History is a dangerous tool, especially for anyone not steeped in the period from which lessons are drawn. Nevertheless, we ignore history at our own peril, for the historical background of the insanity defense is an important part of the context in which it operates. Moreover, even a cursory survey of the historical record reveals an underlying ambivalence between continuing theoretical notions of criminal behavior and the application, in practice, of these notions to offenders brought before the courts.
We know nothing about the first stirrings of the insanity defense in the western world, and little about its use even through classical times. Ancient Hebraic law recognized that deafmutes, idiots, and minors were not responsible for their actions,17 but as to the practical application of such tests we know virtually nothing. Similarly, classical Greek law explicitly distinguished between intentional and unintentional homicide (holding the former the greater wrong). The distinction was attacked on theoretical grounds by Plato18 (who nevertheless admitted the practical utility of the distinction), but justified by Aristotle,19 who reasoned that the capacity for choice was critical to the question of moral blameworthiness, and that this capacity was lacking in animals, children, and insane persons.20 Current experience, however, suggests caution in assuming that theoretical discussions bear much relationship to what actually goes on in the courts. Probably the most that can be said about Hebrew and Greek law is that both recognized that some persons should not be held responsible for actions that would ordinarily be punishable, and that the category of the irresponsible may have been roughly defined by reference to the capacity of children or animals.
Classical Roman law took a generally protective attitude towards those thought to be mentally disordered, relieving them both of the capacity to govern their affairs and of responsibility for wrongful actions.21 Again an analogy is drawn to children,22 but by the time the Emperor Justinian caused a codification of the laws in the sixth century A.D., the shoe is on the other foot: the insane person, who "does not know what he is doing,"23 is entirely incapable of action with legal consequences, whereas the child, although "not very different from a madman," may in some cases be allowed at least to make contracts.24 Again, materials on actual practice in the law courts are scanty indeed; but the Roman materials, intended for the guidance of practicing lawyers or as codifications of existing practice, may safely be assumed at least to attempt a reflection of reality, and not merely of principle. The Roman practice of denying capacity as well as responsibility to the insane suggests that acquittal by reason of insanity would result in mandatory tutelage, if not mandatory commitment- the first explicit appearance of one of the abiding themes of the insanity defense throughout the ages.25 The insane may not be "punished," but neither will they be suffered to wander free of controls not imposed on the ordinary citizen.
With the breakdown of Roman institutions following the decline of the empire, administration of the criminal laws passed to the hands of the local authorities where it existed in any formal sense at all. In England, after the end of Roman control, the legal background seems to have been Germanic customary law, often said to have imposed criminal liability without regard to fault. Here, however, it is important to remember that whatever the theoretical basis of "criminal liability," its most important application was to compensate the victim or his family for an injury done: "older Anglo-Saxon law . . . considers crime mainly as a tort."26 The question therefore is not whether an insane person is to be blamed and punished, but rather which of two (perhaps equally blameless) persons shall bear the cost of an injury. Yet even in such a system, notions of fault were not entirely eliminated. William Holdsworth ascribes to pre-Conquest times the notion that although technically guilty of a crime, a lunatic "may deserve a pardon."27 By the thirteenth century, homicide was a capital matter, no longer capable of satisfaction by payment of weregild (to the family of the deceased) or fines to the court; and in such cases the king was with some regularity granting pardon for homicide committed while of unsound mind, although the standard actually applied "all depended upon the king's 'grace.' "28 Thus in 1278, we see Edward I ordering that one Hugh de Misyn, convicted of hanging his daughter Cicely, be released into the custody of 12 men of the county, because he committed the murder "while suffering from madness." The release, it should be noted, is not absolute, and his bailors (with cash, at least, at hazard) shall see to it that he "shall not hereafter injure anyone."29
As time went on, the king's grace began to be administered as a matter of course, during the trial itself;30 attempts to formulate standards appear to have drawn at the outset on the common law's distinction (formulated for the purpose of determining when the lord could take his vassal's estate in wardship, and the financial incidents of such a taking) between the idiot, born without understanding, and the madman, whose misfortune befell him later in life.31 For purposes of the criminal law, however, account had to be taken of the person who appeared to fluctuate between madness and sanity. Thus in the seventeenth century, Sir Edward Coke speaks of the born idiot and the madman (who "wholly loseth his memorie and understanding"), neither of whom is responsible for his actions, and also of the lunatic, who has intervals of lucidity (during which he is criminally responsible for his actions) as well as periods of insanity (during which he is not).32 Coke's definitions have a pleasing neatness about them, all the more so because the critical question is left untouched: there is nothing to explain how to distinguish the (not responsible) born idiot from a (responsible) born fool, how to distinguish the irresponsible madman from the responsible, vicious criminal.33
