|
Article Excerpt
I. INTRODUCTION
Actus non facit reum nisi mens sit rea--"the act does not make a person guilty unless the mind is also guilty." (1) Few today would disagree with the maxim; the criminal law has long since rejected the idea that causing harm should be criminal regardless of the defendant's subjective culpability. (2) Still, the maxim begs a critical question: can jurors accurately determine whether the defendant acted with the requisite "guilty mind"? St. Thomas Aquinas was certainly skeptical that such mindreading--as cognitive psychologists call it (3)--is within the ken of mere mortals:
[M]an, the framer of human law, is able to judge only of outward acts; because man seeth those things that appear, according to 1 Kings 16.7; but God alone, the framer of the Divine law, is able to judge the inward movement of wills.... (4)
Given the significant cognitive demands contemporary criminal law imposes on jurors, Aquinas's skepticism seems more prescient than ever. The common law, for example, not only asks jurors to distinguish between seventy-eight different terms for mental states, (5) it fails to define the various mental states consistently--there are seven different definitions of willful alone (6)--and combines them in permutations that defy comprehension. (7) Indeed, the common law is such a mess that no less an authority than Justice Robert Jackson once bemoaned the "variety, disparity, and confusion" of the "requisite but elusive mental element." (8)
The Model Penal Code (MPC), adopted by a majority of states, (9) may actually be worse. To be sure, the MPC takes a far more systematic approach to mens rea, winnowing the common law's seventy-eight mental states to four: purpose, knowledge, recklessness, and negligence. (10) Those mental states are differentiated with such "subtlety and precision," however, that it is an open question whether jurors can accurately distinguish them. (11) Consider, for example, the MPC's three subjective mental states: a person acts "purposely" if his "conscious object" is to bring about a particular result; (12) acts "knowingly" if he is "aware that it is practically certain" that his conduct will lead to the result; (13) and acts "recklessly" if he "consciously disregards a substantial and unjustifiable risk" that his conduct will cause the result. (14) Those are fine distinctions, to say the least. Little wonder, then, that scholars have described the MPC as an "elaborate set of precise rules whose operability depends on the jury's willingness"--to say nothing of their ability--"to make artificial characterizations." (15)
The MPC's idiosyncratic definition of negligence only further complicates mindreading. Negligence is defined as the defendant acting "when he should be aware of a substantial and unjustifiable risk" that his conduct will lead to a particular result. (16) Common law negligence is not actually a mental state; the reasonable-person standard "is determined and applied without reference to what the actor was thinking at the moment." (17) The MPC, by contrast, subjectivizes negligence, once again requiring jurors to read the defendant's mind: whether the defendant's conduct was unreasonable must be determined "considering the nature and purpose of his conduct and the circumstances known to him" at the time of the crime. (18)
Nor is that all. Although the common law at least limited each offense to a single mental state, (19) the MPC permits different mental states to apply to different material elements of an offense--what is known as "element analysis." (20) Misdemeanor indecent exposure is an example: "A person commits a misdemeanor if, for the purpose of arousing or gratifying sexual desire ... he exposes his genitals under circumstances in which he knows his conduct is likely to cause affront or alarm." (21) Thus defined, a jury must determine both whether the defendant acted knowingly (with regard to his conduct's potential to cause affront or alarm) and whether he acted purposely (with regard to arousing or gratifying sexual desire)--a far more complicated mindreading task than determining a single mental state.
This discussion also assumes that jurors normally determine whether the defendant is guilty of a single crime. In many cases, of course, jurors will have to choose between multiple closely-related crimes, whether because the prosecution has charged them in the alternative or because they are added in jury instructions as lesser-included offenses. Jurors in a homicide case, for example, may be asked to determine whether the defendant is guilty of first-degree murder, second-degree murder, reckless homicide, or criminally negligent homicide. Those charges are normally distinguished solely by the mental state they require: a "premeditated and intentional" killing for first-degree murder; (22) a "knowing" killing for second-degree murder; (23) a "reckless" killing for reckless homicide; (24) and "criminally negligent conduct" that results in death for criminally negligent homicide. (25)
Finally, in most criminal cases, jurors will have to deal with yet another mental state: namely, the one associated with the defense the defendant uses to negate the mens rea of the charged crime. (26) A defendant may defend himself against a murder charge, for example, by alleging that he was legally insane at the time of the killing. (27) Most often, as in the murder example, the focus will be on the defendant's lack of a "guilty mind." In some cases, however, the defendant may invoke a defense that shifts attention to the victim's mental state at the time of the crime--a rape case, for example, in which the defendant claims that the victim consented to the sex.
