The Ethics of Forensic Practice: Reclaiming the Wasteland

An Appreciation

I took Stone's full-year course in law and psychiatry, co-taught with Alan Dershowitz, during my second year in law school. Dershowitz had a superb analytic mind and dazzling Socratic skill, but if one listened carefully to his quieter co-teacher, one understood at once that Stone was every bit Dershowitz's intellectual equal and was philosophically inquisitive about the issues. I was so impressed that I took every other course that Stone offered, and inveigled him into being my law school thesis advisor and into serving on my doctoral dissertation committee. He was also my first clinical supervisor and no supervisor ever taught me more. In short, I studied at the Stone University of Law & Psychiatry and my entire career manifests the influence. I was taught to be conceptual, skeptical, and always morally attentive.

His influence is, of course, wider than the impact on previous students. The occasion of this symposium indicates how far-reaching his ethical challenge to forensic psychiatry has been. His many other works in mental health law,^4,5 have also profoundly affected mental health scholars and practitioners. This issue's set of papers discussing his work is a fitting tribute to a scholarly life well-lived.

When I decided to pursue a career primarily as an interdisciplinary, academic lawyer, Stone offered his complete support and help. Those were the days when law faculty and legal education in general were much less interdisciplinary than they are today, so my credentials looked nontraditional and, indeed, suspect to the traditionalists. Nonetheless, with his support, I found a wonderful position at my first school, USC. I am forever grateful for what he taught me and for what he did for me. This contribution is a mark of affection and respect.

The General Ethical Issue

The great political philosopher, Brian Barry, says that the problem for justice, which is itself a moral concept, is how we should live together when there is little agreement about how we should live, about what a good life and justice demand.⁶ In a social order such as ours that is dedicated, at least in principle, to the use of public reason rather than the use of force to solve these problems, the dominant response to claims of justice is procedural. Establish a mechanism for deciding such questions that all can live with and then, subject to some substantive constraints, agree to live with whatever decision the procedure produces. So, for example, the United States Constitution places few substantive constraints on the lawmaking power of either the Congress or state legislatures. Except for a few fundamental interests that are specially protected, and as long as the state has adhered to the decision rules, state power is limited only by a requirement that the state action meets the standard of minimal rationality, of being rationally related to a legitimate state purpose.

Now, our legal system is adversarial, not inquisitorial. Might a continental system be more rational? Of course, but it is not ours. In our system, expert consultants and witnesses routinely aid rational adversarial fact-finding. We often need expert help adequately to resolve factual disputes when the matter at issue is beyond the ken of the layperson. We expect adversarial witnesses to favor the party that has engaged them, although not to the extent of distorting their data, going beyond the evidence, or otherwise misleading the finder of fact. We expect truth and justice to prevail through the adversarial presentation of evidence and vigorous cross-examination. This is the nature of the beast. As long as experts stay within reasonable professional bounds, they have done their job, and the law owes them a debt of gratitude.

Forensic psychiatrists are no different in this regard from other experts who offer services to the law. Stone is right that market forces, the desire to please, pre-existing moral and political commitments, and a host of other factors can cause forensic psychiatrists to do unprofessional work, but that is true of virtually all experts. Forensic psychiatrists are no more at the mercy of such variables than are other experts who base their practice on a similarly established clinical or data base. They are less at risk than experts, such as art authenticators, for example, who may have few if any objective criteria to guide them. Forensic practitioners must of course be on their guard, but we do not want to throw out the baby with the bathwater if there really is a baby in the tub. That question I shall discuss in the next section.

Stone is correct that there is no consensus ethical stance towards forensic practice, but my response is that this is entirely expectable and unproblematic. How we should live and live together are fraught questions. Except at the level of the most banal generalities, there is no consensus ethics in any context. Even if a professional group were to adopt a code of ethics, as virtually all have, that might be termed consensual in virtue of its adoption, that code will be open to interpretation about many matters and it will not be immune to the types of arguments undermining its premises that rational people might accept. The history of the shifts in medical and legal ethics confirms this observation. At the level of mundane forensic practice, there are many issues that will be ethically contestable and that permit conflicting reasonable views. My conclusion is that as long as the forensic practitioner, whose work is accepted in general by society, chooses a reasonable, coherent, publicly defensible ethics to guide specific work, the practitioner does not run afoul of his or ethical duty.

The standard model, to which I adhere, argues that the forensic practitioner owes only the duty to act respectfully and honestly towards the subject and to perform his forensic functions with the highest level of professional skill. I perceive no inherent conflict with accepting and acting in that role and with training as a psychiatrist. Different roles place different ethical demands on us. Stone claims, however, that a doctor's duty is always to the patient first. It is a special type of relationship and has much in common with the duties of lawyers to their clients. It is a helping relationship that goes far beyond an economic exchange. Many other experts—think about engineers or art appraisers, for example—have no such duties, no such relationship. Nevertheless, when the psychiatrist engages in forensic practice, he or she is no longer a helper and the subject is not a patient. He or she owes no duty of beneficence to the subject of the evaluation. Indeed, to explicitly or implicitly indicate otherwise to the subject of the evaluation is a grave ethical lapse because it establishes a false basis for the interaction that will follow.

