Elaine Scarry and Edward Peters both teach at the University of Pennsylvania—she teaches English, and he medieval history. Peters has just published a history of torture, and Scarry a book that takes torture as its starting point. There the resemblance ends. Peters has written a straightforward historical account of torture, from the ancient world to the most recent Amnesty International report. Scarry has much greater ambition, as her subtitle announces. She claims to find, in the nature of pain, the key to understanding human creativity and human destructiveness.

It is not easy to explain briefly—or indeed at any length—just what The Body in Pain is about. In her introduction, the author tells us that the book has only a single subject, but that this subject can itself be divided (like the Christian god?) into three different subjects: the difficulty of expressing pain, the political and perceptual complications that arise as a result of that difficulty, and the nature of human creation. The reader may wonder how the third of these subjects is linked with the other two. For Scarry, the connection lies in seeing human creation as a form of expression, and hence as arising from the initial problem of expressing physical pain.

Scarry asks us to picture these three subjects as concentric circles. The “body in pain” of her title suggests the inner space of private sensations, but once we look at how that pain is expressed we find ourselves dealing with a public and, indeed, a political situation; we then find, she claims, that we have all along been standing at the center of human creative activity itself.

Overlying this trinity of subjects is a distinct two-part division—between “unmaking” and “making.” Physical pain inflicted in the process of torture, Scarry believes, is a method of destroying—or, as she prefers to say, “deconstructing”—the language, and indeed the world, of the victim. Yet this process has, she claims, a structure that is part of a general structure of “unmaking.” Next Scarry explores the nature of war, and finds the structure of unmaking present here too. She then discovers, in “unmaking,” the mirror of “making,” and so goes on to examine the nature of creation, both intellectual and material. Much of this latter section is in fact an exploration of two texts: the Bible and the writings of Karl Marx. A final chapter entitled “The Interior Structure of the Artifact” seeks to show that made objects are attempts to make the world “sentient” and that they are projections of those who make them which have a reciprocal effect, “remaking” those who make them. Thus Scarry has traveled from the individual human body in pain to the most universal aspects of human civilization: what it is to live in a created world.

This brief outline may indicate the scope and some of the content of The Body in Pain, but it can give little idea of the author’s distinctive mode of handling her subject. To convey this, we must focus more narrowly on specific issues.

The Body in Pain deals directly with torture for little more than thirty pages. To this we can add another twenty pages discussing one difference between torture and war. Although these sections together amount to less than an eighth of the total length of the book, it is appropriate to examine in detail what Scarry says about torture. It is the subject of her first chapter; moreover torture is, she says, “absolute pain,” the extreme example of pain in which its structure as an experience and its relation to reality can first be discerned. I do, admittedly, have an additional motive for concentrating on this topic. The coincidental publication of Peters’s Torture provides us with an independent account. By comparing the two, we may learn something not only about torture, but about Scarry’s methods of argument.

Scarry’s account of torture makes a large number of assertions, directed toward her larger thesis. Some of these assertions are straightforwardly factual; others are clearly intended to have a deeper philosophical meaning. The best way to convey something of the style of this unusual work is to quote some key passages. Here is the first:

The physical pain [inflicted in torture] is so incontestably real that it seems to confer its quality of “incontestable reality” on that power that has brought it into being. It is, of course, precisely because the reality of that power is so highly contestable, the regime so unstable, that torture is being used.

Scarry suggests that torture is used by unstable regimes in an attempt to assert the “incontestable reality” of their power. The passage is characteristic of The Body in Pain in that it mingles an assertion of fact which we can check—the use of torture is associated with unstable regimes—with the philosophically laden suggestion that physical pain has some special “incontestable reality” which can seem to belong to the power that inflicts it.


