srrALKINGTHE CRI~IINAI~ ~IIND

Psychopaths, 'moral imbeciles,' and free willBy David Kelley

Among the books discussed in this essay:Fatal Vuion, hyJoe McGinniss. G. P. Putnam's Sons. 663 pages. $17.95.

Criminology, IOth ed., by E. H. Sutherland and Donald R. Cressey. J. B. Lippincott. 714 pages. (Out of print.)The Tangled Wing, by Melvin Kemner. Holt, Rinehart and Winston. 543 pages. $19.95.

Vulnerabilities to Ddinquency, edited by Dorothy Otnow Lewis. Spectrum Publications. 327 pages. $37.95.Mask of Sanity, 5th ed., by Hervey Cleckley. C. V. MoshyCo. 596 pages. (Out of print.)

The Criminal Personality, by Samuel Yochelsonand Stanton Samenow. Jason Aronson. Vol. I, 538 pages. $35.Vol. II, 578 pages. $40.

Inside theCriminal Mind, by Stanton Samenow, Times Books. 285 pages. $15.50.Criminal Violence,Crimina/Justice, hy Charles Silberman. Random House. 540 pages. $15.

Crime is a social prob-lem; in a sense it is the social problem, becauseit breaks the bond of trust that makes societypossible. But that's about as far as the consensuson the subject goes. On Match 3, for example,the Justice Department released a study show-ing that 40 percent of the people who enteredstate prisons in 1979 were on probation or pa-role for previous crimes-and thus would nothave been free to commit new crimes had theyserved full terms for their earlier ones. The fol-lowing day, the Eisenhower Foundation issued areport denying the efficacy of punishing crimi-nals and urging that public policy address the"real" causes of crime, such as high unemploy-ment among minority youth.

These two reports nearly illustrate the philo-sophical dispute that runs through the debateabout crime. If our actions are a product ofcauses outside our control, then it is unfair-and ineffective-ro blame criminals for what isreally the fault of society, or their parents, ortheir genes. We must try to alter those causes,and use punishment solely as a means of reha-bilitation. If our actions are freely chosen, how-

David Kelley is the authur uf The Evidence' of theSenses, a philosophica1 work on perception.

ever, then society can hold us responsible forthem and refuse to indulge the kinds of excusesthat determinism offers. Punishing wrongdoersis then a form of retribution, and a way of re-moving them from our midst.

For more than a decade, the public has beenmoving steadily into the free will camp. Out-rage over the trial of John Hinckley led Con-gress to tighten the insanity defense. Earlier, inthe 1960s, the sight of social theorists fiddlingwith determinism while the cities burnedhelped elect Richard Nixon on a law-and-orderplatform. The crime rate, despite a recent dip,is well above the level of two decades ago andremains high on the list of public anxieties.Politicians across the spectrum have long sincelearned the electoral advantages of being (orseeming to be) tough on crime, and on crim inals.

Determinism is more difficult to resist incriminology, however, where the goal is to ex-plain criminal behavior. The most powerfulmodels of explanation we have are drawn fromthe physical sciences. TIle social sciences havenot abandoned the hope of finding laws thatgovern human action in the way that the law ofgravity governs the motion of a stone; and jour-nalists who set out to explain particular crimes,

CRITICISM 53

Theamphetamines

did not move hisarm. How can a

chemical causean intention?

54 HARPER'S I AUGUST

to get behind the "story," are drawn ineluctablyinto the search for causes. But the search alwaysruns into problems, problems that arise fromthe very assumption that criminal behavior issolely a product of causes beyond the criminal'scontrol. Thus to solve the social problem ofcrime, we must first confront a philosophical

one. We need to acknowledge the in-rt'" adequacy of determinism.

• he first 610 pages of Joe McGinniss's FatalVision present two sets of "facts" about JeffreyMacDonald, the Green Beret doctor who wasconvicted in 1979 of murdering his wife, Co-lette, and their two children. The first is thecircumstantial evidence against MacDonald,

who has never confessed to the crime; the sec-ond is a mass of psychiatric- testimony, alongwith McGinniss's own speculations, as to whyand how MacDonald could have committed sobrutal a crime.