The standard model starts with impeccable moral pre-commitments to respect for persons and professional integrity. It suggests that forensic practitioners can serve a socially important and useful function if they adhere to those pre-commitments. The model appeals to public reason. Stone has demonstrated no logical error in the standard model's argument, and indeed it is hard to imagine what such an error might be. The only possible critique is an external one that starts from contestable premises, such as that doctors must always act for the benefit of the subject whenever they are using their professional skills. Stone has not argued for this premise, and, given his apparent non-objectivism about ethics, it might be difficult for him to do so. One can imagine a rational argument for such a premise, say, one based in the most just distribution of social resources. For example, there is a real question about whether society as a whole is better off having trained psychiatrists spending more time in the courtroom and less time helping people who have mental disorders.

All ethical systems must begin with an asserted foundation that cannot be argued for further. If a psychiatrist wishes to adopt Stone's monistic ethics of caring stance, then he or she will behave properly and consistently if he or she does not engage in forensic work. On the other hand, if there is a genuine need for psychiatric assistance to adversarial decision-making, and our society reasonably believes that there is, I see no reason that forensic psychiatrists who adopt the premises of the standard model should have any questions about their ethical bona fides.

This assumes, however, that forensic psychiatrists really have something to offer that makes forensic work a contribution to social welfare. If Stone is right that they have almost nothing to contribute, then forensic work is at the least a large social cost and at the worst unethical conduct because it is not capable of providing what it claims it can. I conclude, in contrast to Stone, that forensic psychiatrists do have something to contribute, thus justifying their work, but that the contribution is considerably less than many believe and the law permits. The next two sections of this article support this argument by explaining the law's criteria that forensic work helps evaluate and by showing how forensic work can be relevant to those criteria.

The Nature and Justification of Legal Criteria

The civil and criminal legal standards that forensic psychiatrists address are entirely behavioral—broadly speaking to include mental states (e.g., perceptions, beliefs, desires) and actions. They must be understood in conjunction with the law's folk psychological view of the person and behavior that holds that mental states—specifically desires, beliefs and intentions—are crucial for causally explaining and evaluating action. It presupposes that genuine agency is possible. For example, if one were to ask you, the reader, why you are reading this article, I trust that no one would tell me a causal-mechanical story about brain states (even if such a story were now possible, which it is not). Instead, you would give a reason along these hypothetical lines: you desire to learn something about the ethics of forensic psychiatry, you believe that reading this article might satisfy that desire and you formed the intention to read it. This is the type of causal account we all use all the time to explain our behavior to ourselves and others.

The law's view of the person is thus the so-called folk psychological model: a conscious (and potentially self-conscious) creature capable of practical reason, an agent who forms and acts on intentions that are the product of the person's desires and beliefs. In Western philosophy, this model has its primary roots in Aristotle's philosophy. We are the sort of creatures that can act for and respond to reasons. The law does not treat persons generally as non-intentional creatures or mechanical forces of nature. It could not be otherwise. Law and morality are action-guiding⁷ and could not guide people ex ante and ex post unless people were the types of creatures who could use rules as premises in their practical reasoning. Law and morality as action-guiding, normative systems of rules are otherwise useless, and perhaps incoherent. Law is a system of rules that at the least is meant to guide or influence behavior and thus to operate as a potential cause of behavior. As the philosopher John Searle writes: Once we have the possibility of explaining particular forms of human behavior as following rules, we have a very rich explanatory apparatus that differs dramatically from the explanatory apparatus of the natural sciences. When we say we are following rules, we are accepting the notion of mental causation and the attendant notions of rationality and existence of norms…. The content of the rule does not just describe what is happening, but plays a part in making it happen [Ref. 8, p 33].

Legal and moral rules are not simply mechanistic causes that produce reflex compliance. They operate within the domain of practical reason. Agents are meant to and can only use these rules as potential reasons for action as they deliberate about what they should do. Moral and legal rules are thus action-guiding primarily because they provide an agent with good moral or prudential reasons for forbearance or action. Unless people were capable of understanding and then using legal rules as premises in deliberation, law would be powerless to affect human behavior.⁹ This view assumes that law is sufficiently knowable to guide conduct, but a contrary assumption is largely incoherent.

