Early Interrogation Law
As we demonstrated in chapter 1, the law of confessions reflects the institutions that identify, or create, deviance. We also surveyed what we called “brakes” that help control those suspected of crime or deviance. Perhaps the most obvious is the concern about false confessions. Another is the autonomy interest of the suspect, consisting of a dignity interest and an interest in being free to make up one’s own mind whether to self-accuse of a crime. And there is a proportionality principle that causes the State to avoid brutal methods to obtain confessions to minor crimes; it also causes judges to draw back from imposing penalties that seem to outweigh the harm a defendant is accused of committing. Most of these brakes, or controls, can be seen as far back as Roman law.
According to German historian Adolf Friedrich Rudroff, torture “was used from time immemorial against slaves, partly to strengthen testimony, partly to extort confessions.”1 The Roman historian Titus Livy in the massive History of Rome recounts an examination of slaves in 210 B.C. during which “they all confessed.”2 Indeed, torture was routinely used during the Roman Republic to obtain confessions, but only from slaves.3
That torture could be used only against slaves during the Republic is evidence of the autonomy control. As Hepworth and Turner note, the Western law of confessions is premised on the notion of “the self-activated rational individual.”4 Roman citizens fit that category; thus, torture would have been deemed a violation of the right to make a rational decision. A slave, by contrast, “was a thing … [that] had no rights.”5 Yet even those who had no rights, who were not considered self-activated rational individuals, would be protected by the false confession parameter that was developing in Roman law.
By the early years of the empire, judges were cautioned to observe safeguards thought to prevent false confessions. Caesar Augustus, who ruled Rome from 27 B.C. to 14 A.D., cautioned that the investigation of crime “should not begin with the application of pain, and that reliance should not be placed entirely on torture.”6 The infliction of pain should be used only on those “so close to being proved [guilty] by other evidence” that the confession “appears to be the only thing lacking.”7 Hadrian, emperor from 117 to 138 A.D., embraced the same notion and added the observation that some “have so little endurance that they would rather tell any kind of lie than suffer torture.”8
During the second century, the Roman Empire’s peak, the use of torture was extended to free persons.9 It is unsurprising that a judicial system ultimately answerable to an emperor would degrade the autonomy of defendants and witnesses. The use of torture not only expanded beyond the class of slaves during the empire, but it also became more frequent. “Judges were always prone to think that torture was a short cut to the truth and to start torturing witnesses and the accused right away; this became all the more common when torture was made applicable to free persons.”10 The emperors “frequently warned [provincial] governors against this tendency, telling them not to start by torturing witnesses but to try to establish a case by argument, and in particular not to torture an accused unless there was a prima facie case against him.”
All sorts of reasons might explain why torture was first applied to free persons during the empire and invoked more frequently. The empire needed efficient judicial outcomes, and torture was an efficient way to produce confessions. It is also conceivable that judges increasingly used torture as the empire grew larger and accumulated more enemies, which made it more vulnerable. Strachan-Davidson concludes that the Roman law courts in the period 300–500 A.D. “share in the general demoralization of the age” and participated in the “machinery of cruelty and rapacity” of the later empire.11 The evolution of the law of confessions is consistent with the hypothesis that Roman judges perceived a threat to their way of life that led to the greater use of torture to obtain confessions.
Better proof is available for why European justice turned to torture in the thirteenth century. In 1215, at Pope Innocent III’s insistence, the Fourth Lateran Council barred the participation of clerics in the ordeals of hot iron and water.12 In that era, the ordeals were the principal means of settling disputes when guilt was not made clear through confession or a defendant caught red-handed. Ordeals were explicit invitations for God to reveal which party was telling the truth. In the ordeal of hot iron, for example, if the burned hand was healing after three days, that litigant was telling the truth. Granting the premise that God would reveal the truth-teller, the ordeal was infallible. Clerics were critical to its implementation because they wrapped the burned hand, unwrapped it three days later, and proclaimed whether the burn was healing.
When clerics were forbidden this role, the proof dimension of the ordeal was suddenly lost. Both English and European justice systems cast about for a mechanism that was as infallible as possible. (We defer the English story to section B of this chapter, noting here only that English law never embraced torture to obtain confessions in routine cases.) In a search for a system that would not convict the innocent, thirteenth-century European law required that no conviction could be entered unless there were two eyewitnesses or a voluntary confession.13 The net effect of this rule of evidence was to make it impossible to convict on circumstantial evidence, no matter the strength of the evidence, unless the defendant confessed. The inability to convict in many cases was a threat to order in Europe. The twin controls—fear of false confessions and value of autonomy—that would normally prevent torture to obtain confessions were not weighty enough at this time in history. As John Langbein puts it, “No society will long tolerate a legal system in which there is no prospect of convicting unrepentant persons who commit clandestine crimes.”14
The law in Europe thus began to accept torture as a means of obtaining confessions in cases where the circumstantial evidence suggested guilt. Again according to Langbein, “The law of torture grew up to regulate this [European] process of generating confessions.”15 There is no reason to assume that Europeans were simply more tolerant of torture than their English counterparts—indeed, the kings and privy councilors who designed and administered the English justice system were largely descended from the Normans. Thus, the more permissive attitude toward torture on the Continent probably resulted from the greater perceived threat to peace and order created by the rigid European two-witness rule.
England avoided resort to the two-witness rule because of reforms implemented a half-century before the Fourth Lateran Council, in Henry II’s 1166 Assize of Clarendon, to which we now turn.