If we begin by scrutinizing the claim that can most easily be checked, we soon find grounds for doubt. Peters’s Torture describes how torture developed in medieval and early modern Europe primarily as a means of gathering evidence in ordinary criminal cases. In a moment we shall examine this development more closely; but all we need to note here is that the regimes he discusses under which this practice flourished were relatively stable. Moreover, since torture was frequently inflicted to decide on the truth of an accusation brought by an ordinary citizen, against another ordinary citizen, for a wrong done by the accused to the accuser, the continuance or legitimacy of the regime itself was usually not at issue. In France, for example, torture had been carried out for centuries by the ancien régime, which cannot have been very worried about its own stability until the late eighteenth century—the very time at which torture in France was abolished. This took place substantially in 1780 and entirely in 1788; in other words, before the Revolution. It is worth noting, too, that the extremely unstable governments which ruled in an atmosphere of suspicion and conspiracy between 1789 and the ascendancy of Napoleon did not revive the use of torture. Torture was also used in criminal procedures in the Ottoman Empire and in Japan during the Tokugawa period—both long-lived forms of government.

Scarry might reply that she had in mind the use of torture in modern times, and not the history of its development. This reply would be a concession; it would detract from the appearance of universal application with which Scarry seeks to invest her account of pain in general, and of torture in particular. The quotation would cease to be a timeless truth about “the body in pain” and would become a limited point about some bodies in pain in some historical circumstances. Even thus circumscribed, however, the claim would be difficult to defend. Was the reality of the power of Fascist Italy, Nazi Germany, or Stalinist Russia truly so “highly contestable”? If the use of torture is related to the “contestability” of the reality of political power, one might have thought torture more likely to be employed in the declining years of the Weimar Republic.

If we are seeking an explanation for the revival of torture in the twentieth century, Peters’s discussion of the complex set of factors involved is more helpful than anything Scarry says. He places special weight on the nature of the new nation-state that emerged after 1917: a state in which a particular political party claimed to be the embodiment of both state and law, and hence able to act without the usual constraints of the rule of law. The attitude was well expressed by Frick, the Nazi minister of the interior: “Law is what serves the German people. Injustice is what injures it.”1 Torture has also been used by liberal-democratic states, for instance by France in Algeria in the 1950s. Peters links this to colonialism and the way in which those working for the colonial power tend to deny fully human status to the population over which they rule. He does not pretend, however, to have an explanation for the prevalence of torture in authoritarian countries today, admitting that it is “hardly possible even to guess at an etiology.”

So the factual basis for the assertion behind the first passage quoted does not stand up to examination. When we start to discuss Scarry’s claims about the “contestability” of the reality of power under particular regimes, we are slipping from the straightforwardly factual to the philosophical. We have to ask what she means when she says that the reality of physical pain in torture is “so incontestably real” or that the reality of the power of a regime is “so highly contestable.” Scarry cannot be using such expressions in their ordinary sense. To those not persuaded by Berkeley’s philosophical idealism, the iron bars of a prisoner’s cell are just as “incontestably real” as any pain that torture might inflict. Nor is it the reality of the power of a regime that its opponents wish to contest; it is the legitimacy of that power which is contested. If the regime did not wield real power, it would be unnecessary to struggle against it. But Scarry is not using terms like “reality” in any strict philosophical sense either, for there is no generally accepted philosophical basis for saying that physical pain is more “real” than ordinary material objects like cell bars. (The nature of ultimate reality is, of course, one of the perennial fields of philosophical contest.) In the absence of any clear meaning for the language which Scarry uses, it is difficult to know what to make of her claim that the physical pain inflicted by torture has some special quality of “incontestable reality.”


The conventional view of torture is that it is usually inflicted during a process of interrogation, in order to obtain information which the person under interrogation will not willingly yield. In another representative passage, Scarry seeks to turn this view on its head:

As the content and context of the torturer’s questions make clear, the fact that something is asked as if the content of the answer matters does not mean that it matters. It is crucial to see that the interrogation does not stand outside an episode of torture as its motive or justification: it is internal to the structure of torture, exists there because of its intimate connections to and interactions with the physical pain…. For the torturers, the sheer and simple fact of human agony is made invisible, and the moral fact of inflicting that agony is made neutral by the feigned urgency and significance of the question…. In compelling confession, the torturers compel the prisoner to record and objectify the fact that intense pain is world-destroying.