People use legal rules as premises in the practical syllogisms that guide much human action. No instinct governs how fast a person drives on the open highway. But among the various explanatory variables, the posted speed limit and the belief in the probability of paying the consequences for exceeding it surely play a large role in the driver's choice of speed. I am not suggesting that human behavior cannot be modified by means other than influencing deliberation or that human beings always deliberate before they act. Of course it can and of course they do not. But law operates through practical reason, even when we most habitually follow the legal rules. Law can directly and indirectly affect the world we inhabit only by its influence on practical reason.

The legal view of the person is not that all people always reason and behave consistently rationally according to some pre-ordained, normative notion of rationality. It is simply that people are creatures who act for and consistently with their reasons for action and who are generally capable of minimal rationality according to mostly conventional, socially constructed standards. The type of rationality the law requires is the ordinary person's common sense view of rationality, not the technical notion that might be acceptable within the disciplines of economics, philosophy, psychology, computer science, and the like.

Virtually everything for which we deserve to be praised, blamed, rewarded, or punished is the product of mental causation and, in principle, responsive to reason. I do not mean to imply dualism. I am simply accepting the folk psychological view that mental states—which are fully produced by and realizable in the brain—do play a genuinely causal role in explaining human behavior. Machines may cause harm, but they cannot do wrong, and they cannot violate expectations about how we ought to live together. Only people can violate expectations of what they owe each other, and only people can do wrong. Machines do not deserve praise, blame, reward or punishment. Machines do not deserve concern and respect simply because they exist. These concepts apply only to potentially acting, intentional agents. Folk psychology is the type of account of behavior that is crucial to judgments about responsibility. We praise and blame acting agents depending on their reasons for action.

Stone admits that the folk psychological model of the person and behavior may be the law's model, but he claims that it is not a “valid account of the human mind.”¹ I interpret this claim to mean that the folk psychological model is pre-scientific and empirically incorrect, but there is no argument presented in support of the claim and it is implausible. Virtually every neurologically intact person consistently has the experience of first person agency, the experience that one's intentions flow from one's desires and beliefs and culminate in action. Indeed, this folk psychological experience is so central to human life and so apparently explanatory that it is difficult to imagine giving it up or what type of discovery could give us good reason to do so, even if it were possible to give it up. As the eminent philosopher of the mind, Jerry Fodor, has written: …if commonsense intentional psychology were really to collapse, that would be, beyond comparison, the greatest intellectual catastrophe in the history of our species; if we're that wrong about the mind, then that's the wrongest we've ever been about anything. The collapse of the supernatural, for example, doesn't compare…. Nothing except, perhaps, our commonsense physics…comes as near our cognitive core as intentional explanation does. We'll be in deep, deep trouble if we have to give it up…. But be of good cheer; everything is going to be all right [Ref. 10, p xii].

Moreover, the folk psychological theory has much explanatory power and is capable of scientific investigation. Indeed, within psychology there is a growing recognition that mental state influence is a fundamental tool of social cognition.¹¹

The plausible theory of mind that might support such explanations is thoroughly material, but nonreductive and nondualist. It hypothesizes that all mental and behavioral activity is the causal product of lawful physical events in the brain, that mental states are real, that they are caused by lower level biological processes in the brain, that they are realized in the brain—the mind-brain—but not at the level of neurons, and that mental states can be causally efficacious.¹²

In addition, there is a perfectly plausible evolutionary story about why folk psychology is causally explanatory and why human beings need rules such as those the law provides. We have evolved into self-conscious creatures who act for reasons. Indeed, the ability to understand the intentions of another creature has been demonstrated in new world monkeys that evolved 40 million years ago.¹³ Practical reason is inescapable for creatures like ourselves who inevitably care about the ends they pursue and about what reason they have to act in one way rather than another.¹⁴ Because we are social creatures whose interactions are not governed primarily by innate repertoires, it is inevitable that rules are necessary to help order our interactions in any minimally complex social group.¹⁵ Human beings have developed extraordinarily diverse ways of living together, but a ubiquitous feature of all societies is that they are governed by rules addressed to beings capable of following those rules. As Fodor notes, one of the most basic, well-justified assumptions about human nature is that we are consciously intentional creatures who are capable of a great deal of rationality. At the very least, we remain entitled to presume that conscious intentions are causal and to place the burden of persuasion at a very high level on those who wish to substitute another account. Stone has much work to do to convince us that the folk psychological view of personhood and behavior is invalid.

Now, let us consider a few simple examples of behavioral legal criteria. Competence to contract, roughly speaking, requires that the agent be capable of understanding the nature of the deal and its consequences. Competence to stand criminal trial requires that the defendant be able rationally to understand the charges and the proceedings and be able reasonably to assist counsel. Prima facie criminal liability requires that a defendant with reasonably integrated consciousness intentionally acted (or omitted, in cases of duty) with the requisite mental state. For a final example, a criminal defendant is legally insane if, as a result of mental disorder, the agent did not understand what he or she was doing or did not understand that it was wrong, or could not control himself. Note that all the legal criteria are not self-defining and are normative. How much understanding and what kind is necessary for competence or legal insanity are normative, legal requirements, not scientific or clinical matters. There is no clinical or scientific answer to these questions.