Prior to the Assize of Clarendon, robbery, theft, and murder were resolved by a variety of proof mechanisms. A person caught “red-handed”—i.e. with proof of the crime in his possession—was “brought before some court ([likely] enough it is a court hurriedly summoned for the purpose), and without being allowed to say one word in self-defense, he will be promptly hanged, beheaded, or precipitated from a cliff, and the owner of the stolen goods will perhaps act as an amateur executioner.”16 In the absence of clear evidence of guilt, cases were decided by confession, testimony, the swearing of oaths, and examination of documents. The early English turned to the ordeal “when other ways of discovering the truth were not available.”17
The ordeals that migrated to England, perhaps as early as the late seventh century and certainly by the tenth century, were developed by the Frankish tribes in Europe.18 Even if one accepted the premise of the ordeals, it was a deeply unsatisfying method of settling disputes. It was, at least sometimes, corrupt by the reign of England’s second king, William II, son of William the Conqueror. He ridiculed the ordeal of hot iron when fifty men accused of taking the king’s deer were, in 1098, made to carry the hot iron and then presented themselves three days later with unburnt hands.19 Perhaps the iron was not so hot.
Another problem with the ordeal was that a specific accusation was “normally decided by battle.”20 Trial by battle, as the name implies, sometimes meant a dawn-to-dusk battle where one party died horribly. Bartlett quotes an account of a trial by battle in which Guy “unhorsed” Herman who then disemboweled Guy’s horse.21 They fought with swords “until both, exhausted by the weight and burden of arms, threw away their shields and hastened to gain victory” by wrestling. Guy seemed to have prevailed but Herman put his hand under Guy’s armor, “seized him by the testicles” and “hurled Guy away from him; by this tearing motion all the lower parts of the body were broken so that Guy, now prostrate, gave up, crying out that he was conquered and dying.” Needless to say, the prospect of facing trial by battle must have discouraged private parties from making accusations.
Crime was rampant in England during the twelfth century, particularly during the contested reign of Stephen.22 “The collapse of the machinery of government in Stephen’s reign had been almost complete.”23 Powerful lords walled themselves off in castles, “levied payments from the villages” and when all the money was gone, they “plundered and burned all the villages.”24 The tyranny of the castle owners led to an English world, as described by contemporary English monks, where “[i]f two or three men came riding into a village, all the villagers would flee, for they thought they were robbers.”25
William the Conqueror’s great-grandson was crowned King Henry II in 1154.26 He quickly asserted a kingly authority that Stephen lacked. He recovered all the territories and rights of lordship that had been lost to Scotland, France and Wales during Stephen’s reign.27 To restore order, he destroyed the castles of the barons who opposed him and claimed royal ownership of all lands that had ever been owned by kings of England.28 Later, he reshaped the administration of justice. To the end of “preserving peace and maintaining justice,” the Assize of Clarendon, 1166, consolidated power in the royal courts and gave sheriffs more power to arrest those who served the castle lords.29
But by far the most significant change was that justice did not have to await the victim or the victim’s family to take action. If a jury of presentment, similar to a modern grand jury, found reason to suspect robbery, theft, or murder, then the defendant was given a chance to prove his innocence by the ordeal of water.30 The jury of presentment, representing the community, displaced private justice at the hands of victims for the crimes of robbery, theft, and murder. The conceit created by the Assize of Clarendon, which continues to this day, is that crimes harm the State. Today, criminal defendants are prosecuted by a state, by the United States, or by the people of a state, as in The People v. O. J. Simpson.
When the ordeals disappeared in 1217, English judges soon took the verdict of the inquest as guilt rather than probable guilt. It was a natural progression. Because juries had been making judgments about probable guilt for a half-century, they could be trusted to determine the ultimate truth about guilt. English law thus never adopted the European rule requiring two eyewitnesses or a voluntary confession.31 English juries could convict on strong circumstantial evidence, and no reason existed to resort to torture to obtain confessions.
Indeed, the Assize of Clarendon implicitly rejects untrustworthy confessions. Chapter 13 provides that “if any one, in the presence of lawful men or the hundreds, make confession of robbery, murder, theft, or the reception [hiding] of those committing them, and should later wish to deny it, let him not have law.”32 To translate a bit, “let him not have law” denied the confessor the ordeal as a means of vindicating innocence. “[T]he hundreds” is possibly a reference to the hundred man, the head of the district responsible for organizing villagers to catch thieves and robbers.33 The requirement that the confession be heard by “lawful men” excludes men whose oath was not accepted because they had been convicted of a crime or had otherwise “diminished their lawfulness.”34 In sum, confessions proved by trustworthy men could be accepted as true, and no need existed for the proof of the ordeal. Untrustworthy confessions were not accepted as proof.
The Assize of Clarendon is now almost 850 years old, but it is an uncannily modern treatment of confessions. Both then and now if the confession is trusted to be true, it disposes of the case. Under Henry’s law, the case was literally over because the defendant was denied a chance to prove his innocence. The confession amounted to a conviction. Today, the effect of a confession is less formal but dramatic nonetheless. If a judge admits a confession, juries rarely acquit. Defendants have the constitutional right to explain to the jury why the confession is untrue.35 But the odds of an acquittal are so low that few defendants proceed to trial once the judge rules a confession admissible.
In Henry II’s day, the effect of a confession turned on the trust that the legal system could put in the one who heard the confession. While this rule will be amended in various ways over the next 850 years, it remains the core of the law of confessions today. We see this idea in the earliest English confession case we found, discussed in the next section. The process is also much the same today. A grand jury decides whether there is probable cause to believe the defendant guilty and then, as a practical matter, the burden shifts to the defendant to disprove the State’s case.36
We begin with a distinction that will be important for the rest of this chapter and the next. Two very different categories of confessions existed in English law. The most common form in the early period was an admission to judges in court that constituted a conviction. This is the descendant of a confession to “trustworthy men” in the Assize of Clarendon. It was the end of the case. The second type is “mere confessions” that had no formal legal effect. These included statements to the justices of the peace, to coroners during inquests, and to victims and bystanders. The “trustworthy men” category from the Assize of Clarendon had, in effect, been narrowed to judges.