Scarry holds that interrogation is a part of torture, rather than the other way around, and that the role of interrogation is to disguise from the torturers the fact that they are inflicting human agony. This is a novel idea, but it appears, again, to have been put forward in ignorance of the history of torture. Let us look at Peters’s well-documented account of how torture developed in medieval Europe.

Until the twelfth century, when a freeman was accused of a crime he generally needed only to swear an oath that the accusation was false. In some cases he might have to bring oath helpers, others who would swear an oath in support of his own oath. (They did not have to be witnesses to anything relevant to the alleged crime; their role was simply to testify to their belief in the oath of the accused.) These oaths normally put an end to the case against the accused. If the accused was of infamous character, however, he might have to undergo an ordeal, such as being bound with stones and thrown into a pond, in order to allow God to judge the truth of the accusation; and sometimes, for the same reason, the accused and accuser might have to engage in combat.

As Romano-canonical law developed in the thirteenth century, these primitive and irrational modes of proof gave way before the development of the inquisitorial process. This process eventually set up a standard of proof which meant that legal proof of guilt of a serious crime could come only from the evidence of two eyewitnesses, or from the accused’s own confession. If there were not two eyewitnesses, and the accused would not confess voluntarily, torture was the only means by which a conviction could be obtained. It was to be applied, however, only when a prima facie case against the accused already existed, and even then the confession made under torture was not itself proof; it had to be repeated away from the place of torture, and ideally it should contain details of which only the perpetrator of the crime could be aware.

Since the actual practice of torture often was at variance with this relatively rational—even if appallingly insensitive—jurisprudence of torture, it might be doubted whether the existence of torture really had all that much to do with the need for legal proof of a crime. Was there perhaps a deeper explanation, as Scarry wishes to suggest? But any other explanation would have to account for the disappearance of torture from the English legal system after 1166. In that year, Peters tells us, Henry II eliminated the power of royal officials to undertake criminal prosecutions on their own, and instead created the grand jury—a group of citizens who presented an accusation to a traveling royal judge. If the accusations made out a case, the trial itself was held before another jury. The rules of evidence were broad enough to allow the accumulation of circumstantial evidence until the ordinary citizens who made up the trial jury were convinced of the guilt of the accused. There was thus no legal need for torture, and—with the exception of canon law, which existed in England as it did on the Continent—torture was not part of the English legal process. This fact gives powerful support to the view that the existence of torture on the other side of the English Channel was related to the forms of proof required by Continental law. The interrogation was crucial; it was nearly always the justification, and probably often the genuine motive, for torture.

What of Scarry’s further claim, made in the same passage, that “the sheer and simple fact of human agony is made invisible” to the torturers by the process of interrogation? This claim is paradoxical, for one might have thought that to be an efficient torturer, one would need to be well aware of when one was causing pain and when one was not doing so. Peters presents a very different picture, citing accounts of the training of torturers under the junta of the Greek colonels, and other testimony to the effect that torturers are themselves beaten as part of their training, so they presumably know what it feels like to be beaten. It is, Peters says, a “fabricated political reality” which the torturers are trained to accept—political, not metaphysical. Instead of attempting to make invisible the agony of the victim, this fabricated political reality puts their victims outside the pale of humanity, thus rendering what is done to them morally acceptable; something very different from, and surely more plausible than, what Scarry would have us believe.

Incidentally, if we did accept Scarry’s claim that for the torturers “the sheer and simple fact of human agony is made invisible,” we might well wonder what to make of her immediately following suggestion that “the moral fact of inflicting that agony is made neutral” by the interrogation procedure. If the agony is invisible to the torturers, how can there be a “moral fact” (whatever that might be) of inflicting it? We are not usually morally troubled by consequences of which we are not aware.