Even the mental disorder criterion, when the presence of mental disorder is a legal requirement, is also behavioral. As mental health professionals know, virtually all Diagnostic and Statistical Manual of Mental Disorders (Text Revision) (DSM-IV-TR)¹⁶ criteria are behavioral. For those diagnoses for which an organic abnormality must at least be posited, there must still be abnormal behavior or there is no warrant for a diagnosis of mental disorder. Finally, people with mental disorders have desires and beliefs and form intentions, just like people without mental disorders. They may not be rational or they may have trouble controlling some of their behavior, but the folk psychological model of explanation applies in full. Irrational or compelled action is still action and not mechanism.

In other words, folk psychological behavioral criteria are always the final legal pathway, the final standard that must be addressed, the ultimate legal question. All evidence, including what caused the behavior, must help answer the folk psychological questions that the law asks. The law concerns acting agents, not mechanisms.

Let me give the example of Roper v. Simmons¹⁷ and the evidence concerning myelination of the adolescent brain. In 2005, the Supreme Court categorically excluded capital punishment of murderers who killed when they were 16 or 17 years old. Many professional organizations, including the American Academy of Psychiatry and Law, urged the Supreme Court to rely on quite recent and indisputable neuroimaging data that indicates that the human cortex continues maturing until the middle to late 20s. The Supreme Court did not cite this evidence. At most the Court referred to it indirectly, but it did cite behavioral science that demonstrated that adolescents are on average less rational than adults.

In my opinion, the Court was right not to rely on the neurological data.¹⁸ The Court held that adolescents who commit capital murder can not be put to death because they are insufficiently rational to deserve capital punishment. The Court thus recognized that rationality, a behavioral characteristic, is the touchstone criterion of retributive desert. Once the rationality difference was shown directly—and the behavioral science simply confirmed what every parent knows, as the amicus brief of the AMA conceded—the neurological evidence simply provides a partial causal mechanism for why this difference is observed. It provides no directly legally relevant information. Even if no neuroimaging data were available, once the behavioral difference was demonstrated, the Supreme Court had a completely sufficient factual basis to hold as it did. If no behavioral differences were demonstrable, then the brain evidence would have been beside the point. Brains do not kill people. People kill people. We do not praise and blame, reward and punish brains. Rationality, not myelination, is the basis of responsibility.

The conclusion to be drawn from this description of the law's folk psychological theory of the person and behavior and of the positive law—the law as it is—is that many of the problems that Stone claims bedevil forensic psychiatrists are really non-problems. There is no looming deconstruction of self and agency. There is a metaphysical free will and determinism issue, but it is irrelevant to the practice of forensic psychiatry.¹⁹ Free will, for example, is neither a criterion for any legal rule nor a foundation for responsibility practices generally. How the brain enables the mind is indeed an extraordinarily difficult conceptual and empirical problem. If we ever discover how this is possible, it is likely to transform our view of biology and ourselves, but for now we have no idea how this happens.²⁰ The mind-brain problem is irrelevant to the practice of forensic psychiatry, because the final pathway is always folk psychological mental states. There is no problem in blending talk of scientific causation with questions about agency. The sophisticated forensic psychiatrist simply needs to understand that any scientific information, such as mechanisms of causation, must somehow be relevant to assessing the folk psychology. Causation per se, at any level of scientific explanation, including abnormal causation, is not an excusing or mitigating condition, and causation is not the same thing as compulsion.²¹ The question is whether folk psychological criteria were met or not.

Again, it will not do to say that folk psychology is an unsophisticated view of human behavior. Any forensic psychiatrist who believes this, however, must nonetheless either remain within the confines of folk psychology as the final pathway or seek alternative means of putting bread on the table. It will also not do to object that the scientifically trained physician often uses the language of mechanism to understand the patient's disorder and to communicate with colleagues. All that is well and good. But if the psychiatrist wishes to practice forensics, he or she must be willing to translate such language into folk psychological concepts. Until agency is shown to be an illusion, folk psychology will be with us. Indeed, rule-following makes no sense without folk psychology. Forensic psychiatrists, as informed citizens, may wish to convince the legal system that agency is an illusion, but they should not do it by trying to smuggle mechanistic thinking and conclusions into their forensic work.

The Proper Role for Forensic Psychiatrists

As the examples of mental health law provided thus far demonstrate, the forensic practitioner's central task must be to help legal decision-makers evaluate mental states. My conclusion is that forensic psychiatrists have a modest but genuinely helpful role to play if they understand the questions and are sensitive to the limits of their clinical judgments and data base. I shall use the insanity defense as my example of the general role we should play because it is familiar to everyone, Stone uses it, and Stone suggests that there is nothing I could have added to the resolution of the Weinstein/Cystkopf case (described later) as a forensic psychologist. The next section suggests more specific forensic do's and don'ts.