Though the psychological account has manyloose ends, its outlines are clear enough. Mac-Donald was a hardworking, dedicated, and (toall appearances) compassionate doctor. But hehad an intense desire to control others andshowed great hostility toward anyone who stoodin his way. or even disagreed with him. He wasremarkably unperceptive about his own behav-ior, and in several haunting scenes in thebook-as when he worries about which uni-form he should wear to meet the press after be-

Illustration by Marshall Arisman

ing cleared of murder charges by the military-he reveals an incredible poverty of feeling.McGinniss caught MacDonald in a number oflies, many of them serving no ostensible pur-pose. And despite MacDonald's protestationsthat his marriage was happy, he had had a seriesof casual affairs, apparently fueled by worriesabout his masculinity.

MacDonald had no history of violence. Butif, on the night of the murder, he and his wifehad an argument (there is some evidence thatthey did, and that their relationship had grownincreasingly tense in the preceding months);and if the argument reached a pitch of feeling,with Colette stepping out of her usual passiverole; and if MacDonald saw her as a threat to hismasculinity, to his very sense of self (defined asit was by his ability to control others); and if,having worked for thirty-six hours straight, hewas unable to exert his normal control over hisfeelings-if all this was so, then perhaps we canbegin to understand the murder (of his wife, atleast) as a form of self-defense.

On page 610, however, we are given anotherexplanation, a physiological one: MacDonaldhad been taking diet pills containing amphet-amines. McGinniss conjectures that MacDon-ald took the pills in doses larger than he hasadmitted-doses large enough, according tothe medical literature, to cause psychosis, hal-lucinations, and delusions of persecution.

Yet rieither rhe psychological nor the physio-logical explanation of the crime is very satisfy-ing-not, at least, if we are looking for causalexplanations. Psychology tends to explain anaction by reference to underlying beliefs andgoals. The great advantage of this approach isthat it can make an action intelligible. If Mac-Donald believed that his wife was challenginghis authority, and if his goal was to avoid suchchallenges at all costs, then the act of removingthe threat follows by a kind of brutal logic. Of-ten, an explanation of this sort is enough. Butin the case of a murder it is not. The violentdestruction of one's wife and children is notonly awful and repellent; it wears its awful andrepellent character on its face, visible to anyonenot wholly deranged. For most of us, the enor-mity of such an act would function as a kind ofbarrier reef: the tides of personality would crashagainst it and rebound, shaking loose the grip ofwhatever desires had tempted us.

So we turn with relief to the diet pill hvporh-esis, not only because it is clear and simple butbecause it gives us a real cause, one that mighthave compelled MacDonald to act. No one,after all, chooses the way his neurons react to achemical. But our relief is temporary, for wehave purchased causal necessity at the cost ofintelligibility. We do not yet understand why or

how an amphetamine could trigger an act ofviolence. We do not even know what kind ofexplanation to look for. The amphetamines didnot move MacDonald's arm in the way the windmight move a branch; his arm was guided by hisintention to kill. How can a chemical cause anintention?

If the various explanations of the crime in Fa-tal Vision are finally unsatisfying, the problem isnot literary but metaphysical. We expect therelation between cause and effect to be bothnecessary and intelligible. In the case of a hu-man act, physiology can give us the first, andpsychology the second, but we cannot put thetwo together until we understand (and we do

not) the causal intercourse betweenmind and body, matter and spirit.It has commonly been assumed that science is

the natural ally of determinism. Science, afterall, trades in causal explanations. ImmanuelKant, two centuries ago, argued that the scien-tific perspective leads inevitably to determin-ism, that freedom could be defended only byopposing the authority of science. In WaldenTwo, B. F. Skinner claimed that the increasingsuccess of a science of behavior would make de-terminism more and more plausible. But theprogress of science has not borne out Skinner'sprediction. The problem is not that scientistshaven't discovered any causal influences on hu-man behavior. The problem is that they havefound too many.

