AI-generated from sources
Refuted as Proof, Kept as Intelligence, the Long Life of Torture as an Engine of State
In Brief
- In the legal tradition of Europe and America, torture served two distinct ends: completing the proof of guilt with a confession, and drawing from a prisoner the names of his confederates in treason or conspiracy.
- Cesare Beccaria, whose 1764 essay on crimes and punishments attacked torture as a test of guilt, showed that pain selects for endurance rather than truth: it teaches the robust to stay silent and the weak to confess.
- Treason was where the old limits gave way: in Roman law, the rules that shielded masters and restrained torture all carried an exception for it, and on that charge anyone, whatever his rank, could be tortured.
- A colonial American law that banned torture to obtain confessions yet allowed it to uncover conspirators drew the same line as a 2005 American legal memorandum, which held that the condemnation of forced confessions in ordinary criminal cases said little about “enhanced” interrogation of suspects believed to hold vital intelligence.
Main authors
A colonial American law, reproduced in a source book of American history before 1787, begins with a refusal. No man is to be forced by torture to confess a crime, whether against himself or against anyone else [1]. Then comes the exception. In a capital case, once the prisoner has been fully convicted on clear and sufficient evidence, and if it is apparent that there were “other conspiratours, or confederates with him,” he may be tortured after all, provided the torments are not barbarous . In a few lines, the law performs a separation that would outlive it. Torture loses its role as proof, since guilt must be established before it begins. It keeps another role, which has nothing to do with the verdict: drawing from a condemned man the names of others.
That separation runs through the long history of torture in European and American law. When Cesare Beccaria attacked torture in 1764, he attacked it above all as a test of guilt, and on that ground his case proved hard to answer [2]. Torture's other use rested on something else. Uncovering confederates in treason and public danger never claimed to prove anything; it answered to the state's sense of its own peril [3, 4]. This is why the refutation of torture as proof left that second use largely standing. It also explains the form in which the practice resurfaces: rarely as a rule of ordinary criminal law, more often as an exception at its edges, defended in the name of necessity [5, 6].
Slaves, traitors and the Roman exception
Torture presupposes a public. In the early codes that Henry Charles Lea studied, a criminal could defend himself with the sword against his adversary's feud, or settle his guilt with money [7]. To torture him for a confession would have seemed an absurd violation of his rights. Crime was an injury between individuals, and the idea that society at large had a stake in discovering and punishing it was, in Lea's verdict on what he calls a barbarous age, too abstract to shape the law . Rome had that public, and it distributed torture according to status. Under the republic, by the account James Williams gave in the Encyclopædia Britannica, torture seems to have been confined to slaves, and Roman law held the torture of slaves to be the most efficacious means of obtaining truth [8, 9].
Treason was where the hierarchy dissolved. With the empire, free men accused of crimes became liable to torture; on a charge of treason, everyone was, whatever his rank, “for in treason the condition of all was equal” . The same exception ran through the rules that protected masters. Slaves could be tortured against their owners only in treason and in offences regarded as a species of it, such as coining or frauds on the revenue . Even the safeguards of procedure bent at this point. Torture was meant to come last, when nothing but the confession was missing, and a single unsupported witness did not suffice to order it—except in treason [10]. Each limit of the Roman system carried the same reservation for crimes against the state.
Torture as proof was unstable from the start, and it gravitated toward crimes held to be hidden. In his history of the Spanish Inquisition, Lea notes that no one could define evidence “insufficient for conviction yet sufficient for torture” [11]. One authority held that a single witness, even an accomplice, justified it; another required three accomplice witnesses . Witchcraft shows where this uncertainty led. Because it was thought peculiarly difficult to prove, and its testimony impalpable, torture became an unfailing resource for puzzled tribunals, while every safeguard was refused to the accused [12]. Secrecy, elusive proof, a danger that seemed to demand severity: the same features would later be claimed for conspiracy.
Under the Tudors, the case for torturing conspirators was already one of state rather than of proof. A justification preserved in Lucy Aikin's memoirs of Elizabeth's court reasons a fortiori. If the general laws of nations had allowed torture in lesser cases, and in a harsher manner, it could surely be used on persons whose conspiracies it so much mattered to have disclosed . The goal is disclosure, not a verdict. The historian A. F. Pollard gives this logic its frankest form. In Tudor times, he writes, expediency rather than justice became the supreme test of public acts; high treason was a political crime that need involve no breach of morals, and traitors were put to death because they were dangerous . Torture aimed at confederates belongs to that order of reasoning. It measures danger, and danger does not wait for proof.