Let us start with why we excuse some people with mental disorders who commit crimes. It is not simply because they have a disorder and it is not simply because a mental disorder may have played a causal role. The reason that we excuse some defendants with disorders is that they were sufficiently incapable of rationality or incapable of controlling their behavior in the context in question. (Indeed, I believe that these are the only two reasons that some people with mental disorders sometimes receive special legal treatment in any area of the law.) Stone may not agree that lack of rationality is the best interpretation of the insanity defense, but then he must provide an alternative interpretation buttressed by an argument in favor of that interpretation. Whatever criteria are ultimately adopted, they must be folk psychological and cannot be simply answered by the presence of a mental disorder. They must be assessed independently within the folk psychological model.

The role of the forensic psychiatrist is therefore to help the finder of fact understand what the defendant's mental state, broadly understood, was at the time of the crime. The finder of fact can then use this information to decide whether the law's normative criteria are met. Consider Weinstein/Cystkopf, whose defense team wanted to use neurological evidence of frontal lobe damage from a subarachnoid cyst pressing on the frontal cortex to support a claim that Cystkopf was legally insane because he could not control himself. The implication was that Cystkopf was just a mechanism when he killed his wife, and that he was not an acting agent. The case was plea bargained right before the trial, but the trial judge had already ruled that all the neurological data would have been admissible to support the claim of legal insanity.²²

Stone finds my previously published counternarrative of agency in the Cystkopf case²³ compelling, but claims that I have no theory of legal insanity. He is right and wrong. A story about agency is compelling because it is the only possible story within the law's folk psychological model. There may be normative disputes about the proper folk psychological criteria in any legal context. For example, should or should not the insanity defense contain a “control” prong? What kind of rationality and how much are necessary for responsibility? All such disputes must be resolved within the folk psychological model of the person. If Stone wishes to abandon the folk psychological model for another, it will require both an argument—not an assertion—that folk psychology is invalid and a normatively desirable argument for what should replace it.

Stone is wrong about my needing to have criteria for legal insanity in my role as a forensic psychologist. In that role, I do not need such criteria because deciding what the legal criteria for legal insanity should be is not a scientific or a clinical question and normative standards always leave room for discretion. Contrary to Stone's assertion, who is legally insane is a question that simply cannot be answered with psychological or scientific precision because it is not a psychological or scientific question. I may have such a view as a citizen or as a lawmaker, but not as a forensic psychologist. All that I (or any other forensic practitioner) need is an understanding of the basis —lack of rational capacity—for why legal insanity excuses and then our task is to provide evidence that addresses this question and helps the finder of fact resolve it. If the forensic practitioner gives an ultimate legal conclusion, it implies that the practitioner does have legal criteria that he or she is applying and should be prepared to defend. As I argue later, however, forensic practitioners should not give such a conclusion and I do not do so. Therefore, once again, I need not have any criteria for legal insanity in my role as a forensic psychologist.

In many cases, the forensic psychiatrist evaluating the defendant weeks or even months after the crime, may not have much to add to the observations of contemporaneous observers concerning the defendant's rationality. Even then, however, a good clinician may be helpful in retrospectively assessing the defendant's mental state. Clinicians can ferret out hitherto unidentified perceptions or beliefs and help make a coherent whole of the defendant's mental experience.

Moreover, the forensic clinician's empirical knowledge base can often be helpful—assuming, of course, that the asserted facts are clinically or scientifically warranted. Consider, for example, Eric Clark, the defendant involved in the Supreme Court's recent decision upholding the constitutionality of both Arizona's extremely narrow insanity defense and Arizona's exclusion of most expert mental state evidence introduced for the purpose of negating mens rea.²⁴ Mr. Clark indisputably had paranoid schizophrenia, but there was dispute about whether he delusionally believed that his victim was a space alien impersonating a police officer. If Mr. Clark's evidence about these beliefs had been legally admissible to negate mens rea and had been believed, he would have lacked the charged intent to kill a human being with knowledge that the victim was a police officer. After all, Mr. Clark allegedly intended to kill a space alien. Although no one other than Mr. Clark—assuming infallible memory—can be certain what he believed the night of the homicide, the statistical probability that paranoid schizophrenia would produce that particular type of delusional belief is both probative evidence and not the sort of information a layperson would have.