No category of human action has been stud-ied in as much depth, or from as many angles, ascrime, Here is some of what we 'have learnedfrom that inquiry:D Young males are disproportionately responsi-ble for crimes of violerice and property crimes.The Baby Boom partly explains the massive risein crime from the early 1960s to the early1970s. Rut only partly. In some areas of thecountry, the murder rate in those years went upten times faster than demographic changesalone would have led one to predict.D Psychologists have found that criminals tendto fall outside the normal range on a number ofpersonality traits. These include some we mightexpect, such as disrespect for authority and di-minished capacity for empathy. But amongthem are also such unexpected traits as hyper-activity and slower response to aversive stimuli.D There is a link between poverty and crime,but it is a complex one. Crime rates are higherin poor areas than in wealthy ones (for violentcrimes, at least), and poor people are more like-ly to be arrested and convicted. But the ratesare higher in urban slums than in rural areas ofequal poverty, and they vary widely among eth-nic groups of the same economic status; poverty

It is not thatscientists havefound no causa!influences onhumanbehavior, butrather too many

CRITICISM 5;

Social normsencourage

violencein boys,

discourage it ingirls. But isn't

gender aphysiological

conditionas well?

56 HARPER'S I AUGUST

per se may not be the crucial variable. There isalso some evidence that crime rates fluctuate inaccordance with the business cycle, suggesting acorrelation, if a weak one, between crime andunemployment.D Delinquents are much more likely to havebeen abused as children than nondelinquents.D The incidence of alcoholism-and, espe-cially since the 1960s, of drug use-is muchgreater among criminals than among the popu-lation at large. There is also some evidence thatabout a third of all serious crimes are committedby people under the influence.D When a criminal has a twin, that twin is atleast twice as likely to be a criminal himself if heis an identical rather than a fraternal twin. Andamong adopted children who commit crimes,the biological parents are more likely to becriminals than the adoptive parents.

This criminological sampler, brief as it is,shows that no single factor is sufficient to ex-plain criminal behavior. This should not comeas a surprise: no social scientist expects to find asingle explanation for any human action. It isprecisely the job of theory to explain how var-ious causal influences interact. But this raisesanother, deeper problem. The factors men-tioned above are of diverse types: economic,cultural, psychological, physiological, genetic.It is far from clear how one should go about ex-plaining the interaction of causes at such differ-ent levels.

The existing theories typically solve thisproblem by denying it. A good example is E. H.Sutherland's theory of "differential associa-tion," which evolved through the ten editionsof Criminology, by Surherland and D. R. Cressey(Sutherland, who died in 1950, published thefirst edition in 1924; the tenth edition was pub-lished in 1978). This theory is still perhaps thebest known in the field. Sutherland and Cresseyhold that criminal behavior is determined bvone's participation in a number of groups: fam~ilv, school, neighborhood clique. We tend toadopt the attitudes of groups we belong to, inproportion to the strength of our ties to thesegroups. We are all pulled in different directionsby competing attitudes toward criminal behav-ior, by different "definitions" (to use Suther-land's term) of the law as something to berespected or flouted. Thus "a person becomesdelinquent because of an excess of definitions fa-vorable to violation of law over definitions un-favorable to violation of law" among the groupshe belongs to.

The theory of differential association cantake into account a number of factors relevantto crime, but only those having to do with socialconditions. The theory is essentially a form oferivironmental determinism, based on the same

model of causality as Skinner's behaviorism. Tomaintain such a reductionist, "single-level" ex-planation of human action, any causes that arenot social must be explained away.