It is tempting to ascribe some of this unconcern for definition to the nature of the evidence that would be presented at trial: lay testimony from friends and neighbors focusing on the accused's behavior, rather than expert testimony presenting conclusions about the inferences that should be drawn from that behavior. Such an hypothesis, however, loses much of its force on examination of the writings of Sir Mathew Hale (Chief Justice of the Court of King's Bench), found among his papers at his death in 1676. Hale's theoretical basis is a familiar one: human beings are "naturally endowed with these two great faculties, understanding and liberty of will . . . and consequently obnoxious to guilt and punishment for the violation of . . . law, which in respect of these two great faculties he hath a capacity to obey: The consent of the will is that which renders human actions either commendable or culpable. . . . And because the liberty or choice of the will presupposeth an act of understanding to know the thing or action chosen by the will, it follows that where there is a total defect of the understanding there is no free act of the will. . . ."34
Hale, however, explicitly recognizes the difficulty in devising rules for the practical application of his test: "for doubtless, most persons, that are felons of themselves, are under a degree of partial insanity, when they commit these offenses."35 He concludes only that division between the responsible and the irresponsible "must rest upon circumstances duly to be weighed and considered both by the judge and jury" in each case; the "best measure that I can think of" is whether or not the accused "hath yet ordinarily as great understanding as ordinarily a child of fourteen years hath."36
Except perhaps for his "somewhat fatuous"37 concluding suggestion that the accused be lined up beside an ordinary child of 14 for judgment, Hale's discussion has a distinctly modern and practical ring. The tidy legal categories must be applied, in practice, to the very untidy defendants appearing before the courts. And the difficulty of application, to Hale, is not merely a problem of proof, arising from the fact that courts must decide cases on the basis of what evidence they can get, rather than what evidence they would like to get. It is a problem inherent in the nature of man, for this "insanity," whatever it be, whatever its causes, is not an entity absent from all responsible men and present in all the irresponsible. It is "doubtless" present to some degree in most felons, and perhaps even (though Hale never explicitly takes this step) in some or many of the good citizens who nevertheless manage to avoid criminal behavior entirely. To Hale, then, the problem of definition is not merely a problem of defining whnt will excuse, but a problem of how much. Given the imprecision of language, close cases will exist no matter how fine the definition, or how full the evidence.
Hale's treatise was not published until 1736. In the meantime, trial judges had apparently gotten in the habit of charging the jury that a defendant was to be held responsible for his actions unless he lacked the capacity to distinguish "good from evil"38 Later cases continued this formula, occasionally substituting the words "right from wrong."39 What was meant by these words, however, is far from clear, and the apparent assumption that this was a narrow test may be more a reflection on the capabilities of counsel for the accused (where he had one) than upon considered choice made by the judges. Good lawyers apparently felt free to argue for a broad reading of the rules, and were at least occasionally successful. In May, 1800, one James Hadfield fired a shot at King George III as the King was entering the royal box of a theater. Hadfield apparently believed that he had been commanded by God to sacrifice himself for the world's salvation, and settled upon an attempt to murder the King (high treason) as the surest way of bringing about his own demise. His counsel, Thomas Erskine, argued to the judges that language in prior cases treatises indicating that a "total deprivation of memory and understanding" was necessary to relieve a defendant from criminal responsibility could not be taken literally. "[Delusion], where there is no frenzy or raving madness, is the true character of insanity," and Hadfield would be exempt from criminal responsibility if he were suffering from such a delusion, and if "the act in question was the immediate unqualified offspring of the disease."40
Erskine's arguments apparently were persuasive, although the evidence he presented clearly indicated that Hadfield understood he was committing a capital felony (indeed, that was his sole reason for the act). Erskine had not finished presenting his case when both the court and the Attorney General (for the prosecution) agreed that the trial should be stopped. The jury was urged to return a verdict of not guilty by reason of insanity, and promptly did so.41 Hadfield was remanded to Newgate Jail, and the Parliament quickly passed an act requiring (as was apparently the current practice; Parliament seems just to have wanted to make sure) that a special form of verdict (not guilty by reason of insanity) be returned if the jury should acquit the defendant on grounds of insanity. The act further stated that upon such acquittal the accused should be kept in custody "until His Majesty's pleasure shall be known; and it shall thereupon be lawful for His Majesty to give such order for the safe custody of such person, during his pleasure, in such place and such manner as to His Majesty shall deem fit."42
Hadfield's Case settled only the question what should be done with Hadfield himself. Judges, in the rare case where insanity was alleged as a defense, continued to charge juries that the question was whether the accused had knowledge of good and evil, or sometimes of right and wrong. What was meant by these terms is unclear, and was rarely if ever explained to the jury. And juries, acting under such charges, continued to return verdicts of not guilty by reason of insanity even where the evidence (increasingly, the testimony of medical men), as in the case of Hadfield, suggested that the accused knew that his actions would be considered violations of moral and secular law by his countrymen.