Whether a defendant could have controlled himself, independent of whether the defendant was not rational, is a particularly fraught issue about which I believe forensic clinicians often do go beyond their expertise.²⁵ (I prefer to treat these cases as rationality problems, but that's another matter.) Forensic clinicians need to be especially cognizant that they are not talking about mechanisms. In these cases, there is no question that the defendant acted. Where the defendant does not act at all, such as in the relatively rare cases of automatism or sleepwalking, prima facie liability is defeated ab initio, and there is no question of a control problem. The genuine control question then becomes largely a matter of common sense because we have no valid scientific measure for distinguishing did not versus could not. We need to ask questions of the following sort: How does this defendant describe his subjective experience of not being able to help himself, and how do others with similar problems describe their subjective experience? Clinicians can be very helpful in assessing this. Does the person act at considerable cost to himself, even when the chance of detection is substantial? Are there any scientific data to support the behavioral observations and inferences? Again, compared to lay witnesses, forensic clinicians have special observational skills, scientific knowledge, and ability to organize these elements into a coherent, folk psychological account.

Let's apply this analysis to Cystkopf, who claimed he could not control his behavior when he was enraged. Had he ever been enraged before? What happened? Enraged since? What happened? Here's what I can tell you about this prosperous businessperson. If he had severely and derangedly lost behavioral control on previous occasions, he would have been in trouble with the law or would have been worked up prior to killing his spouse. There was no such history, however. Alas, even otherwise good people sometimes lose it. Now, Cystkopf also had a benign subarachnoid cyst that was pressing on his left frontal cortex and that was apparently producing some physiological abnormalities in that region. Experts, and not laypeople, know that lesions in this area can predispose people to problems with inhibition and judgment, and they may have data about the relation of such lesions to antisocial behavior. Indeed, validated neuropsychological assessment techniques can help pinpoint subtle behavioral abnormalities flowing from suspected lesions that may be relevant to folk psychological criteria. Thus, the consequences of the cyst may have been a bit of further confirming evidence of Cystkopf's difficulty controlling himself, if, and this is a big if, it was consistent with rich behavioral evidence. If the homicidal rage was a one-time occurrence, then the neuroevidence tells us nothing terribly probative and it is likely to be misleading. Moreover, even if frontal lobe abnormalities played a partial causal role, Cystkopf was still an acting agent when he killed his spouse.

Now let me turn to a major counterexample to Stone's assertion that we lack legally relevant science: the question of predicting future violent behavior, which is important for involuntary civil and quasi-criminal commitment, capital punishment, and civil liability. Forensic psychiatrists and psychiatrists in general have had a vexed history with this issue. On the one hand, on the ground that they can reasonably successfully make such predictions, they generally desire the ability to hospitalize potentially violent people. On the other hand, they have resisted claims that they can predict violence with sufficient accuracy to support imposition of capital punishment or to support liability for failure to warn potential victims of a patient's threatened violence. Now, the degree of accuracy normatively necessary to support any of these predictions can vary from context to context, but the law will continue to ask forensic practitioners for predictive judgments and accuracy will always be a problem.

Prediction is such an important matter with such profound civil liberties implications that it is fortunate that research over the past half century has taught us a great deal about predicting violence. Forensic psychiatrists traditionally made predictions based on seat-of-the-pants clinical reasoning, but ever since the 1954 publication of Dr. Meehl's slim but epochal volume, Clincal vs. Statistical Reasoning,²⁶ it has been clear that actuarial or statistical prediction methods are superior to clinical prediction and that very infrequent events, such as extreme violence, are difficult to predict accurately by any means. Research performed since Meehl's book appeared has confirmed his conclusion, although it has also shown that clinical prediction is often better than previously thought.²⁷ Statistical methods are especially superior when the behavior to be predicted is violence.²⁸

At this point, there are numerous actuarial tools available to be used to make predictions,²⁹ and we have learned a great deal about the magnitude of the contribution of various measurable variables, such as psychopathy, to the risk of violent conduct among people with mental disorder.³⁰ Using actuarial methods, forensic practitioners can predict violence at well above chance levels, especially among high risk populations. Predictive accuracy does not yet approach perfection—false positives and negatives are still a problem—but it is vastly better than formerly. It is well nigh inconceivable that this valid research base is not relevant to forensic psychiatric practice when the law requires a prediction. I will presently describe how I believe forensic practitioners ought to present such data.

Before I leave the topic of prediction, let me clarify an important conceptual point about predictability, agency and responsibility. Predictability is not inconsistent with the folk psychological account of agency and responsibility. An acting agent who is responsive to reason may be entirely predictable. When I do demonstrations at professional meetings that require audience cooperation, it is completely predictable that audience members will be cooperative and polite, but they surely are agents who deserve my thanks for their efforts. Even if the prediction technique in question uses demographic or other non-folk psychological variables, it is human action and not mechanism that is being predicted.

In sum, forensic experts have much to contribute to a thick description of the folk psychological mental states that are the criteria to be addressed and on occasion they may have clinical or scientific data that are relevant to the folk psychological account. I agree with Stone that we often have less science relevant to legal criteria than some forensic practitioners suppose, but as the examples I have provided demonstrate, there is more than Stone admits. Finally, it is hard to imagine what argument would persuasively indicate that clinicians do not have special skills in understanding mental abnormalities and the role they play in folk psychological accounts.