We see examples of both types of confessions in the Year Books, a massive collection of the law reports of medieval England. Our source for the Year Books cases is the website created by Professor David Seipp, who has indexed and translated over 22,000 of these reports that “are our principal source materials for the development of legal doctrines, concepts, and methods from 1290 to 1535, a period during which the common law developed into recognizable form.”37
The earliest English confession case we found was in 1295, in the twenty-third year of the rule of Edward I. A tall “immensely powerful” man who was a “born soldier,” Edward led English armies against Wales and eventually, incorporated Wales into England in 1284.38 He then turned his attention to Scotland, seeking a diplomatic and political takeover. When that failed, he was the first English king to attempt to conquer Scotland.
On the judicial front, the measures taken by Edward’s great-grandfather, Henry II, in the Assize of Clarendon had established a stable legal order. Henry III, Edward’s father, had also done much to regularize the legal order. While the local shire courts continued to function, the royal courts heard all cases where the jury of presentment found reason to suspect defendants.39 The threat from crime, and thus from that form of deviance, was far lower in 1295 than it had been in 1166 when Henry II established the Assize of Clarendon. Thus, our thesis predicts that courts would be more willing in 1295 to take seriously a claim that a confession was given unwillingly or was untrue.
We see this in Robert Butler’s case from 1295. While in prison awaiting trial, he confessed to a coroner that he was “a thief, murderer and robber, etc.”40 Coroners functioned as investigating magistrates, gathering evidence in homicide cases and reporting that evidence at the defendant’s trial. In Butler’s case, the coroner provided the written confession to the royal justices, who called Butler to appear before them to see if he would acknowledge his confession in court, thus in effect pleading guilty. He did, but he claimed that he made the confession to the coroner “because of the rigour and harshness he had sustained in prison, in order to be relieved from his distress.”
Notice the procedural posture here. The confession to the coroner was not itself a conviction. To convict him in the king’s court, the justices required his appearance before them “to see if he was willing to confess what he had admitted before the coroner.”41 If he had denied the confession, then a jury would determine his guilt or innocence. When he admitted the confession, his first line of defense was to claim that he confessed because of prison conditions.
While little is known about prison conditions in England in the thirteenth century, it is safe to assume that prison was a horrid place. The two major London prisons of the time were Fleet and Newgate, both built in the twelfth century.42 Newgate was built later than Fleet, and we know more about it.
Very soon after it had been opened Newgate Gaol had acquired an especially terrible reputation for its unwholesome environment, the distresses of its prisoners and the cruelties of its staff. … The smells of Newgate were renowned and for centuries they polluted the whole of the surrounding locality. Lack of hygiene and faulty sanitation caused frequent outbreaks of fever and the annual death-rate was extremely high.43
The case report from the Year Books does not explain why Butler thought that the prison conditions should relieve him of the legal effect of his confession. It is possible, we suppose, that he was making a dignity or autonomy claim. Six centuries later, an American defendant named Ernest Miranda would make a similar dignity-autonomy claim—that the rigor and harshness of police interrogation should relieve him of the consequences of confessing. But we are doubtful that, in 1295, autonomy or dignity claims would cause a court to disregard a confession made to a coroner. Indeed, we are skeptical that an English subject, in 1295, would be aware of a right that could require the king to provide decent prison conditions. Thus, the more likely underlying claim is that the rigor and harshness of the prison conditions compelled Butler to confess falsely. A false confession, as the Assize of Clarendon recognized, is of no value to a legal system.
To decide Butler’s claim, the justices sent for three prisoners and asked them if the conditions were as Butler described. According to the report: “They say no.”44 Not satisfied, the justices then sent for the “keeper of the prison” and made him swear to tell the truth. Unlike today, in medieval times oaths were taken very seriously. It is fair even to say that an oath created a sacred obligation to comply with the oath or face God’s damnation. The jailer testified that Butler’s account of the prison conditions “was never true.” The justices then asked Butler for a response. He repeated his claim that he confessed because of the “harshness of the prison,” but he quickly retreated to a second argument—that he was a clerk and thus had a clerical privilege not to be tried in the royal courts. He lost on that claim, too, for technical reasons that need not concern us.
The significance of Butler’s case is the court’s response to the argument that the confession should not “count” as a guilty plea because it was induced by prison conditions. One possible response was to deny the legal premise of the claim, to say it did not matter whether you confessed to escape the rigorous conditions. This the court did not do. Apparently accepting the legal premise that a false confession should not be deemed a conviction, the justices went to the trouble to call three fellow prisoners as well as the keeper of the prison before they ruled that the factual premise was false. It appears that, by 1295, English justices accepted the proposition that harsh treatment might compel a defendant to confess falsely. Under the Assize of Clarendon, a confession would not operate as a guilty plea unless it was made to a trustworthy witness. In 1295, English courts continued to recognize that untrue confessions should not be treated as guilty pleas.
To be sure, the crimes Butler faced were capital offenses, and the proportionality principle suggests that judges should tread lightly and not be too eager to accept a confession. But in a way that misses the point. It was the judges’ willingness in 1295 to take seriously Butler’s argument that a false confession should not be credited that makes his case significant.