Then, on January 20, 1843, Daniel M'Naghten43 put a pistol to the back of Edward Drummond, private secretary to Prime Minister Robert Peel, and fired; a second shot was misdirected when a policeman seized M'Naghten's arms. Drummond died five days later, and M'Naghten was indicted for murder.
The trial (commencing March 3, before Lord Chief Justice Tindal, Mr. Justice Williams, and Mr. Justice Coleridge) was the subject of considerable public interest. Early newspaper reports of the shooting (not originally expected to be fatal) had been entirely at a loss to explain the reasons for it. Although M'Naghten had stated at his preliminary hearing two days after the event that "The Tories in my native city have compelled me to do this; they follow and persecute me," the London Times concluded on January 25 that M'Naghten had shown so much prudence and deliberation "that [the act] can hardly have been the conduct of a madman." Before the trial began, however, the papers were reporting "symptoms of mental aberration" that had begun "upwards of two years ago," and M'Naghten's counsel arranged to have him examined in prison by three doctors.
Court, counsel, and spectators were thus prepared to hear a claim of insanity, and lines were drawn in the opening arguments. For the Crown, the Solicitor General urged a narrow view- "To excuse [the prisoner] it will not be sufficient that ... he had a morbid delusion of mind upon some subjects, which could not exist in a wholly sane person; that is not enough, if he had that degree of intellect which enabled him to know and distinguish between right and wrong, if he knew what would be the effects of his crime, and consciously committed it, and if with that consciousness he wilfully committed it." Mr. Cockburn, for the defense, gave a different emphasis: "The question is not here, as my learned friend would have you think, whether this individual knew that he was killing another when he raised his hand to destroy him, although he might be under a delusion, but whether under that delusion of mind he did an act which he would not have done under any other circumstances, save under the impulse of the delusion which he could not control, and out of which delusion alone the act itself arose." The Court, for the moment, said nothing.
After the Crown had presented evidence of the killing, Mr. Cockburn introduced the doctors- Dr. Monro, Dr. Sutherland (jun.), and Dr. Bright. They had examined M'Naghten in prison on February 18, and he had described his persecution "by a system or crew at Glasgow, Edinburgh, Liverpool, London, and Boulogne . . . this crew preceded or followed him whereever he went ... he had no peace of mind, and he was sure it would kill him ... it was grinding of the mind . . . physicians could be of no service to him, for if he took a ton of drugs it would be of no service to him." He was followed, watched, and snickered at; beastly and atrocious allusions to his conduct were forever appearing in the newspapers; on one or two occasions "something pernicious" had been put into his food; and the crew was systematically destroying his health. The prisoner, the doctors said (all three were present at the examination and their testimony is substantially the same), "imagined the person at whom he fired at Charing Cross to be one of the crew- a part of the system that was destroying his health."
No, the doctors agreed, the delusions were not feigned; to M'Naghten, they were real. On cross-examination, the Solicitor General raised the question of productivity:
A person may have a delusion and know murder to be a crime?
- If there existed antecedent symptoms I should consider the murder to be an overt act, the crowning piece of his insanity. But if he had stolen a 10 pound note it would not have tallied with his delusion.
But suppose he had stolen the note from one of his persecutors?