If there is a link between alcohol and crimi-nal violence, for example, Sutherland andCressey suggest that perhaps the offenders have"learned from associations with others certainways of acting when intoxicated." There mayactually be something to this It is very difficultto reproduce in the laboratory the types of be-havior that alcohol induces; social setting doesseem to be a factor. But it is a sign of sociologi-cal desperation to claim that it is the only causalfactor. In some cases, at least, the criminal's in-tent seems to come first; drinking or takingdrugs is a way of nerving himself for the act. Itis also likely that alcohol use and crime are re-lated effects of underlying psychological causes,And what of the effect of alcohol on the brain?

Another example: Many social scientistswould explain the fact that males commit ahigher proportion of violent crimes than fe-males by treating gender as something purelysocial. Social norms encourage violence inboys, discourage it in girls. That is surely part ofthe explanation. It is just as surely not thewhole explanation; gender is a physiological aswell as a social condition. As Melvin Konnerpoints out in The Tangled \X'ing, studies withanimals have shown that testosterone levelsduring key periods in maturation affect the de-gree of aggressiveness in adults. And an in-crease in testosterone lowers the threshold offiring in a nerve bundie called the stria iermina-lis, which is part of a neural circuit known to beinvolved in violent behavior.

Or consider the appalling incidence of childabuse in the families of delinquents. DorothyOtnow Lewis and her associates divided theyouths at a Connecticut correctional schoolinto two groups, according to the severity oftheir crimes. As she reportsinVu!nerabilities toDelinquency, 75 percent of the more violent of-fenders had been abused as children, as against33 percent of the less violent. And 79 percentof the first group (compared with 20 percent ofthe second) had witnessed extreme violence:they had seen their mothers slashed,· their sib-lings burned with cigarettes. It would be hard tofind more compelling evidence that one's envi-ronment can have devastating effects. But it islikely that the violent behavior of these youthsflowed from physiological as well as emotionaldamage they suffered as children. Virtually allof the more violent offenders had neurologicaldisorders, and 30 percent of them (as againstnone of the less violent offenders) had grosslyabnormal electroencephalograms and/or histo-ries of grand mal seizures.

Environmental determinists are aware thatbetween the stimulus and the response lies avery complicated piece of equipment-the hu-man organism. But they regard the internalproperties of that organism as mere "interven-ing variables," to use Skinner's phrase: conduitsthat pass along, unmodified, the stream of envi-ronmental forces. The Tangled Wing is an ex-haustive demonstration that this view is false.Pulling together evidence from genetics, bio-chemistry, ethology, and neurophysiology,Konner shows that the intervening variables arein fact the controlling ones. An animal's genet-ic endowment determines which stimuli it canrespond to and the kinds of responses thosestimuli are most likely to elicit. Even among hu-mans, action flows in large part from emotionsthat have their origins in the interplay of hor-mones and neural structures that were shapedby selection pressures over the course of a mil-lion years,

This is not to say, of course, that the envi-ronment is irrelevant, merely that environmen-tal determinism is as narrow and simple-mindedas genetic determinism. As Konner writes,"Any analysis of the causes of human naturethat tends to ignore either the genes or the envi-ronmental factors may safely be discarded."

But such ecumenism has a cost that Konnerdoes not fully appreciate. The evolution of thenervous system, from the simplest reflex arc tothe human brain, has been a process of inter-posing longer and more complex loops betweenstimulus and response. As control of behaviormoves inward, action replaces reaction, the or-ganism becomes an agent, and we have to con-sider the possibility that the whole is more thanthe sum of its parts.

Despite its sophistication, The Tangled Wingis still reductionist, filled with confidence that ascientific understanding of the parts will add upto an understanding of the whole. "When wehave characterized the biology of moods," Kon-ner suggests, in one of many such statements,"we will have characterized the major forces be-hind behavior." But the riches that await us inbiological research, and doubtless they aremany, will leave Konner's account of humanbehavior overdrawn, his confidence in reduc-tionism insufficiently funded by the evidence.Indeed, a number of prominent biologists, suchas Nobel Prize winner Roger Sperry, have con-cluded that behavior will never be understoodfully at the neuronal or biochemical level; andthey have revived the view that qualitativelynew and irreducible properties emerge in a bio-logical system as it becomes more complex.