A different dimension of the evolving law of confessions can be seen in a case that took place in 1348, the year the Black Death reached its deadly peak in England.45 The criminal procedure of the time required defendants who had been found probably guilty by the jury of presentment to relate their version of the case to the royal justices. In doing so, defendants would sometimes state facts that admitted guilt. In those cases, the justices treated the defendant’s statements as a guilty plea and entered a conviction. The only time the question was submitted to the trial jury was when the defendant’s presentation did not admit guilt.
On arraignment for the death of W.C., the defendant, Tailour of S., told the justices that he would “tell the Court the case as it was.”46 He wanted to strike someone else with his knife but “W.C. came between them to settle (appeaser) the fight, and the defendant struck him with his knife, not willing or knowing, so by misadventure [accident].” But this particular claim of accident was, under English law, an admission of the felony of homicide. Justice William de Thorp, Chief Justice of the King’s Bench, said that the defendant’s explanation was
cause enough to adjudge him to death, because it was the law that if two persons fought together, and one came between them to settle (appeaser) the contest, and one or the other struck and killed him, he was guilty of this felony, because he began the wrong (le mal), and did it of his own wrong, and not by misadventure.
To his horror, Tailour’s defense had collapsed into a guilty plea to a capital crime! But de Thorp said that if the defendant wished to put himself “on the jury” (have a jury decide guilt), the court would ignore his guilty plea and “receive him of grace.”47 The latter likely means that the court was free to accept the guilty plea and sentence him to death but chose, in its discretion, to let the matter be settled by the jury.48
We see here a new wrinkle in the English law of confessions. It was not enough that the confession was the truth, as was the issue in Butler’s case and in chapter 13 of the Assize of Clarendon. Nothing suggests that Tailour’s statement was false. Yet the court permitted him, in effect, to withdraw his confession because it had been made without knowledge that he was pleading guilty. Tailour’s case is evidence that English justices weighed whether admissions were made with knowledge of their consequences before entering a conviction. This, of course, is recognition of the defendant’s autonomy in a context where the proportionality principle suggests caution in finding the defendant guilty.
Getting to the truth might be the first value served in the law of confessions. But getting to the truth in a way that recognizes the suspect as a free-willed autonomous actor with dignity interests is another value served by the English law of confessions. Perhaps the horror of the Black Death contributed to the creation of autonomous English subjects. It had been “widely known” since 1346 “that a plague of unparalleled fury was raging in the East” and making its way westward.49 Europe and England succumbed in 1347 and 1348.50 A natural reaction to watching an agonizing death fall remorselessly on a quarter of the population would be to take the side of a defendant who had unwittingly condemned himself to the gallows.
To understand the next case requires that we explain the doctrine of “approvement,” which gave judges discretion to permit a defendant who had confessed his crime to accuse or “approve” others of the crime.51 If the approved defendants were convicted, the original defendant was banished but would escape the death penalty. We know that approvement was recognized in English law as early as the mid-thirteenth century because Bracton mentioned it in his treatise.52
An unnamed defendant in 1411 attempted to approve others to obtain release from jail in a homicide case.53 Because this was a “gaol release” case and not a trial, the judge who heard the case, Stourton, was not one of the king’s justices. Indeed, the court clerk noted that he was not even a serjeant, the highest rank of lawyers. Though the record is far from clear, it appears that the defendant asked Stourton if he could confess his felony in court and then approve others of the crime. Stourton told him that he should have confessed in prison rather than in court. Once he confessed in court, he could not approve others because a confession in court would result in his being “hanged forthwith.” The judge then told the defendant that he “ought not to confess the felony in hope of saving his life.” The outcome of the case is not given, but no grounds appear to permit his release from jail. He likely was brought to trial later in the king’s court.
The case is best read, we think, as evincing Stourton’s desire to protect the defendant from confessing his guilt while ignorant of the consequences. To be sure, Stourton’s advice that a confession in court would preclude becoming an approver seems to be incorrect.54 Perhaps this is why the clerk recorded that Stourton was neither a serjeant nor a justice. But if Stourton was sincere, it is another example of a judge seeking to prevent an admission of guilt from someone who did not know the consequences of confessing. It is an example of the evolving autonomy of the English subject.
But the autonomy sword had two very sharp edges in that era. The edge that could hurt defendants was that a guilty plea made without coercion and with knowledge that one was pleading guilty was conclusive proof of guilt. A defendant named Gower had earlier confessed in court that he had “consented to and abetted the death of A.”55 He was convicted and permitted to abjure the realm. In 1345, he was found in the realm, taken before the King’s Bench, and asked “why they should not proceed to execution” on the earlier conviction for his failure to stay out of the realm.
Gower’s defense was that the earlier conviction was unfounded because he only aided in the death and could not be convicted until the one who committed the killing, the principal, was convicted.56 As the principal had not been convicted, Gower argued that his earlier conviction was void. While his defense was valid in law, the court ruled that it applied only to defendants who put their case to the jury. It did not protect defendants who confessed. Indeed, the justices made the remarkable statement that even if there had been no murder, “still his confession condemned him.”
Executing innocent men who have chosen to confess to a capital crime seems to take autonomy too far. Yet in an eerie parallel, some judges and prosecutors today have argued that a wrongful conviction should stand if the time for appealing or challenging it has passed. Justice Scalia remarked: “There is no basis in text, tradition, or even in contemporary practice (if that were enough) for finding in the Constitution a right to demand judicial consideration of newly discovered evidence of innocence brought forward after conviction.”57 Scalia seems to say that once the legal system has spoken and the time to review the outcome has passed, it becomes a permanent legal fact in the universe, at least as far as the United States Constitution is concerned. That approach to legal finality could justify the court’s view in Gower’s case that a false confession, willingly made, led to a valid conviction that could not be challenged.