- [Answer lost in the laughter.J
After the doctors had testified, the court inquired whether the Crown had any contrary evidence to produce; if not, the judges thought the case should be stopped. The Solicitor General had none, and the case was accordingly submitted to the jury, the Solicitor General informing them, "I cannot press for a verdict against the prisoner." In view of the language that has subsequently come down as the M'Naghten Rule (or Rules), Chief Justice Tindal's charge to the jury deserves some attention. He did not use the now-familiar formulation. Instead, he told the jury, "If [the prisoner] was not sensible at the time he committed that act that it was a violation of the law of God or of man, undoubtedly he was not responsible for that act, or liable to any punishment whatever flowing from that act. Gentlemen, that is the precise point which I shall feel it my duty to leave to you. ... I cannot help remarking, in common with my learned brethren, that the whole of the medical evidence is on one side, and that there is no part of it which leaves any doubt in the mind. ... I am, however, in your hands; but if on balancing the evidence in your minds you think the prisoner capable of distinguishing between right and wrong, then he was a responsible agent and liable to all the penalties the law imposes." Two things are significant about this charge. First, Tindal (and his brethren, for whom he spoke) did not believe that he was deviating from the ordinary formulation of the test. The question was whether the prisoner knew that his act "was a violation of the law of God or of man," whether he was "capable of distinguishing between right and wrong." Allowing for the usual variation in a day when "standard" jury charges were unknown and judges usually spoke extemporaneously, this is the same formula in use since the eighteenth century. Second, and perhaps more important, the judges thought it clear that M'Naghten met the test. He knew that he was firing a pistol, he knew that the object of the discharge was a human being, and his fondest hope at the moment of firing was to put an end to the life of the human being at whom he shot. When asked by the doctors why he had fired at the unfortunate Drummond, he answered that it was because Drummond had given him "a scowling look. At that moment all the feelings of months and years rushed into his mind, and he thought that he could only obtain peace by shooting him." Either the judges thought that such a state of mind was comprehended within the "right and wrong" test, or else they were unwilling to apply to the poor man before them a restrictive test devised at least in part to quiet hypothetical fears.
The outcome of the case caused considerable public controversy. Queen Victoria (who three years before had been shot at by a man later acquitted by reason of insanity) was furious·. "The law may be perfect, but how is it that whenever a case for its application arises, it proves to be of no avail." After debate in the House of Lords, the judges were called on the carpet and asked a series of questions relating to the insanity defense. In reply, they formulated what has come down to us as the M'Naghten Rule: the jury should be told that "to establish a defence on the ground of insanity, it must be clearly proved that, at the time of the committing of the act, the party accused was laboring under such a defect of reason, from disease of the mind, as not to know the nature and quality of the act he was doing; or, if he did know it, that he did not know he was doing what was wrong." This, they thought, was essentially the same as the charge actually given, with the slight advantage that the usual form, "put generally and in the abstract," might in some ways be misleading.
The Lords (and the newspapers) approved the judges' formulation, but the London Times presaged much future comment by observing that it was difficult to reconcile "with what took place at the late unfortunate trial." To say that M'Naghten himself was not acquittable under the rules bearing his name- as is often done, even today- is to make an assumption about the meaning of the words contained in the rules that is, I think, unwarranted. The judges, at trial, had found the case a clear one under the "right and wrong" test, and the jury apparently agreed. When called to explain, the judges said that they were only slightly restating the test, for clarity. Perhaps they were engaged in nothing more than an elaborate face-saving gesture; but given the ambiguity of the M'Naghten language, and the context in which it was formulated, it would seem more plausible that the judges were simply attempting to formulate more precisely (and cautiously, given the audience) what it was about Daniel M'Naghten that had convinced them he was not a fit subject for punishment.44
Notwithstanding public satisfaction, the M' ? lighten test was criticized almost from its inception- and even before, if we accept the proposition that M'Naghten was intended more to restate than to modify the "right and wrong" test, for Isaac Ray's Medical Jurisprudence of Insanity, first published in 1838, had called that test flatly "fallacious." Ray's argument, often since repeated in substance, was that "the insane mind is not entirely deprived of [the] power of moral discernment, but in many subjects is perfectly rational, and displays the exercise of a sound and well balanced mind." Therefore, presumably reading the "right and wrong" test as requiring incapacity to make any such discernments, he argued that it was unduly narrow. How accurate was Ray's reading of the American application of the "right and wrong" test is unclear. The American cases are as rare as the English ones, and even more sparsely recorded. Further complicating the issue is the question of burden of proof, for American judges were even more inclined than their English brethren to emphasize the then-prevailing rule that the defendant, in order to escape liability, must establish his irresponsibility "clearly," "beyond doubt," or "overwhelmingly." Some of Ray's objections may more suitably be seen as objections to this burden (or disagreement about whether it had been met) than to the standard actually applied.45 My own reading of the cases suggests that if they stand for anything, they stand for the application in practice of a sliding scale: strict for the poor and unwelcome, broad for the rich and well-connected. But the evidence is admittedly sparse.