In the human brain, the massive expansionof the frontal lobes made possible two traits thathave always seemed to distinguish man from

other animals: the capacities for self-awarenessand for abstract, conceptual thought. Konnerhas almost nothing to say about these capaci-ties, or about the fact that they enable us tomodify and override the more primitive re-sponses of evolutionarily older parts of thebrain. Yet if Sperry and his allies are correct,these capacities are examples of "emergent"properties beyond the reach of any reductionistexplanation. The only hope of understandingthem lies in a more holistic approach, one thatcrosses the mind-body divide and examinesthem as traits of a conscious self.

The result of such an examination could bemerely a more complex deterministic accountof human action. But it could also be that thecapacities for conceptual thought and self-awareness represent an evolutionary change ofkind, not merely degree. They do not, ofcourse, break the bonds of determinism alto-gether: we are still constrained by our geneticand physiological equipment, and can hardlyremain unaffected by uur social environment.But if the human agent, the self, is more thanthe sum of its parts, then our actions may bemore than the sum of their antecedents; we may

have room to maneuver within thecausal net,

CriminOlOgy texts routinely denounce thesearch for criminal man-for a set of personal-ity traits peculiar to criminals. And no doubt agood deal uf nonsense has been perpetrated inthe course of this search. But there is in fact apersonality syndrome that one encounters atevery turn in the literature on crime. The typewas formerly known as the psychopath or socio-path; in the current edition of the psychiatrists'Diagnostic and Statistical Manual, the syndromegoes by the anemic name of "antisocial person-ality disorder."

The psychopath is not the bug-eyed psychot-ic who serves as a wild card in Hollywood crimedramas. Psychopathy does not involve any clearpsychosis or neurosis; that is why it is classifiedas a personality disorder. Perhaps the most re-vealing name for the syndrome is the oldest.Psychologists in the nineteenth century identi-fied a disorder that seemed to involve no cogni-tive impairment-those who had it were oftenquite intelligent and clearheaded-but rather agross deficiency in what used to be called themoral faculties: the capacities for deep feeling,working toward goals, living according to stan-dards, cooperating with others. These peopleseemed profoundly amoral. Despite their intel-ligence, they were unable to look beyond theimpulse of the moment. They seemed constitu-tionally incapable of empathy and lacked eventhe most elementary sense of fairness and reci-

If the self ismore than thesum of its parts,then our actionsmay be morethan the sumof theirantecedents

CRITICISM 57

The psychopathi.~perhaps best

seen as aprototype to

which criminalsconform moreor less closely

58 HARPER'S / AUGUST

procity. It was as if a human intelligence hadbeen planted in the brain of an innocently pred-atory animal. The psychologists called these in-dividuals "moral imbeciles."

The classic clinical portrait of the psycho-path was drawn by Hervey Cleckley, who was atherapist in private practice, in Mask of Sanity;some researchers still use his list of sixteen traitsas a diagnostic tool. Cleckley's subjects exhibit-cd the normal range of intelligence; many werewell informed, many were talented. They werenot delusional, and seemed entirely free ofanxiety. Yet they seemed unable to learn fromexperience, making the same mistakes over andover, even after they had recognized them. Hissubjects were chronic liars, even when no cleargain was involved. Cleckley came to believethat the intelligence they showed in conversa-tion was merely verbal; as he studied them moreclosely, he was struck by the concreteness andfragmentation of their thinking, reflected intheir complete lack uf interest in long-rangeplanning for their lives.

