Introduction

Philip Zimbardo’s 1971 Stanford Prison Experiment randomly assigned young male volunteers to the roles of guards or prisoners for a two-week period. Zimbardo was interested in how an individual adapts to a new environment and role. In 1971 he set up a mock prison in the basement of the Stanford University psychology building. The guards soon dehumanized and abused the prisoners to the extent that many of the prisoners wished to withdraw from the experiment and he had to call off and shut down the experiment after only six days. The simulation raised philosophical and social science questions about the nature and source of evil, whether humans can liberate themselves from the constraints and alienation of society, and the capability of the social sciences to answer the question of what makes people do what they do. It also started a debate over human nature, the social sciences, and research ethics that has lasted almost 50 years.

The widely used tutorial website Collaborative Institutional Training Initiative (CITI) module on history and ethics identified the ethical problems with the Zimbardo study as harm to participants and a degree of detachment on the part of the researcher:

“This landmark psychological study of the human response to captivity and, in particular, prison life, involved assigning roles to normal male student volunteers to create groups of ‘prisoners’ and ‘guards.’ The research became so intense, as physical and psychological abuse of ‘prisoners’ by ‘guards’ escalated, that several of the subjects experienced distress less than 36 hours after the study began. Dr Philip Zimbardo, the researcher, did not stop the experiment/simulation until six days had passed.”¹

The simulation was enhanced by the role-playing of five participants. 8612 became a rebellious political prisoner and Arnett, a sociology grad student, became one of the most sadistic/ extremely authoritarian guards.² In response to a request by prisoners for religious services, Zimbardo recruited a priest to play prison chaplain, had an ex-convict advise him on prisons and serve as chair of the parole board, and allowed a public defender to hold official lawyer’s visits with prisoners. All three acted as they would with real prisoners.

In general, investigators file a conflict of interest form when submitting their application to their Institutional Review (IRB) board for approval to carry out their proposed research. While most research conflict of interest statements focus on financial interests or personal and family member gains, some non-financial considerations may influence or compromise professional judgment in a clinical trial or research experiment. In Zimbardo’s case, he had a conflict of professional interest in his dual role of Principle Investigator and Prison Superintendent. He finally ended the experiment when confronted by Christina Maslach, his former teaching assistant and current girlfriend, whom he asked to come to the mock prison on the fifth day. She was shocked by what she saw and, as a result, he ended the experiment the next morning.

In April 2018, Thibault Le Texier, a Ph.D. economist and associate researcher at Nice Sophia Antipolis University in France, published Historie d’un mensonge Enquéte sur l’expérience de Stanford (History of a Lie: Investigation of the Stanford experience) which he called “one of the greatest scientific deceptions of the 20th century.”³ Le Texier found a conversation taped on day three of the simulation in Zimbardo’s archives at Stanford University in which Zimbardo told his staff that two prisoners had come in the day before and said they wanted to leave. Zimbardo had told them “no,” and informed them that the only way they could leave was a medical or psychiatric reason. Zimbardo added that he thought the prisoners really believed they can’t get out. Le Texier noted that the informed consent form that Zimbardo’s subjects signed, which is available online from Zimbardo’s own website, contain no mention of the safe phrase “I quit the experiment.”^4,5

Under today’s guidelines, this would violate research ethics since investigators are required to inform participants that they are free to quit the experiment at any time and may or may not require a safe phrase. After interviewing Zimbardo, science reporter Brian Resnick⁶ thought that the results of the prison experiment seemed unscientific and untrustworthy and doubted that it should be the basis for enduring lessons in psychology.

The Rise of Policy to Protect Human Subjects

The experiment was carried out in August 1971. This was just before the Institutional Guide to DHEW Policy on Protection of Human Subjects,⁷ known as ‘The Yellow Book,’ was published on December 1, 1971. The Guide reflected several years’ experience with an earlier public health service policy. It was designed to be flexible on what can or should be done and depended upon common sense and sound professional judgment. It explicitly stated that one basic element of informed consent is an instruction that the subject is free to withdraw his consent and to discontinue participation in the project or activity at any time. It mentioned that compensation to volunteers should never be such as to constitute an undue inducement. However, Zimbardo’s simulation was funded by the Office of Naval Research, Department of Defense which, at the time, was not a signatory to the DHEW policy. The simulation was approved by a Stanford University Human Subjects Research Review Committee, the Stanford Psychology Department, and the Group Effectiveness Branch of the Office of Naval Research. In 1973 the American Psychological Association concluded that all existing ethical guidelines had been followed.⁸

Congress passed the National Research Act (PL 93-348) in July 1974, officially giving DHEW the authority to establish regulations in this area. The law specifically limited the scope of the regulations to biomedical and behavioural research. This was followed by the 1979 Belmont Report⁹ which established the ethical principles of respect for persons, beneficence, and justice and provided guidelines for research involving human subjects.

Like Stanley Milgram’s Obedience to Authority,¹⁰ Zimbardo’s Stanford Prison Experiment found that systemic and group forces can influence an individual’s behaviour to violate a social norm or commit a crime, often without her or his personal awareness at the time. These studies, then, implant fear or anxiety and, like a cautionary tale, warn us of danger and use the Holocaust or Abu Ghraib prison as real-life examples of what might happen to good people who find themselves in an evil place. When Zimbardo first spoke publicly about the Stanford Prison Experiment, Milgram told him, “I would soon be diffusing some of the critical heat off him regarding the ethics of such ‘dark side of human nature’ research.”¹¹

Human behaviour is almost always a function of the interaction of person and social context, and Zimbardo’s prison environment encouraged several participants and groups to intentionally or unintentionally take on specific roles and behaviours. Rather than focusing on a person-centred analysis which starts with the dispositional factors of individuals to explain their behaviour, this paper will examine group dynamics and the situational factors in which individual participants are embedded. It will review how the experiment was conceived and developed, describe the then existing ethics review process, and focus on how participants attempted to get out of the experiment or developed coping mechanisms and behaviours to endure their painful or difficult situation.