Enough has been said to show that The Body in Pain does not contain an accurate description of torture. This defect can be traced to the author’s blithe disregard for the ordinary canons of argument. For instance, on so peculiar—and yet verifiable—a contention as the invisibility of agony to the torturers, the reader naturally looks for supporting evidence. None is offered. Although she cites many accounts of torture from such countries as Greece and Brazil, Scarry makes no attempt to glean supporting information on this point from the considerable corpus of testimony by torturers or by their victims.

Then we come to the further contention that torture seeks to destroy the world of the victim and to objectify this destruction in confession. Again, no evidence is provided; indeed, Scarry disarmingly concedes: “For the most part, neither actual nor fictional accounts of torture focus on the process of disintegrating perception brought about by great pain and objectified in confession.” She then purports to find “a partial description of the process” in Sartre’s short story “The Wall,” a story which in fact is about execution, not about torture. This is, however, sufficient excuse for three or four pages of discussion of the story, complete with references to Sophocles, Shakespeare, Beckett, Wilde, Zola, Christ, and Wuthering Heights. This style of “argument,” if it can be called that, is characteristic of the entire book. Fictional writing may, of course, provide valuable insights into what people experience, what they may do, and the motives from which they act; but fiction cannot serve as the sole basis for claims about the way things actually are. Perhaps Scarry has forgotten that she is not writing a work of literary criticism. It is difficult to think of any other explanation for her failure to support her specific claims about torture by providing—in addition to her discussion of fiction—evidence drawn from the real world in which torture takes place.

I shall quote just one more example of Scarry’s methods of attempting to carry conviction. It occurs in a passage in which she is exploring the differences between war and torture. In contrast to the possibility of intelligent argument on behalf of war, she says:

There has never been an intelligent argument on behalf of torture (and such an argument is a conceptual impossibility).

To make such a statement is either to fly, once again, in the face of evident fact; or to abuse the word “intelligent” by making it refer to a quality closer to moral sensitivity than to any intellectual ability. The historical perspective provided by Peters is once more a useful corrective to Scarry’s hyperbole. From the thirteenth century until 1780, when Pierre Muyart published his defense of torture against Cesare Beccaria’s attack on the practice, some of the most learned of European jurists had argued on behalf of torture. There have been other, more recent defenses, including the controversial defense of torture in Algeria by General Jacques Massu in his memoirs, La Vraie Battaille d’Alger.

Scarry is aware of such defenses; she even has a long footnote discussing a defense of torture by the American philosopher Michael Levin.2 Curiously, in view of her statement that an intelligent defense of torture is a conceptual impossibility, Scarry never attempts to show that the objections to Levin’s argument are conceptual. In fact, Scarry makes heavy going of the relatively simple task of refuting the moral argument for accepting torture, because she confuses legal and moral culpability. She attempts to show that whatever the extraordinary circumstances—a nuclear bomb planted in a city building by a madman who will reveal its location only if tortured—that might lead to a jury exonerating a person who inflicted torture under the pressure of this situation, there is no need for any guarantee of exoneration to be provided before the fact. But of course a defense of the moral legitimacy of torture does not need to be linked to prior legal exoneration. The real problem with the use of such hypothetical cases is, as Peters notes, that they are hypothetical, and when torture is allowed in the real world, it is never so tightly restricted.

I would grant, very willingly, that none of the defenders of torture has shown sufficient concern for the pernicious effects of an institutionalized practice of torture. Probably all of them have in mind an idealized practice of torture which does not, and within the limits of human nature could not, correspond to the actual practice of torture; but to say that none of these defenses is intelligent is to be slack with words. To say that it is a “conceptual impossibility” that there should be an intelligent defense of torture suggests either a lack of understanding of the nature of conceptual impossibility, or an attempt to bluff one’s readers into accepting a claim which manifestly cannot withstand close examination.

If now we move away from the specific topic of torture to the broader purposes to which Scarry wishes to put her account of pain, we find the thesis of her book built on statements like this:

To acknowledge the radical subjectivity of pain is to acknowledge the simple and absolute incompatibility of pain and the world. The survival of each depends on its separation from the other.