It is fair to say, in short, that contemporary criminal law requires jurors to be latter-day Kreskins--to not only reliably distinguish nearly-indistinguishable mental states, but also to accurately determine which of many possible mental states the defendant actually possessed at the time of the crime. (28) Is such mindreading possible? Or is Aquinas correct that "God alone ... is able to judge the inward movement of wills"?
Given the centrality of mens rea to criminal responsibility, we would expect legal scholars to have provided a persuasive answer to this question. Unfortunately, nothing could be further from the truth. Most scholars simply presume that jurors can mindread accurately, and those that take mindreading seriously have uniformly adopted "common sense functionalism," (29) a theory of mental-state attribution that is contradicted by a vast amount of research into the cognitive psychology of mindreading. Common-sense functionalism assumes that a juror can accurately determine a defendant's mental state through common-sense generalizations about how external circumstances, mental states, and physical behavior are causally related. (30) Research indicates, however, that mindreading is actually a simulation-based, not theory-based, process. When a juror perceives the defendant to be similar to himself, he will mindread through "projection," attributing to the defendant the mental state that he would have had in the defendant's situation. And when the juror perceives the defendant to be dissimilar to himself, he will mindread through "prototyping," inferring the defendant's mental state from the degree of correspondence between the defendant's act and his pre-existing conception of what the "typical" crime or defense of that type looks like.
The goal of this Article is to provide a comprehensive--though admittedly speculative--explanation of how jurors use projection and prototyping to make mental-state attributions in criminal cases. The Article is divided into six Parts. The first two provide the necessary background: Part II sketches the traditional functionalist explanation of mindreading, and Part III explains why jurors are unlikely to use a functionalist method in a case that focuses on the defendant's mens rea. The next two Parts are descriptive: Part IV introduces projection and prototyping and discusses the evidence that jurors use them to make mental-state determinations, while Part V explains the cognitive mechanism--perceived similarity between juror and defendant--that determines which technique a juror will use in a particular case. The final two Parts are then analytic: Part VI explains why projection and prototyping are likely to result in inaccurate mental-state determinations, and Part VII discusses de-biasing techniques that could be used to improve their accuracy.
II. THE TRADITIONAL EXPLANATION
Rebecca Dresser has argued that "the general legal reaction to the proof issues raised by other minds judgments is either to ignore or to dismiss them." (31) It is difficult to disagree. Aside from Dresser's own article and an article by Adam Candeub, (32) one searches the legal literature in vain for a sustained discussion of the psychology of juror mindreading. Most scholars, it appears, simply presume that accurate mindreading is possible--a position that dates back at least to 1882, when Bowen famously dismissed Oliver Wendell Holmes's skepticism toward mindreading (33) by claiming that "the state of a man's mind is as much a fact as the state of his digestion." (34) Little has changed in the intervening century. In fact, as recently as 1987, Richard Singer could assert without argument that "[i]t is certainly within the jury's ken to find that a typical self defender did not intend (purpose) a killing." (35)
The question, of course, is why legal scholars take it for granted that jurors can make accurate mental-state determinations. As discussed above, (36) given the significant cognitive demands that contemporary criminal law imposes on jurors, it is far from obvious that they can. The answer seems to be that legal scholars embrace, implicitly or explicitly, a commonsense theory of mental-state attribution in which mindreading seems neither particularly complicated nor particularly problematic.
