which some but not all emotional harms are legally cognizable will produce at least two types of ‘losers’: (1) defendants who sincerely believe that they must defend against claims of non-compelling emotional harm, and (2) plaintiffs who sincerely believe that the regime has failed to recognize their compelling emotional harm. Those in group (1) will have the good-faith complaint that they have been unjustifiably hauled into court, and those in group (2) will have the good-faith complaint that they have been unjustifiably deprived of their day in court. To determine what rule for drawing lines between legally cognizable and non-cognizable emotional harm tort should choose, one can ask the following question: what is the most morally attractive explanation or rationalization that tort can give to people in groups (1) and (2) about why it has put them in a ‘losing’ position? If tort uses one of the traditional rules, then it must tell the unhappy defendants that the reason why they have been hauled into court is that the regime has determined that the plaintiff’s claim has a physical nexus (one can conduct an analogous analysis with respect to the unhappy plaintiffs). By contrast, if tort uses one of the non-traditional rules—say, the Sullivan rule—then it must tell the unhappy defendants that the reason for their trouble is that the system has determined that the plaintiff’s claim is objectively corroborated and reasonable. By considering which of these explanations is normative-morally superior—and perhaps even by polling unhappy plaintiffs and unhappy defendants about which explanation they would prefer to receive—it may be possible to make progress in determining which rule tort should use to draw lines between legally cognizable and non-cognizable emotional harm.

Moral-philosophical tools such as Rawlsian and Scanlonian contractualism have non-trivial limitations, one of which is that they can identity only partial or pro tanto—and not complete or all-things-considered—considerations in favor of choosing one rule over others. For example, the preceding Scanlonian-contractualist approach would be open to the counter that, although it accounts for plaintiffs’ and defendants’ interests, it fails to account for those of judges or of elected legislators. This means that, although these moral-philosophical tools can identify discrete considerations in favor of choosing one rule over others, they cannot identify all that one would need to consider in order comprehensively to determine where to draw the line between legally cognizable and non-cognizable emotional harm. With respect to this recalcitrant problem, ‘good enough’ may be the best that any judicial regime can do.