Forensic Do's and Don'ts

Let me conclude with a short list of forensic do's and don'ts, all of which flow from the model I have presented and should insulate the forensic clinician working within the standard model or any other against virtually any ethical criticism. Space constraints do not permit me to lengthen the list and explain many of the recommendations more fully, but the recommendations and arguments presented will give you a sense of the modest but important role I believe we should play. I first developed this way of approaching forensic work in 1978,³¹ and continue to believe, based on experience and research, that it is a valid guide to ethical practice because adhering to it will cause forensic practitioners to convey the most legally useful information and to stay within the bounds of their expertise.

• Before offering any consultative or testimonial evidence, be sure that it is legally relevant. You should always be able to explain precisely why any bit of evidence is legally relevant.

• Never talk about free will in reports or testimony. It is not a legal criterion and we have nothing useful to add to the metaphysical debate. Too often, when we use the language of free will we appear to be using it as an explanatory tool, but doing so is simply a way of begging the question of responsibility. Free will or its lack does no explanatory work in forensic practice.

• Always try to give the thickest possible description of the subject's relevant mental states. At the relevant time, what did the defendant perceive, desire, feel, believe, intend? What were his or her reasons for action or inaction? If particular behavior is in question, when did it begin and is it situationally variable? And so on. What behavioral data are necessary for the thick description follow logically from the nature of the folk psychological question being asked.

• In appropriate cases try to obtain, if possible, observations of the subject by others who had a good opportunity to observe, such as family members, friends, coworkers, and the like. In other words, it is virtually always valuable to triangulate by checking the account you derive from a clinical interview with data from observers who had real world contact with the subject. This is especially valuable if you are doing a retrospective competence or responsibility evaluation.

• Avoid giving diagnoses, if possible. I know from long experience that most forensic practitioners viscerally reject this recommendation. Diagnoses are central to clinical work and professional communication. Why reject their use in forensic work? As Loren Roth memorably said to me when, in the wake of United States v. Hinckley,³² I recommended in congressional testimony that diagnoses should be excluded in federal insanity defense cases, “You've got to let psychiatrists be psychiatrists.” Well, of course I do when the psychiatrists are working in the office, clinic, and hospital, but hear me out nonetheless. These are the reasons I wish we would not use diagnoses in forensic work. The first is that the criteria for all diagnoses are behavioral. Second, a diagnostic label is vastly underinformative as a proxy for the subject's behavior. As DSM-IV-TR¹⁶ recognizes in its introduction, each diagnostic category can encompass extremely heterogeneous behavior. Third, as DSM-IV-TR also warns, these categories were not created to answer legal questions. Thus, the diagnostic label conveys little if any legally relevant information. The law needs to understand the subject's behavior, not the label we professionals place on the agglomeration of behavioral signs and symptoms that we call a disorder entity. There can of course be disputes about the underlying behavior, but the underlying behavior is more concrete, empirical and less inferential than the diagnosis, so resolving such disputes should be easier than resolving diagnostic disputes. Even if these arguments are correct, you might ask, what's the harm? In my view, the harm is potentially misleading and confusing the jury and wasting the law's time. Laypeople consider the signs and symptoms of diseases to be mechanisms and mechanisms are never responsible or competent, for example. The use of diagnostic labels facilitates the lamentable slippage from folk psychological understanding and criteria to mechanistic thinking, thus threatening to beg the legal question. Further, what is the point of opposing experts taking time to quibble over diagnoses, when even a correct diagnosis conveys little relevant information? Why does it matter whether John Hinckley, Jr, for example, had schizophrenia, schizotypal personality disorder, schizoaffective disorder, or any other? The real question is what he really believed and what he really intended to accomplish by killing President Reagan. Asking the judge or jury to find the pea in the diagnostic shell game simply distracts them from the primary task of focusing on the behavior. Moreover, the skilled forensic practitioner does not need diagnostic labels to convey further, broader information about a subject. Suppose, using clinical or more rigorously established data, the practitioner is trying to explain how people with psychotic mental states behave generally. He or she can simply say, “people like the subject who (hear voices, have strange beliefs …, [fill in the blank]) also do or do not in general behave the following ways …”. For example, “people like Eric Clark who have unfounded and profoundly abnormal beliefs that they are in danger do/do not usually also believe that their supposed enemies are space aliens.” In sum, the recommendation is simply to describe the legally relevant behavior and further legally relevant clinical and scientific information without using a diagnosis. Try it. You'll like it.

• Avoid using jargon and technical terms. If you must use them, define them in common-sense terms, try to find an ordinary language synonym, and stick to the synonym. This avoids confusing the finder of fact and avoids charges that you are trying to hornswoggle the judge or jury by pretending to greater technical or scientific precision than actually exists.