That theory relies on three interrelated assumptions. The first is that there are fixed and unchanging causal relationships between external circumstances, internal mental states, and physical behavior. Jerome Hall's response to Holmes--echoing Bowen--is illustrative:
[T]he inevitable limitations on our knowledge do not support the conclusion that, in the vast majority of judgments based on rational methods of investigation, there is no reasonably accurate correspondence. The whole law of evidence is a studied effort to cope with this fundamental problem and its justification rests on the high probability of a "sufficiently" accurate representation of inner states by external conduct. This premise is accepted not only in law, but throughout every avenue of social life. It rests ultimately on the essential uniformity of human nature, and is so deeply ingrained in our daily experience that it is hardly possible even to conceive of social intercourse founded on its rejection. (37)
The second assumption, dependent on the first, is that jurors possess an intuitive theory of the specific causal relationships that exist between circumstances, mental states, and behavior. Wigmore divides those relationships--what H.L.A. Hart calls "common-sense generalizations about human nature" (38)--into two basic categories: (1) "circumstances tending to excite, stimulate, or bring the emotion in question into play"; and (2) "outward conduct expressing and resulting from the emotion in question." (39) Hart's insistence "that men are capable of self-control when confronted with an open till but not when confronted with a wife in adultery" (40) and Oliver Wendell Holmes's belief that "[d]etached reflection cannot be demanded in the presence of an uplifted knife" (41) are examples of commonsense generalizations in the first category. Rex v. Shaw's assertion that strangling someone to death with a rope indicates malice aforethought is an example of a commonsense generalization in the second category. (42) As the examples indicate, both categories of generalizations take a common "if x, then y" form, where x is the circumstances or behavior and y is the mental state that can be inferred from them. (43)
The third and final assumption then flows naturally from the second: jurors determine a defendant's mental state by applying their commonsense theory of the specific relationships that exist between circumstances, mental states, and behavior to the evidence in the case. Hall makes that point explicitly in his discussion of how fact-finders rely on "the essential uniformity of human nature" to determine what kinds of behavior indicate that a defendant acted "intentionally":
[H]uman conduct that is associated causally with certain harms proscribed by law, is labelled [sic] "intentional" by triers of the material facts on the basis of knowledge of certain external data.... Given certain facts, we must conclude that any and every rational human being in those circumstances did or did not intend the results.... (44)
Judge Posner offers a more specific example of this reasoning process--what he calls, somewhat idiosyncratically, a "behaviorist account of deliberation" (45)--when he discusses how jurors infer premeditation from the evidence in a case:
[I]n deciding whether a crime is premeditated ... we examine the circumstances of the crime: Was it concealed? Had the criminal made arrangements for a getaway? Had he obtained the means of committing the crime in advance? Were those means suitable to the end (suitably lethal, in the case of a murder)? Did the criminal have much to gain from the crime? From these circumstances a model of a deliberating criminal could be constructed--an "objective" reconstruction of the criminal's motivational experience, created by attributing to him a certain type of rationality. (46)
Taken together, these three assumptions constitute what cognitive psychologists have described as a functionalist account of mindreading. (47) We can summarize that account as follows:
[O]ur mindreading capacity is implemented by an intuitive theory of mind, a body of knowledge or belief about the causal relations between mental states on the one hand and behavior, environment and other mental states on the other.... In mindreading, we use these generalizations and our data about the target's circumstances or behavior to infer some of the target's mental states. (48)
Two additional aspects of functionalist theorizing are worth noting. First, the theory assumes that jurors acquire their "intuitive theory of mind" in a manner that approximates the scientific method: via "a gradual process of theorizing, making generalizations based on the available evidence, revising ... theories to handle incongruent evidence, etc." (49) The process, in other words, is empirical, (50) grounded in jurors' real-world observation of causal relationships between external circumstances and mental states and between mental states and physical behavior. (51)
Second, as conceptualized by functionalist theorizing, mindreading does not involve any kind of introspection on the part of the juror--the juror mindreads by applying his commonsense generalizations to the evidence in the defendant's case; he does not ask what mental state he would have had in the defendant's situation. (52) The distinction is a fine one: given that the juror's generalizations are ostensibly based on the psychology of "any and every rational human being," they should apply no less to him than to others. Nevertheless, although theorizing and introspection may well lead a juror to attribute the same mental state to the defendant, the processes they employ are completely different: functionalism views mindreading as the application of a theory, not as an exercise in phenomenology.
III. PROBLEMS WITH THE TRADITIONAL EXPLANATION
There is no question that functionalist theorizing provides an intuitively plausible explanation of how jurors mindread. (53) Jurors do not have direct access to the defendant's mind, but they do know what the defendant's physical behavior was--the crime itself, the actus reus--and have at least some sense of what the external circumstances were that led to that behavior. It thus seems eminently sensible to assume that a juror would infer the "missing" mental state from the evidence by relying on an intuitive theory of how circumstances, mental states, and behavior are causally related.