The autonomy of English defendants in that era did not extend to the right not to plead to the charge. Defendants who did not plead their cases could be found mute by malice and hanged. Incentives thus abounded for defendants to talk before the justices and the jury. In a 1406 case, the inquest found the defendant mute by malice and “[a]warded that the felon be hanged,” but the justices said it was not necessary to enter judgment on that charge because the defendant had acknowledged the felony in an earlier proceeding.58
An alternative to being hanged for refusing to plead was the quaint peine forte et dure, or “hard and forceful punishment.”59 According to Matthew Hale, a defendant subjected to peine forte et dure was “put into a dark, lower room” in the prison and “laid naked upon the bare ground upon his back … his legs and arms drawn and extended with cords to the four corners of the room, and upon his body laid as great a weight of iron, as he can bear, and more.” He was given three morsels of bread without drink and afterwards fed nothing and allowed only to drink standing water “next [to] the door of the prison” until he died. It is an odd view of autonomy, to modern eyes, to recognize the right not to be tried until one has entered a plea but then allow the crown to kill the defendant who refuses to plead.
Other sources indicate that if the defendant relented and agreed to plead and to be tried by a jury, the weight would be removed.60 This description of peine forte et dure is consistent with cases we found where the justices pleaded with defendants to enter a plea, explaining what would happen if they did not. In one case from 1492, Justice Fairfax told the defendant to “take good care of yourself, my friend, and think carefully what a great offence it is against God to murder [yourself] voluntarily through obstinacy. … In any case, twelve men will perhaps find you not guilty.”61 The defendant pleaded not guilty. Again, we see concern with autonomy.
One way for a defendant to challenge a confession would be to show that it was produced by coercion. We saw a foreshadowing of that argument in Butler’s claim that the prison conditions caused his confession. The argument appears, more fully formed, in an early civil case about twenty years after Butler’s case. In 1313, a defendant sought to avoid having to pay a bond by claiming “that he executed the bond under coercion of imprisonment.”62
In 1353, a defendant used the coercion defense successfully in a criminal case, albeit not in a confessions context. Arraigned for stealing two shillings worth of bread, the defendant said that she did it “by the command of the one who was her husband at that time.”63 The judges refused to rule in her favor and allowed the case to go to the jury, which “found that she had done it by the coercion of her husband against her will … and thus she went acquitted.” The theory here, expressed in modern form, is that the will of the coercer substitutes for the one who is coerced, rendering her guiltless. Implicit in the sketchy case report, of course, is the notion that in the fourteenth century, wives did their husband’s bidding without question.
Most of the early confessions cases featured admissions that were both freely given and knowing. About one-third of the confessions cases from this century involved a defendant who confessed freely and then claimed he was a cleric.64 If he proved he was a cleric, by reading from the Bible, the court would deliver the defendant to the ordinary—the bishop acting in his ecclesiastical capacity—for punishment.65 About another third of the cases involved a freely given confession and then a request to become an approver of others.66 The majority of the final third involved a confession that was merely mentioned in passing. For example, in 1492, Archer confessed to the coroner that he killed a man.67 The issue at trial was whether he would plead to the charge or stand mute.
The argument that a coerced confession was questionable as evidence in a criminal case does not appear in anything like the modern form until we arrive at a period that John Langbein called the “century of torture”: 1540–1640.68 As we will see, even under rulers like Henry VIII, “Bloody Mary,” and Elizabeth, torture was sufficiently rare that it was an outlier rather than the norm. Indeed, the challenge for Langbein was to explain why if the English used torture occasionally to obtain evidence, its use did not become routine as it had in European justice.
Following Langbein, our interest in torture is only when used to obtain confessions.69 Violence was endemic in English society of this time, but its use was almost always to dispatch one’s enemies, to force a defendant to submit to a jury (peine forte et dure), or to punish defendants convicted of particularly dangerous or loathsome acts. For example, the Year Books contain an account of the Earl of Kent’s attempt to overthrow Henry IV in 1399.70 Though the account has, in modern times, been dismissed as a forgery,71 it would not have been included in the Year Books if it lacked verisimilitude. In the Year Book account, two of the defeated earls were beheaded on the spot. A third escaped but was later captured and beheaded in front of a large crowd. Then “his head was taken to London and set up on London bridge.”
A relatively common form of execution was drawing and quartering. The point to this barbarous practice was to make the condemned man suffer as much as possible prior to death and to desecrate his body. Here, for example, is the sentence pronounced after Sir Walter Raleigh was convicted of treason:
[You are] to be hanged and cut down alive, and your body shall be opened, your heart and bowels plucked out, and your privy members cut off, and thrown into the fire before your very eyes; then your head to be stricken off from your body, and your body shall be divided into four quarters, to be disposed of at the King’s pleasure. And God have mercy upon your soul.72
None of these practices involved confessions. But we know, because they kept good records via the Privy Council, that the English did sometimes use torture to produce confessions. An advisory body to monarchs since the Norman kings, by the sixteenth century the Privy Council included judges, barons, king’s serjeants, and the king’s attorney.73 Torture to obtain evidence required written authorization by the Privy Council or the monarch, and many of those warrants survived.74 The warrants spanned the years 1540–1640. There is every reason to believe that the warrants that survived disclose most of the use of torture to obtain evidence. As Langbein puts it: “If there had been any massive use of torture before the records began to evidence it in 1540, or after they cease to disclose it in 1640, or omitted from the records for the intervening century, literary and other sources would alert us.”75
If the English monarchs were ever to use torture routinely, it would have been in the period from 1540 to 1640—a period that saw three monarchs who were more than willing to use terror and repression to rule England.76 Henry VIII’s often despotic rule is well known. His daughter, Mary I, reinstituted Catholicism as the state religion, which made her the target of the Protestants who had ruled England under her father.77 She was known as “Bloody Mary” because 300 Protestant “heretics” were burned at the stake during her reign. Elizabeth’s reign was calmer than that of her father and sister, but she obviously felt threatened—two-thirds of all surviving torture warrants were issued while she was queen.78 Indeed, upon Elizabeth’s death, torture warrants became very infrequent, sputtering to an end during the reign of Charles I in 1640. Two years later, the English civil war would begin, and yet only one torture warrant was issued in the years leading up to the war.