Beccaria and the muscles of truth
Beccaria's refutation, published in 1764, is built as a dilemma aimed at the trial. If the accused is guilty, he should suffer only the punishment the law ordains, and his confession is unnecessary . If he is not, torture falls on an innocent man, for in the eye of the law every man is innocent whose crime has not been proved . Beyond the dilemma lies a deeper charge. To make a man his own accuser, and pain the test of truth, is to reason “as if truth resided in the muscles and fibres” of the sufferer . Every step of the argument concerns proof, presumption and verdict. Its target is torture as a witness.
Pain, Beccaria went on, does not extract truth; it sorts bodies. No interrogation, he observes, is “more suggestive than pain” [13]. It teaches the robust villain an obstinate silence, so as to trade a greater punishment for a lesser one, and teaches the feeble man to confess to end the present agony . The same line of reasoning, as set out in James Anson Farrer's Crimes and Punishments, draws the perverse conclusion: under torture the innocent man can only lose, while the guilty may gain [14]. Beccaria also gave the practice a genealogy. He called it a remaining monument of the ancient legislation of ordeals, the trials by fire and boiling water once named Judgments of God, with the sole difference that the outcome now depended on the will of the accused [15]. On this reading, torture was no method of inquiry at all, only an ordeal in disguise.
The reformers did not forget the accomplice. The text in Farrer's volume takes up torture used to discover accomplices and turns the main argument against it. If torture is unfit to discover the truth, how can it disclose accomplices, which are part of the truth to be discovered [16]? A man who accuses himself, it adds, will accuse others all the more readily; and is it just to torment men for the crimes of others ? An episode George Grote recounts in his history of Greece shows what that fear looks like. Suspected conspirators were tortured to reveal their accomplices; one of them deposed against 30 conspicuous citizens, whom the people put to death after a hasty trial [17]. Yet the objection remains framed as a question of truth. It treats names as evidence to be weighed, like a confession. It has less to say to an authority that wants those names not to convict anyone, but to act.
Refuting torture as proof also left its usefulness open to dispute. Albert Leffingwell, in a book on the ethics of scientific experimentation, held it beyond question that torturing accused men whose guilt was almost certain had, in innumerable cases, furthered the ends of justice, and that mysterious crimes had come to light through the mere threat of torment [18]. Where abolition came, it often rested on grounds other than the reliability of confessions. A papal bull of 1816, as Williams reported in 1888, decreed not only that torture should cease but that proceedings should be public and the accuser confronted with the accused [19]. Nearly two centuries after that bull, South Africa's Constitutional Court recalled that a British colonial administration of the Napoleonic era had abolished torture because it was regarded as barbaric in itself and degrading to the society that practised it [20]. Publicity and dignity are arguments that reach the closed room. The case against false confessions does not necessarily get there.
The sweat-box and the secret jail
Once disowned by the law, torture did not vanish; it moved beside the law. In a defence addressed to a jury, published as The Trial of Theodore Parker, Parker argued that in England torture had never been conformable to common or statute law, yet had often been practised by a corrupt administration and wicked judges . The same gap between rule and practice appears in American criminal justice as Winifred Louise Taylor described it. The accused, she wrote, could be subjected to the tortures of the sweat-box, so that he might convict himself before being tried, while the law professed to hold him innocent until proven guilty [21]. This qualifies any neat division of labour. The coerced confession that Beccaria had discredited survived as well, but as a practice the law refused to own.
The argument for stepping outside ordinary proof always came back to conspiracy. Writing in 1888 on the suspension of habeas corpus during the War of the Rebellion, Sydney G. Fisher set out the dilemma with care. The power to hurry a man secretly to jail, he recalled, had often been called “more dangerous to freedom than all the other engines of tyranny” [22]. Yet it was generally admitted, he added, that a government attacked by rebellion cannot protect itself from conspirators and assassins if each must be proved guilty beyond reasonable doubt before a court . Fisher is writing about detention, not torture. But his dilemma has the same shape as the one that shelters the torture of confederates: ordinary proof is cast as a luxury that conspiracy denies the state.