For a number of reasons, however, projection and prototyping provide a far better account of how jurors make mental-state determinations than functionalist theorizing. To begin with, research indicates that the "default starting point" (54) of mindreading is projection or prototyping, not functionalist theorizing. In terms of prototyping, for example, Fiske and Neuberg have found that "[p]erceivers attempt category-based impression formation before they use more attribute-oriented impression formation, and if relatively category-oriented processes are successful, then the perceiver goes no further toward more attribute-oriented processes." (55) Similarly, in terms of projection, Krueger has shown that "[w]hen the responses of others are not known, people project their own as a first bet." (56)
Krueger's conclusion indicates, of course, that jurors may shift from projection to functionalist theorizing if the "responses of others" are known. Indeed, Ames argues that, in general, "[c]umulative behavioral evidence supercedes extra-target strategies: [p]rojection and stereotyping will drive mindreading when behavioral evidence is ambiguous, but as apparent evidence accumulates, inductive judgments will dominate." (57) That makes sense--if a mindreading target's behavior is unambiguous because we know how the target normally reacts to certain situations or because one and only one mental state is consistent with the behavioral evidence, there is no reason to disregard that knowledge in favor of "extra-target strategies" like projection and prototyping.
Few criminal cases that focus on mens rea, however, will provide jurors with sufficiently unambiguous behavioral evidence. First, voir dire will ensure that the defendant is a stranger to the jurors, (58) while the character evidence rule will normally prevent the prosecution from compensating for that fact by bringing information about the defendant's character to their attention. (59) Jurors will thus usually have no information concerning how the defendant normally responds to an ordinary situation, much less to an extraordinary situation like the one at issue in the case.
Second, it is very unlikely that the behavioral evidence in the case--the actus reus--will be consistent with one and only one mental state. Contrary to functionalist theorizing, "there is no one-to-one correspondence between events and other people's reactions to these events; people do not always react the same way to a given event, and the same psychological reaction can be elicited in many different ways." (60) That does not mean a juror will see all of the possible mental-state inferences in a criminal case as equally reasonable; the behavioral evidence may make some mental-state inferences more plausible than others. But it seems very unlikely that a juror trying to determine the defendant's mental state in a case involving a murder or a rape, for example, will view the behavioral evidence as unambiguous, given that such complex acts are particularly difficult to mindread. (61) Moreover, the juror will not be mindreading in a vacuum: his determination will be framed by the competing descriptions of the defendant's mental state offered by the prosecution and the defense, a conflict that cannot help but reinforce the ambiguity of the behavioral evidence. (62)
Both the default primacy of extra-target mindreading strategies and the general ambiguity of the actus reus in a criminal case indicate, in short, that jurors are likely to use projection and prototyping, instead of functionalist theorizing, to determine a defendant's mental state. Indeed, that is exactly what a significant number of empirical studies of mindreading in the legal context have found. That research is discussed in the next Part.
IV. PROJECTION
A. PROJECTION DESCRIBED
As we have seen, the functionalist model of mindreading makes two basic assumptions: (1) that jurors have an intuitive theory of mind that consists of commonsense generalizations about the causal relationships between external circumstances, mental states, and physical behavior; and (2) that jurors mindread by applying their intuitive theory of mind to the evidence in the defendant's case and then imputing the resulting mental state to the defendant.