Thus, while the period from 1540 to 1640 was almost certainly not the totality of English use of torture to obtain confessions, it was almost certainly the period of its most frequent occurrence. If we find few torture warrants during this period, it must mean that the English rarely used torture to obtain evidence. And we do find few torture warrants. Langbein records eighty-one Privy Council torture warrants, or an average of fewer than one per year.79 Even if the records show only one-third of all torture warrants that would be an average of three uses per year of torture to obtain confessions. Langbein concludes: “Relative to the thousands of felony investigations each year, the number of torture cases was minuscule.”80 Blackstone draws an instructive distinction when he concludes that the rack “was occasionally used as engine of state, not of law, more than once in the reign of queen Elizabeth.”81 Torture, then, was not part of the English system of proof.
If torture was not part of the standard proof in England, as it was on the Continent, why was it used at all? Langbein’s study of the torture warrants persuades him that it was used principally to protect the state from treason and counterfeiting and, occasionally, when a powerful figure had been the victim of crime.82 The first two connections fit our thesis that when a State feels threatened, and there are no threats greater than treason or counterfeiting, it will respond with whatever is necessary to ensure its survival. The use of torture when a powerful figure had been the victim of a crime is simply an artifact of political power.
To say that the English did not view torture as a routine method of extracting confessions is not, of course, to say that the English did not use coercive interrogation methods. As we will see in a moment, compulsion, force, and violence were standard interrogation methods, at least by 1547 in treason cases. We are back to the question asked at the end of the last section: Why would a legal culture that permitted convicted defendants to be drawn and quartered not routinely permit torture to obtain evidence, as was done on the Continent? Langbein’s answer is that the English legal system obtained a sufficient number of convictions in jury trials at which much evidence was admissible, while on the Continent, the two-witness rule made convictions difficult to obtain.83 Thus, by 1628, torture to obtain evidence in England “was falling into desuetude,” an “English experiment” that “left no traces.”84
Langbein’s account seems right, as far as it goes, but still leaves the bigger question unanswered. If the English viewed torture as simply another tool to uncover truth, it would have been used routinely. There must have been a suspicion that torture did not produce truth, a belief that can be traced back to Roman law. A manifestation of the reluctance to rely on tortured confessions appeared early in the “torture century.” In 1547, Parliament sought to cleanse treason prosecutions of torture when it forbade an indictment or conviction of treason unless the defendant “be accused by two sufficient and lawful witnesses, or shall willingly without violence confess the same.”85
The 1547 Treason Act can be explained, of course, as a limit on the power of the monarch to bring down his enemies. That Parliament enacted it only months after the death of Henry VIII manifests this motive. It seems likely, though, that Parliament was also acknowledging the unreliability of confessions produced by violence. Staundford’s treatise, published in 1557, noted that when a defendant is charged with a felony, confessions to the judge at arraignment are “the best and surest answer that be in our law for quieting the conscience of the judge and for making a good and firm condemnation” as long as the confession “did not proceed from fear, menace, or duress.”86 This concern seems pretty clearly directed at the unreliability problem. Thus, Staundford’s principle and the motivation behind the 1547 Treason Act probably come from a common source. Unfortunately, that source is lost in what we call the “missing century.” But first a brief detour to the Star Chamber.
When interrogation methods are the issue, the Star Chamber is perhaps the most well-known English institution. Leonard Levy wrote about it extensively in Origins of the Fifth Amendment, which won a Pulitzer Prize in history.87 Levy’s thesis is that the reaction to the Star Chamber helped develop the English principle against self-accusation.88 In 1638, John Lilburn refused to take an oath to tell the truth to the Star Chamber, which sought to question him about his unlicensed publishing activities. As we will see in chapter 7, the Court in its famous Miranda case relied heavily on Lilburn’s persecution by the Star Chamber to expand the privilege to cover custodial police interrogation.
Yet, despite the fascination of this mostly secret court, it is largely beside the point for our enterprise, with one exception that we note in the next section. First, this history has been thoroughly, perhaps obsessively, documented. We have little to add, though we would remind readers that John Lilburn’s fight against the Star Chamber was not about its substantive power to compel confessions but, rather, the procedural rule that those who appeared before the Star Chamber could be compelled to swear an oath to tell the truth. Lilburn, and many others, argued that the crown lacked the power to compel anyone to swear an oath to tell the truth. As noted earlier, oaths at the time created a solemn obligation to tell the truth. Thus, if someone could be compelled to take an oath to tell the truth, then he was forced to be his own accuser. But it was self-accusation that critics of the Star Chamber condemned, not involuntary confessions. Thus, the fight over the Star Chamber has nothing to say substantively about whether a confession is admissible.