Most of his patients, especially those withcriminal records, were able to size up peoplequickly; they were good at manipulating othersand mimicking conventional feelings and atti-tudes when it served their purposes. Yet at othertimes their actions revealed an inability to an-ticipate how others would react. (One woman,in applying for jobs, routinely gave as referencespeople whose trust she had repeatedly violat-ed.) Cleckley was most struck by the poverty offeeling these people exhibited. Primitive emo-tions-spite, vanity, sentimental affection,flashes of violent anger-came and went likeNew England weather, but there was no indica-tion that they experienced deeper, more com-plex emotions, such as grief, pride, joy, despair,or love. His patients were often witty, but neverrevealed any genuine sense of humor. Theiregocentricity was so profound as to differ inkind from ordinary self-centeredness. Yet de-spite their indifference to the suffering theycaused others, their obliviousness to moral stan-dards, and their incapacity for feeling shame,humiliation, or regret, they were quick toblame others and to defend themselves whencriticized. Moral evaluation mattered to themin a way that belied the appearance ofamorality.

Cleckley's patients were not all criminals,nor do all criminals fit the pattern he de-scribed. The degree of overlap is hard to esti-mate. Researchers using the Minnesota Multi-phasic Personality Inventory have found thatprison inmates score well above the generalpopulation on the "psychopathic deviate" scale;however, that scale is a fairly crude measure.The psychopath is perhaps best seen as a proto-

type to which criminals conform more or lessclosely.

Any doubts about the existence of a link be-tween crime and psychopathy have heen dis-pelled by the work of Stanton Samenow andSamuel Yochelson, who conducted a fifteen-year study of criminals at Sr. Elizabeths Hospi-tal, a federal psychiatric facility in Washington,D.C. Their two-volume work, The CriminalPersonality (summarized in Samenow's more re-cent Inside the Criminal Mind), offers a clinicalportrait remarkably similar to Cleckley's. Yo-chelson and Same now found the same con-creteness in thinking that others have noticed.Their patients' short attention spans made itnext to impossible for them to take a long-rangeview of their lives. They rarely learned from ex-perience. Their non-integrative cognitive stylemade it difficult for them to see any contradic-tion between their violent, predatory behaviorand the sentimentality they often expressed to-ward the helpless. The career criminals Yochel-son and Same now studied tended to view"straight" life as a series of concrete acts, mostof them boring. These people lived in the mo-ment, and did not see the value of the long-term rewards of a family or a career. Thesecognitive traits have a common root; an anti-conceptual mode of thinking. For it is the pow-er to conceptualize that makes us ahle to act onprinciples, to think in terms of long-rangegoals, and to learn from experience.

The psychopathic syndrome also involves acertain self-conception. The psychopath wastraditionally considered less susceptible to fearand anxiety than other people. That, indeed,was the basis for one explanation of the syn-drome: psychopaths' insensitivity to punish-ment hinders the process of socialization. In thecourse of their interviews, however, Yochelsonand Samenow found their subjects to he in-tensely fearful.

Their greatest fear, Yochelson and Samenowfound, was that of "the zero state." This sense ofcomplete and profound worthlessness was some-thing all of their patients had experienced, andwent to great lengths to repress. They protectedthemselves against it by a kind of grandiosity, aconception of themselves as supermen, as ef-fortless heroes able to achieve great ends by un-conventional means. Their chief method ofsustaining this self-image was to exert controlover others. By forcing others to bend to hiswill-intimidating them, manipulating themthrough lies and cons-the psychopath makessociety affirm a view of his potency that he can-not affirm by looking within.

Conversely, anything that suggests a lack ofcontrol over the world threatens to bring on thezero state. According to Samenow, "The threat

of being less than top dog, the possibility thathe won't achieve unusual distinction, thechance that things will not go as he wants con-stitute a major threat to the criminal, almost asthough his life were at stake. From his stand-point it is, because the puncturing of his inflat-ed self-concept is psychological homicide."Anyone trying to understand the case of JeffreyMacDonald should find that a chilling obser-vation.

Theories about the causes of psychopathy-like those about the causes of crime-are nu-merous and varied. Most if not all of the traitsof the psychopath have been observed in peoplewith neurological damage. And it is hard to be-lieve that neural damage had nothing to dowith the violent behavior of the delinquentsstudied by Dorothy Ornow Lewis. But as theeminent neurologist Frank A. Elliott has noted,"organic disorders tend to produce a 'partial'psychopath rather than the fully fledged classi-cal picture."