Here we have a central theme, the attempt to tie together the overcoming of pain and the “making” of the world. There are, of course, many ways in which this theme could be stated. It could be said, for instance, that extreme pain involves the “unmaking” of the world, in the sense that extreme pain drains everything we value of all significance. Such a statement would presumably be too banal for Scarry; she cannot resist the more dramatic, the seemingly more profound idea of the “simple and absolute incompatibility of pain and the world.” Unfortunately this idea is plain nonsense. Pain does exist in the world. That is precisely the problem.

There is another, more familiar way, in which the overcoming of pain (broadly understood) has been linked to the “making” of the world. Marx argued that the human world is a product that arises from the way in which human beings set out to satisfy their material wants—their need for food, shelter, clothing, and so on. If we accept that human beings who cannot satisfy these needs are in pain, then the idea that the “making of the world” starts from pain is merely another way of stating one of Marx’s most basic beliefs.

Something like this appears to be at least one of the things Scarry wants to say in her discussion of Marx. If so, it is a point which has already been more clearly made in many previous accounts of the central role of material production in Marx.3 Scarry does, however, say other things about Marx. She is keen to see him as providing an “analysis of the material artifact.” She even describes him as “our major philosopher on the nature of material objects,” an accolade that must cause shudders to students of Hobbes, Locke, Hume, and Kant. It would rightly have been rejected by Marx himself, since he never made any attempt to develop a systematic philosophical position about materialism. That task fell, on Marx’s death, to Engels, to whom we owe the very term “dialectical materialism.”

As part of this project of making Marx into a thinker interested in material objects per se, rather than in the effects such objects have on our way of living, Scarry says things that betray a failure to grasp what is essential in Marx’s thought. She writes, for instance:

The tool has an important place in Marx’s writing. It restores the referent because it mediates between worker and artifact, and thus when the image of the tool is held steadily visible, the original site of human projection is held visible as well. For this reason the tool is often taken as a summarizing sign of Marx’s work.

The real reason why the tool has an important place in Marx’s work is that Marx’s theory of history asserts that the productive forces—that is, the tools—with which we seek to satisfy our wants determine the relations of production, on which the whole political, social, religious, philosophical, and moral superstructure of our society is built. As Marx put it, in perhaps his crudest formulation: “The handmill gives you society with the feudal lord; the steam mill, society with the industrial capitalist.”4 Could anyone who understands this absolutely elementary point of Marxism have written the passage I have just quoted?

Another point revealed by that passage is that, whether Scarry understands Marx or not, she has succeeded in creating a new style for the discussion of his ideas that is notably more difficult to understand than the original. Here is another example:

If the monumentally complex substance of Capital were to be described in a single sentence, it could be described as an exhausting analysis of the steps and stages by which the obligatory referentiality of fictions ceases to be obligatory: it is an elaborate retracing of the path along which the reciprocity of artifice has lost its way back to its human source.

It is not, however, for the details of its account of Marx—or of its equally lengthy interpretation of the Bible—that The Body in Pain most needs to be criticized. The real problem is the manner in which the author treats her subject. The attention given to the Bible and Marx is, like the preference for examining a short story by Sartre rather than the usual source materials familiar to historians, characteristic of the entire project. The book does contain suggestive insights on particular points: for instance, on the way in which a soldier, considered as a worker, becomes inseparable from the materials of his labor; or on nuclear war as in some respects closer to the model of torture than to the model of conventional war. There are also occasional passages of fine writing which can be placed alongside the turgid sentence on Marx just quoted (Scarry’s description of the soldier merging with his unit and his environment is again an example). But the book is cavalier in its disregard for the hard work of providing either factual evidence or serious philosophical argument for what it says. One of the quotes on the dust jacket gives extraordinary praise to the book as a high point in a recent “resurgence of innovative writing in the literary-critical mode.” If this book exemplifies such a mode of writing, I hope that the resurgence will prove short-lived. Given a choice between this particular kind of “innovative writing” and the solid virtues of a careful, scholarly book like Peters’s Torture, I will do without innovation every time.

This Issue

February 27, 1986