Second, its mostly secret nature means that historians lack the kind of records that are available in the Year Books and the records of the Old Bailey. Third, as an arm of the crown that did not have to follow the common law,89 its machinations tell us nothing about the common law of confessions. As Bentham put it, “In the common law courts, these [Star Chamber] enormities could not be committed.”90 Indeed, defendants were not permitted to testify under oath in common law courts, and no issue of refusing to take an oath could ever arise in these courts.
Finally, whatever value the Star Chamber adds to the English confession story, that court was abolished by the Long Parliament in 1641.91 Most of the movement toward the modern law of confessions occurred after 1641. This is the story to which we now turn.
The “missing century” is actually a century and one-third of another (but that does not make a very good title). The Year Books ceased in 1535, and the Old Bailey trial records do not begin until 1674. There are, to be sure, individual case reports in various volumes published by the Selden Society for the period from 1535 to 1674, but these are scattered and difficult to search. We found several statements of concern about how confessions were obtained in this period, two of which we have already noted. The earlier is the Treason Act of 1547 that required two witnesses unless the defendant “shall willingly without violence confess” the treason. The second, ten years later, is Staundford’s treatise, which broadened and deepened the principle. Staundford noted the value of confessions of felonies to the judge, as long as they “did not proceed from fear, menace, or duress.” Compared to the Treason Act, Staundford’s principle is broader because it applies to all felonies and not just to treason. It is deeper because it cannot be fully explained as an attempt to limit the power of the king.
Another early mention of concern about confessions appears in the November 20, 1606 Articles, Instructions, and Orders contained in the 1606 Charter from James I that established the two Virginia colonies. In setting out the right to a jury trial by twelve “honest and indifferent persons,” the document continues: “and that all and ever person or persons, which shall voluntarily confesse any of the said offenses” shall be convicted as if found guilty by a jury.92 This is the earliest use of the “voluntary” terminology that we found in English law. The same locution shows up in a 1662 Virginia statute that criminalized certain religious observations.93
We found another reference to voluntary confessions in a treatise about the Star Chamber. William Hudson wrote, in 1635 or earlier, that when “dangerous persons” acted in ways that “may be very likely to endanger the very fabric of government, these persons [could be] apprehended by a pursuivant or messenger, and privately examined [before the Star Chamber], without oath, or any compulsory means, concerning the fact.”94 If the person denied the accusation, then the Star Chamber had to proceed by written evidence. “But if he confesses the offence freely and voluntarily, without constraint, then may he be brought to the bar” and examined orally. Like the Treason Act of 1547, the Star Chamber procedures recognized an exception to the formal means of proof—two witnesses testifying to treason or the requirement of proceeding by written evidence only—when the accused confessed the act “willingly without violence” or “freely and voluntarily.”
Coke’s Third Institutes of the Law of England, published in 1644, also contained a reference to voluntary confession. After a felon took sanctuary in a church, his “voluntary and particular confession of the felony before the coroner” justified a judgment that the felon abjure the realm.95 Here, of course, “confession” is used as a plea of guilty. The reference to voluntary confessions in sanctuary cases is oddly narrow. Coke speaks more than once in his Third Institutes of “confession” as a plea,96 but the only other reference we found to a “voluntary” confession was when Coke was paraphrasing the Treason Act of 1547.97 Perhaps the hallowed place of sanctuary in English law, drawing on God’s protection, required more of confessions given while in sanctuary. More likely is that, by 1644, Coke just assumed that pleas to judges had to be voluntary. That was, after all, Staundford’s view almost a century earlier.
To be sure, Coke was the king’s prosecutor in the Gunpowder Plot cases as well as in Sir Walter Raleigh’s prosecution, to which we shall turn shortly, and both featured torture to obtain a confession. The Gunpowder Plot was a plan to blow up the House of Lords.98 Guy Fawkes was found guarding thirty-six barrels of gunpowder on November 5, 1605. He admitted that he had “been at the bottom” of the plot and “that there were fellow-conspirators whom he refused to name.” In seeking more details about the conspiracy, King James I ordered by letter that, if necessary, Fawkes be tortured into confessing: “‘If he will not otherwise confess’ … ‘the gentler tortures are to be first used unto him, et sic per gradus ad ima tenditur’ [and thus by steps extended to greater ones].”99
Coke was one of the commissioners to whom the king’s letter was directed. It is doubtful that he personally inflicted the torture that Fawkes suffered, but he certainly knew it had been authorized. But what should we make of Coke’s involvement in an examination that included torture? First, the Gunpowder Plot trials occurred almost forty years prior to the publication of his Third Institutes. Moreover, Coke did not authorize torture of Fawkes and, as far as we know, might have thought it wrong. Finally, the thesis of this book holds that a government or a prosecutor could approve torture in a case where the State’s existence is threatened, as in the Gunpowder Plot, and still believe that a confession in an ordinary criminal case had to be voluntary before a judicial officer could accept it. As for the threat to the crown posed by the Gunpowder Plot, Gardiner’s account, published in 1897, offered a partial excuse for the torture given the very real possibility of a general Catholic uprising. “We cannot deny that, at that particular moment, they had real cause for alarm.”100
Notice, however, that Staundford’s principle is limited to confessions to the judge, Hudson’s rule to confessions before the Star Chamber, and Coke’s rule to confessions to a coroner. None would apply to “mere confessions” made outside a formal setting. Whether the Treason Act of 1547 prohibited the use of out-of-court confessions turned out to be controversial. Because confessions were used in lieu of two witnesses who otherwise would testify in court, the Act appears to make witnesses unnecessary only when a defendant confessed in court. But political realities being what they are, within a few years of Henry’s VIII’s death, English courts felt pressure to relax the treason rule so that confessions made prior to trial could be used instead of two witnesses. Coke was a particularly adept proponent of this interpretation, as we shall shortly see. Parliament would later, in 1695, make clear that the confession to treason, given “willingly without violence,” had to be in open court to avoid the two-witness rule.101