There are also sociological explanations. InCriminal Violence, Criminal]ustice, Charles Sil-berman describes the brutality of crimes com-mitted by juvenile delinquents, often withoutremorse. Silberman attributes this to the factthat they "have been so brutalized in their ownupbringing." More generally, he suggests thatcrime usually springs from an impoverished self-conception, caused in tum by economic pover-ty: "In a society that rewards success andpenalizes failure ... to be poor is to live withcontinual self-doubt." But this cannot be thewhole story, unless we assume-and the as-sumption is often made by social scientists, usu-ally without benefit of evidence-that the in-dividual derives his self-esteem exclusively fromthe responses of others. That assumption leavesno way to account for the fact that people differin precisely this respect: the autonomy of theirself-estimates.

Cleckley, for his part, held that psychopathy isa deeply rooted disorder, an abnormality moreprofound even than schizophrenia. Though thepsychopath presents a mask of sanity to theworld, his actions reveal that the mask "disguisessomething quite different within, concealing be-hind a perfect mimicry of normal emotion, fineintelligence, and social responsibility a grosslydisabled personality." Yochelson and Samenowmaintain that the problem lies much closer to thesurface, in patterns of thinking that are accessibleto consciousness and-with some effort-sub-ject to conscious control.

They have discovered, for example, a phe-nomenon they call "cutoff," a severe form ofanti-conceptual thinking that allows someoneon the threshold of committing a crime to blankout all of his fears and doubts. This act of blank-

ing out is voluntary: "Even though cutoff is sorapid and automatic, it is still a mental processthat is under the criminal's control. Whetherhe invokes the cutoff is his choice." As evi-dence of volition, they note that criminalslearn not to shut off their fears roo soon, lestthey dull themselves to signs of danger.

By the time of Yochelson's death in 1976,thirteen of the thirty patients in the specialtherapeutic program at St. Elizabeths were lead-ing responsible lives-a major achievement,given the dismal record of criminal rehabilita-tion, and a sign that patterns of thinking areamenable to change. The agent of change, asthe authors describe it, is the insistence that apatient learn to monitor his thoughts and toprevent his fragile self-image from blocking outhis awareness of what he has done, of who he is.

Yochelson and Samenow came to believe thatcrime is a voluntary act for which the criminal isfully accountable. This is not, to say the least, themajority view in criminology, but it is not surpris-ing that they adopted it. For in tracing the roots ofcrime to problems in the criminal's ability tothink conceptually and to form a self-concep-tion, they arrived at the two uniquely human

traits to which anti-determinists in allT fields have always appealed..he conflict between free will and determin-ism first arose in philosophy, and most of thephilosophical arguments for human freedomhave been variations on a common theme. Be-cause we are capable of self-consciousness, it isclaimed, we can focus attention on an impulseor feeling and examine it from a kind of innerdistance that can weaken its aura or grip. Be-cause we are capable of conceptual thought, wecan evaluate these impulses and feelings-theirconsequences, their effects on others, theircompatibility with our principles-and choosewhether to act on them. We are free agents he-cause those capaci ties give us veto power overthe forces that move us.

Determinists have always found this argu-ment naive: science, they say, will show thatbehavior is governed by causes beyond thereach of conceptual thought and self-awareness.But in the case of crime, at any rate, the trail ofscientific inquiry keeps circling back to thosevery capacities. It would be too much to saythat science can establish human freedom.That will always be a philosophical issue. Butthe old assumption that science is a witnessagainst free will is not true, either-it will notsurvive a close look at what scientists have actu-ally discovered. Human beings have turned outto be far more complicated than the sciences ofman anticipated. We may just turn out to be ascomplicated as we always thought. •

Because we arecapable of self-consciousness,we can focusour attentionon an impulseand examineit from aninner distance

CRITICISM 59