• Whenever you assert a statement about a “general psychiatric or social scientific fact,” make clear the database from which the statement is derived. Is it your clinical experience? Large-scale epidemiological studies? Valid experiments? And so on. If there is conflict in the literature or between your clinical experience and the empirical literature, explain why you accept one position or why you reject the larger database.

• Always translate your evidence, from whatever level, into the language of folk psychology. Be especially careful to do this when citing confirmed empirical data. When the issue is the subject's ability to control his or her relevant conduct, be especially careful not to argue by label, to smuggle in the language of mechanism, or to beg the question. The process of deciding, for example, whether an addict had control over his or her seeking and using conduct is not concluded simply by saying the subject is an addict. Nor will it do simply to adduce a genetic predisposition or frontal cortex abnormalities associated with addictions. You must be able to show behaviorally in folk psychological terms that the subject had serious difficulty controlling his or her behavior. To do this properly, you need a good folk psychological model of what a control problem is. This is a very fraught question. Tread very gingerly when addressing control issues.

• When a prediction is in question, you should be aware of the data and preferably should use an actuarial method for making the prediction. Use frequentist rather than probabilistic language when possible, because virtually all people are better at understanding and using the former. Be clear about the limits of the prediction, with attention to base rates. If the prediction involves a very broad category of behavior—such as violent conduct, which can range from a slap to a homicide—reveal it. Where base rates are important, make sure you are attentive to them. Very infrequent behavior will be exceedingly difficult to predict accurately even if a variable is present that substantially increases the risk. Suicide, for example, is extremely infrequent, even though it is vastly more frequent among those with major depressive disorders. Thus, it will be difficult to predict accurately even among those with such disorders.

• Do not give an ultimate legal opinion. Virtually everywhere and for all questions, we are permitted and often encouraged to give an ultimate legal opinion as a matter of reasonable psychiatric (or medical, or psychological) certainty. An exception, of course, is the prohibition against giving an ultimate legal opinion of insanity in federal criminal trials. The reason for this recommendation, with which Jay Katz also agreed, is that the ultimate question is not a scientific, psychiatric, clinical, or medical question: it is legal and normative. There can be no reasonable psychiatric certainty about the opinion because it is not psychiatric. The expert offering his or her opinion is doffing the white coat and entering the jury box as the 13th juror. Of course, you may have such an opinion as an ordinary citizen and could vote that way if you were a juror, but, in my view, you cannot hold that opinion as a matter of professional expertise. It does not matter how much better you understand the legal criterion in question than the lay finder of fact. It does not matter how many cases of this type you have handled. The opinion is not an expert opinion. Period. Give the judge or jury all the underlying, legally relevant evidence they need. That's your professional expertise; that's your job. Then let the judge or jury make the normative legal judgment. That's their expertise; that's their job. If you are forced by local law to give an ultimate legal opinion, try locutions of the following sort: “That is really a question for the jury/judge, but, in my experience, people like this person are/are not found to meet the standard.”

This list of do's and don'ts is of course aspirational and there is a practical problem in adhering to it. Most lawyers who retain you will insist that you violate these admonitions and will rightly claim that it is not an ethical or legal violation to do so. Thus, if you inform counsel who seeks to consult you that you will not offer a diagnosis or an ultimate legal opinion, for example, the lawyer may decide not to retain you. Indeed, I follow my own admonitions and a very frequent response from the attorney is to ask for a referral. I can afford this because, like the example Stone cites from Forensic Ethics and the Expert Witness (FEEW),³³ I have a tenured position and do not depend on forensic work for my living. Consequently, it is not difficult for me to adhere to these recommendations. But for those for whom forensic work is a substantial part or the entirety of their livelihood, following these recommendations can be a real problem. All I can advise is to try to follow as much of it as you can. The more of us who behave this way, the more normative it becomes.

Conclusion

I often begin talks to forensic psychiatrists that recommend a modest but important role with the following provocative riff. Let me understand this. You went to medical school for four years so that you could learn to take care of suffering people. You then went for three to four years of postgraduate specialty training in psychiatry so that you could learn to take care of suffering people who had mental disorders. And you did all this so you could spend your time [pause] consulting with lawyers and testifying in a courtroom?

The point, of course, is that there are losses to suffering people and to society when professionals trained to take care of patients spend their professional energies in nontherapeutic pursuits. It is indisputable, however, that many doctors engage in professional activities that contribute to social welfare but that do not involve taking care of patients. Indeed, some doctors even spend much of their professional lives teaching students of the law! Similarly, many people trained to take care of the legal problems of others do not practice law. This is all well and good. Professional training and skill can be used for many good purposes that do not contribute to the welfare of individual patients and clients.

In short, be of good cheer. Forensic psychiatry has an ethical basis, and if your contribution is suitably modest and qualified, you have much to contribute.