Thus, the best reading of Staundford, Hudson, Coke, the 1606 Virginia Charter, and the Treason Act of 1547 is that these expressions of concern with confessions produced by violence were not limited to confessions made in formal proceedings. The “mere confessions” that had no formal legal effect were beginning to be viewed skeptically during this period. Larry Herman uncovered three cases from this period evincing concern about the voluntariness of confessions made outside of the trial itself.102 In the 1551 treason prosecution of the Duke of Somerset, the prosecutor asked witnesses about out-of court confessions that the crown assured the jury “were true” because “what was sworn was without any kind of compulsion, force, or envy, or displeasure.”103 When the Bishop of Ross accused the Duke of Norfolk of treason in 1571, he was examined in the Tower of London; the crown assured the jury that the bishop was “examined freely, and without any compulsion.”104 In 1600, Sir Christopher Blunt and four other defendants were accused of raising an army to demand concessions from Queen Elizabeth.105 Three of the defendants had confessed to treason, and the queen’s counsel argued that additional proof was unnecessary because the confession “came voluntarily from every man examined, no man being racked or tormented.”106
Three years later, the crown used a confession from the alleged co-conspirator of Sir Walter Raleigh made to the Privy Council to prove Raleigh guilty of treason. Coke was the crown’s lead prosecutor. Raleigh demanded, unsuccessfully, that his accuser be produced in court to repeat the confession. Raleigh claimed that his accuser had been held “close prisoner” for eighteen weeks and “was offered the rack to make him confess.”107 The judges denied that the threat occurred, thus signaling that the argument would have had force if it had been true.
Raleigh’s claim is now understood as the right to confront one’s accusers, a right enshrined in our Sixth Amendment. But at the heart of the claim is the notion that an involuntary confession is not good evidence because it is untrustworthy. The reason he wanted to confront his accuser in court was to obtain a noncoerced and thus more trustworthy version of events. By the time we arrive at Hawkins’s Pleas of the Crown, a little more than a century later, the common law had evolved a crude method to deal with out-of-court confessions of accomplices. They were simply inadmissible. Indeed, the only rule that Hawkins states in connection with out-of-court confessions is that they were admissible only against the maker.108 This solved the problem of the unavailable accomplice confessor because his confession was not admissible in the trial of the principal.
If Raleigh had been able to persuade the judges that the confession was not made voluntarily or willingly, the legal effect during most of this early period was not the modern remedy of suppression but, rather, was for the judge to instruct the jury that it should consider that the confession had been produced by coercion or out of fear and thus might not be reliable. (We will cover this in chapter 5.) If that was the appropriate remedy in 1603, it would not have done Raleigh much good even if the justices agreed that the confession of the alleged accomplice was coerced. Coke told the jury that the king’s safety required a conviction.109 Protecting the king from those who might be plotting his death was probably a higher-order value than giving defendants the right to prove their innocence.110 The jury deliberated a mere fifteen minutes before finding Raleigh guilty of high treason.111
Outside of treason cases, we found few out-of-court confessions during this period. The confessions that appeared in early English case law were mostly admissions made in court when defendants told their side of the events, claimed benefit of clergy, or sought to approve (accuse) others of the crime. Part of the explanation for the lack of out-of-court confessions in the run of the mill cases was that no law enforcement mechanism, in the modern sense, existed in this period. Police did not exist. Sheriffs and constables could make arrests in fresh pursuit but did not investigate “stale” crimes. Grand juries were not as aggressive as Henry II contemplated in the Assize of Clarendon, and it generally fell to the victim or the victim’s family to prosecute crimes.
Another reason for the lack of confessions in the modern sense is that juries of the period were “self-informing” juries: men chosen from the neighborhood precisely because neighbors would know the character of the parties and might even know something of the crime. Thus, the juries served as witnesses as well as triers of fact, and there was little need for other evidence. Langbein notes that “[a]s late as Fortescue (1460s), it was being boasted that the English jury merged witnesses and triers.”112
Although the notion of self-informing juries persisted beyond the fifteenth century,113 the transition to juries as impartial fact-finders was well underway by the turn of the sixteenth century. Impartial fact-finders require evidence, which would include out-of-court statements. If self-informing juries were in decline by the sixteenth century, we should see evidence of evidence-gathering taking place before trial.
And we do. By the terms of the Marian bail and committal statutes of 1554 and 1555, justices of the peace were required to transcribe in writing the statements of the accuser and other witnesses, under oath, and the accused, not under oath, before deciding whether to admit the accused to bail.114 “The JP’s job was to help the accuser build the prosecution case, rather than to serve as a neutral investigator seeking all the evidence, inculpating and exculpating.”115 The statements were used by the victim of the crime, often in consultation with a court clerk, to frame an indictment.116 The unsworn examination of the accused was “routinely read in evidence at his trial.”117
By 1555, the justice of the peace had become the focal point for gathering evidence of crimes. Confessions were now a routine part of the crown’s case. Yet the lack of organized case reports during this period makes it difficult to assess what kinds of arguments were offered when those statements were introduced at trial. We see at least the beginning of a claim that “mere confessions” must be voluntary to be fully credited in Raleigh’s trial and in the sixteenth century cases that Herman uncovered.118 Seventy years after Raleigh’s trial, similar claims would begin to appear in English trials held in the Old Bailey. The records of trials in the Old Bailey, available online, are a rich repository of evidence about English criminal trials. This is the story to which we turn in chapter 3.