In any event the Ray's criticismand later, of other medical critics of M'Naghten- bore little judicial fruit. With the exception of New Hampshire (which, largely under the influence of Ray, adopted a rule that would acquit if the act were the "offspring" of a mental disease), all of the American jurisdictions ultimately approved M'Naghten as a restatement of (and perhaps an improvement on) the "right and wrong" test previously used. In this connection, however, it is worth emphasizing again that nineteenth-century trial judges were accustomed to charging the jury not from standard books of jury instructions, as is now regularly the case, but on the basis of the evidence and issues as the judge saw them in the case at hand; accordingly, what at least some recent commentators would regard as critical differences in formulation often crop up in the reports. As time went on, an increasing number of states (more than a third of the whole by 1960) began to supplement the M'Naghten language, which began to assume a more rigid form. The added language called upon the jury to acquit the defendant if he had a mental disease that kept him from controlling his conduct. This rule, often called (and sometimes phrased as) the "irresistible impulse" test, for a time muted psychiatriccriticism where it was adopted. The honeymoon, however, did not last; and by the 1950s, psychiatric criticism of the control test was as harsh as that of M'Naghten.46
I have spent time on the early history of the insanity defense not out of simple antiquarianism, but because the persistence of certain characteristic features and, more important, of certain recurring problems is to me most suggestive. Where a problem persists under varying formulations of the defense over a period of several centuries, one may strongly suspect that reformulation of the applicable test (however advisable that may be on other grounds) is not going to make the problems disappear. And from the historical record as I have sketched it above, I would draw the following conclusions.
First is the point, perhaps obvious since its forceful presentation in recent times by Professors Joseph Goldstein and Jay Katz,47 that the defense of insanity has never been a defense resulting in freedom for the "acquitted" defendant. First kings, and later courts, have routinely operated on the assumption that a defendant acquitted by reason of insanity should be restrained, and not released. In the modern context, perhaps the most important consequence of this fact is that the insanity defense can be seen (and often is, particularly by the participants) as simply one of several dispositional alternatives, which may be utilized or not depending on the relative convenience of the several routes.
Second is that the insanity defense has always been a rare thing; but even so, where its use came to public attention, it has aroused interest and controversy far out of proportion to the number of cases in which it is of immediate practical importance. One may compare to the uproar over M'Naghten's Case the calm with which the country accepted the "benefit of clergy," a doctrine which by the eighteenth century allowed large numbers of persons a one-time exemption from punishment after conviction of crime. At times it could be (and was) used to allow a convicted murderer complete exemption.48 Yet benefit of clergy remained available until 1827, without substantial outcry- and even then was available to peers for 14 more years, until finally abolished in 1841. The defense of insanity, which if successful negated blame but resulted in confinement, aroused heated debate. Benefit of clergy, which admitted moral guilt but negated or mitigated (to a maximum of one year's imprisonment, plus forfeiture) punishment, did not.
Third is what appears to me a tendency, at least since the time of Hale, to phrase rules that appear restrictive on reading, but to apply those rules more loosely in practice. (This tendency, incidentally, may in part explain the frustration of psychiatric writers with the several legal tests, at least if one assumes that psychiatrists share the understandable view that lawyers ought to mean what they say.) Such, at least, seems to me the lesson of Hadfield's case, and of M'Naghten's. And it suggests that some of the demons often seen in broad formulations of the insanity defense are more a reflection of hypothetical, or symbolic, fears than of practical dangers attendant upon a broad use of the defense.
Fourth, and perhaps related to the third, is that problems of "productivity" (that is, relationships between the insanity and the unlawful action, ultimately to become Durham's bane) are at least as old as M'Naghten, and may well be almost as old as the use of expert medical testimony to establish the defense. This, in turn, suggests two hypotheses: first, that problems of productivity are by no means unique to tests of responsibility that explicitly require such a relationship and second, that the problem may inhere less in the formulation of the standard than it does in the inevitable problem of translation from psychiatric to legal terms. If this last is so (and I shall not here attempt more than to set it forth as a tentative hypothesis), then attempts to remove or palliate this tremendous obstacle may be better directed at the presentation of evidence than at the basic formulation of the test.
I began this section with the suggestion that history is a dangerous tool, and repeat that admonition here. The conclusions I have drawn above are put forth as nothing more than hints towards promising lines of inquiry with regard to more modern administration of the insanity defense. It is to the experience since Durham, and the problems exposed by that experience, that I now turn.