The projection model of mindreading conceptualizes mental-state determinations very differently. It does not assume that jurors possess an intuitive theory of mind; indeed, interest in the model was originally sparked by suspicion that functionalist mindreading "is too information rich--that it attributes too much knowledge and sophistication to mindreaders, even if the knowledge is only tacit or unconscious." (63) Instead, it assumes that individuals are "information poor" but possess a "special skill, namely, constructing pretend, simulated or imaginary mental states ... in simulated processing." (64) That skill, according to the projection model, is at the heart of mindreading:
In using the simulation heuristic to detect a target's mental state ... an attributor begins by taking the target's "perspective." That is, he pretends to be in certain states the target is in, as suggested by his previous information about the target.... He feeds these "pretend" starting states into an appropriate cognitive mechanism of his own ... and lets it operates on them. This mechanism outputs a new state ... and the attributor then attributes that decision to the target. In other words, the attributor tries to make his own mind "emulate" the mental sequence the target will go through.... The heart of this procedure is that the attributor tries to reproduce or match what transpires in the target.... (65)
This form of projection is prospective: the mindreader uses simulation to move forward from information about the target's situation--his physical environment and whatever is known about his psychological characteristics--to the target's mental state. (66) Projection can also be retrospective, using simulation to move backward from the target's physical behavior--the criminal act itself--to the mental state that caused it. Goldman describes this use of projection as a "generate and test strategy":
The attributor begins with a known effect of a sought-after state, often an observable piece of behavior. He generates one or more hypotheses about the prior mental state or combination of states that might be responsible for this effect. He then "tests" (one or more of) these hypotheses by pretending to be in these states, feeding them into an appropriate psychological mechanism, and seeing whether the output matches the observed evidence. When a match is found (perhaps the first match, or the "best" match), he attributes the hypothesized state or combination of states to the target. (67)
Neither form of projection, it is important to note, is completely atheoretical. In particular, "there might be theory in deciding which pretend inputs to construct, and in transferring out simulated state or behavior from ourselves to the target." (68) The former moment is particularly important, because the goal of projection is not to determine the mental state the mindreader would have had in the target's situation, but to determine what mental state the mindreader would have had in the target's situation if he were the target. Accurate projection thus depends, as explored in detail below, (69) on the mindreader's ability to adopt the target's characteristics and to prevent his own characteristics from influencing the simulation process.
Despite these theoretical moments, however, the core of projection remains the mindreader's imaginative act of simulating, experiencing, and reacting to the target's situation: "acknowledging that accuracy of inputs requires informational guidance doesn't undercut the simulational aspect of the cognitive performances in question; the inputs are inputs for simulation." (70) Indeed, the act of projection could not be more different than functionalist theorizing, given the latter's emphasis on the logical and mechanical application of commonsense generalizations about the causal relationships between external circumstances, mental states, and physical behavior. In particular, there is no room in functionalist theorizing for projection's transfer of one of the mindreader's mental states--the mental state generated by the act of simulation--onto the target; by definition, the commonsense generalizations that the functionalist mindreader applies are derived from his experientially acquired intuitive theory of mind, not from introspection.
If jurors use projection to mindread, they most likely rely on a hybrid form of projection, one that is both prospective and retrospective. In criminal cases that focus on mens rea, jurors will usually have two different kinds of information: (1) evidence concerning the defendant's situation, such as the circumstances in which he committed the crime and the mental state that he was in prior to finding himself in those circumstances; and (2) a critical "observable piece" of the defendant's behavior, namely, the actus reus of the crime. They will thus have the raw material they need to project both prospectively (from the defendant's situation to his mental state) and retrospectively (from the actus reus to the defendant's mental state).
There is, moreover, another critical difference between general mindreading and juror mindreading that we need to consider. Whereas mindreaders usually have to generate the hypotheses they test through projection on their own, the prosecution and defense in a criminal case generate those hypotheses for them. In some cases, there will be only two competing hypotheses--a murder case, for example, in which the only issue is whether the defendant was sane or insane at the time of the crime. In others, there may be far more than two, such as a murder case in which the prosecution charges the defendant with various degrees of murder and the defense argues both self-defense and provocation.
What, then, does projection-based mindreading look like in the legal context? The mechanics of projection and the nature of criminal trials suggest a model in which a juror uses a two-stage process of projection to test the various mental states offered by the prosecution and defense. In stage one, the juror runs a retrospective simulation in which he tries to imagine having a particular mental state in the defendant's situation. If he cannot imagine having that mental state, he rejects it and repeats the first stage with the next one. But if he can imagine having it, he then proceeds to stage two and runs a prospective simulation in which he tries to imagine that mental state leading him to commit the actus reus of the crime. If he cannot imagine committing the crime, he rejects the mental state and returns to stage one with the next possible mental state. But if he can imagine committing it, the projection process stops and he imputes that mental state to the defendant.
Variations on this model are certainly possible. Perhaps, instead of ending the projection process as soon as he finds a matching mental state, a juror tries to simulate all of the mental states offered by the prosecution and...
|