The answer to that dilemma was given in court, against the government's own plea. Speaking for defendants charged with conspiracy, Robert Green Ingersoll told a jury that conspiracy is proved exactly like larceny or murder, and rejected the suggestion that its secrecy should lower the evidence required [23]. Nearly all crimes are secret, he observed; very few men steal publicly, “with a band of music and with a torch in each hand” . The law also had a route to accomplices that needed no pain. Harper's New Monthly Magazine noted in 1852 that in almost all countries, when a crime could not be traced by ordinary proofs, an accomplice willing to inform against the rest was customarily exempted from punishment [24]. Promise, rather than pain, was the customary key to a conspiracy. That left the torture of confederates with a single justification: the claim that the danger is too great, or too near, to wait.
Actionable intelligence, an old exception renamed
In the early 21st century, American legal memoranda drew the old line again. A 2005 memorandum by Steven G. Bradbury on United States obligations under the Convention Against Torture argued that the American tradition of condemning torture, and the use of force to obtain confessions in ordinary criminal cases, said little about the CIA's interrogation practices . Those practices, it explained, were reserved for the relatively few terrorists believed to possess “vital, actionable intelligence” that might avert an attack . Written nearly two and a half centuries after Beccaria, the memorandum takes the condemnation of forced confession in ordinary justice for granted. What it claims is the other territory, the one a colonial American code had already set apart when it banned torture for confessions and kept it for confederates . Proof was surrendered; intelligence was kept.
The justification no longer speaks of guilt but of prevention. A 2003 Working Group Report on Detainee Interrogations invoked leading scholarly commentators for whom interrogation methods that might violate the law would be justified as self-defence: a combatant who had culpably caused the danger could be hurt if that was the only means of preventing death or injury to others, and such torture, in their words, should be permissible [25]. In a 2008 radio address, George W. Bush put the premise plainly. The best source of information about terrorist attacks, he said, is the terrorists themselves, and restricting the CIA to the methods of the Field Manual could cost vital information [26]. The prisoner is no longer an accused, to be found guilty or innocent. He is a source, valued for what he knows about others.
The exception presents itself through its restraint, and the restraint is old. Bradbury's memorandum stressed that the techniques would be used “as little as possible on as few detainees as possible,” on a determination made at CIA headquarters [27]. Roman law, as an encyclopedia summarized it in 1911, likewise hedged torture with conditions: nothing but the confession was to be lacking, a personal enemy's evidence was to be received with caution, and repetition was allowed only for inconsistent depositions or denial against strong evidence . Such rules of measure, Roman or modern, guarantee little once torture seeks networks rather than verdicts. An anonymous account published as Germany, Turkey, and Armenia tells of an old priest tortured to extract a confession until, hoping the pain would stop, he cried out that they were revolutionists [28]. The soldiers took the words from his own lips as proof, and instead of choosing their victims, the officials had all the Armenians tortured . One man's forced confession had become a warrant against a whole community.
The objections that carry weight against this use of torture are, accordingly, not about truth. Quoted in a 2007 presidential debate, General Vessey held that any information gained through torture could never be counterbalanced by the damage to America's reputation, and by the risk that enemies would cite it as an excuse to torture captured American soldiers [29]. The argument leaves open whether anything is learned; it weighs the cost to the state that tortures. Voltairine de Cleyre made a harsher calculation about the penalties once inflicted on traitors in England [30]. There were never so many traitors, she wrote, as when they were drawn and quartered; the frightful punishments bred hatred of the government until the traitors grew strong enough to decide what treason was . She was writing of punishment rather than interrogation, but the lesson bears on both: cruelty in the name of the state's safety can manufacture the very danger it claims to meet.
Torture had two lives in the law of Europe and America. One made it a means of proof, completing with a confession a case that evidence had left unfinished . The other made it a means of discovery, drawing from a prisoner the names of his confederates in treason or conspiracy . The critics of the 18th century destroyed the first: pain tests endurance, not truth, silencing the strong and making the weak confess . They touched the second only by treating an accomplice's names as one more piece of evidence . Yet that second use had never rested on proof. From the Roman rule that made all ranks equal before a charge of treason, to Tudor expediency, to memoranda on actionable intelligence, it rested on danger . And it lived less inside ordinary law than in its exceptions, its secret jails and its disowned practices .
What held against it were arguments of another kind. Abolitions grounded in the dignity of the society that tortures and in the publicity of proceedings, warnings about the hatred that cruelty breeds and the reprisals it invites, reached the closed room where the case against false confessions did not . The question was never only whether torture works. It was what a state makes of itself when it keeps, for its own safety, a power its courts have given up. Torture could be refuted as a witness; as a servant of the state, it can only be refused.