Zimbardo’s Review Application

Zimbardo’s experiment grew out of a class project. In the spring of 1971, he asked students in his Social Psychology in Action class to investigate the changes an individual undergoes in the process of adapting to a new environment and role. Options included seniors entering retirement homes, people joining cults, and prisoners and guards socializing into their roles. Several students ran a mock prison in their dormitory over a weekend. While Zimbardo was aware this had taken place, he did not know what had happened until the students gave their in-class report.

The report revealed participants experienced intense feelings of anger, frustration, shame, and confusion about their behaviour during the project. Zimbardo held a debriefing with all of them. He realized that only a random assignment to the roles of guards and prisoners could separate dispositional factors, that is, behaviour-based primarily on participants’ attitudes and personalities, from situational factors, that is, behaviour resulting from the institutional rules and role expectations in a new environment.¹²

Zimbardo had previously studied deindividuation, a concept developed by Festinger et al.,¹³ that refers to the psychological state of group members who lose their individualism because they are treated uniformly within their group. Deindividuated persons may become unrestrained, and the group may generate antisocial acts. Zimbardo had found that participants who were deindividuated were more likely to inflict pain on others than those who felt more individuated.¹⁴

While a fellow at the Center for Advanced Study in Behavioral Science at Stanford, Zimbardo had several contracts with the Office of Naval Research. The US Navy and Marine Corps were concerned about conflicts between guards and prisoners in naval prisons. Zimbardo proposed to demonstrate that prison conditions were not the result of the type of individuals working and incarcerated in them, but rather emerged from the prison environment, rules, and role expectations. If successful, he hoped to help the Navy develop training programs to eliminate conditions which elicit counter-productive conflict.¹⁵ The Office of Naval Research agreed to a no-cost extension to pay for the prison experiment. 

Zimbardo completed his application for “Role Playing in a Simulated Prison” to the Stanford University Human Subjects Research Review Committee on July 31, 1971. Within two weeks the review committee approved the application, Zimbardo set up the mock prison, and the experiment began on Sunday, August 15^th. It was an exploratory study to document as completely as possible the emergence of a prison environment with respect to roles and behaviour. This meant that his application did not have research hypotheses, but rather identified a few basic parameters for the simulation and then proposed to observe and document how it evolved.

The Stanford Review Committee had many of the same concerns that would be addressed in regulations first published by the Department of Health Education and Welfare in 1974 and subsequently incorporated in the 1991 Federal Policy for the Protection of Human Subjects, known as the Common Rule (45 CFR 46)¹⁶ adopted by 15 federal departments and agencies including the Department of Defense. Part C §46.306(a)(2)(ii) requires Institutional Review Board (IRB) approval for a study of prisons as institutional structures or of prisoners as incarcerated persons, provided that the study presents no more than minimal risk and no more than an inconvenience to the subjects.

On his application, Zimbardo wrote: “There may be some potential emotional stress created by this temporary loss of freedom among the prisoners and by the need for surveillance and control among the guards. We will maintain careful observation of both groups, with professional staff living in [rooms] adjacent to the experimental ‘prison cells.’ Subjects will be notified of the potential stressfulness of the experience and, during a preliminary interview, will be encouraged not to participate if they are at all anxious.”¹⁷ In addition, the project would require physical activity, including exercise, cleaning rooms and halls, and number counts. 

Zimbardo answered “No” to the questions “Is deception to be used at any point?” and “Does deception affect the informed consent?” The application then asked if the participants would be submitted to humiliation, harassment, irritation, or public or private embarrassment. Zimbardo answered “Yes,” the prisoners’ privacy will be minimized, they will have to wear prison uniforms, and have to follow orders and rules.

Zimbardo also indicated that a background questionnaire and personality inventory would probably be used in the selection of participants to exclude extreme responders and deviants since he wanted the guards and prisoners to be as homogeneous and “normal” as possible. With respect to confidentiality, Zimbardo promised that all information would be coded, with a master list available only to the principal investigators, and that release forms for any film footage would be obtained from each participant before such films were shown.

Ads were placed in The Stanford Daily (the student-run independent newspaper) and The Palo Alto Times seeking male college students for a psychological study of prison life. Those who responded were sent an information sheet.¹⁸ They were to be paid $15 a day and would be randomly assigned to play the roles of either prisoners or guards for the duration of the study. Those selected would remain in the study from five days to two weeks, depending on the prisoner’s ‘sentence,’ and the guards would continue based on their work ‘effectiveness.’ Food and accommodations would meet minimal standard nutrition, health, and sanitation requirements. Medical and psychiatric facilities would be accessible should any of the participants desire or require such services. Participants would agree to have their behaviour observed, to be interviewed, and take psychological tests and that films of the study could be taken and shown, assuming the content had scientific value.

The information sheet, which was not the informed consent form, stated that “It is obviously essential that no prisoner can leave once jailed.”¹⁸ Zimbardo’s justification was presumably that, like real prisoners, those in the Stanford simulation ought not to have been given an easy way out. The concept of bail did not come up until Wednesday. The guards were told they must report for their eight-hour shifts promptly and that failure to fulfil the contract would result in partial loss of salary accumulated.

Finally, the applicants were told that two of the problems to be studied were (a) the development of norms which govern behaviour in a novel situation, and (b) the differential perceptions of “the prison experience” from participants who are initially comparable but arbitrarily assigned to play different roles. 

Respondents completed a questionnaire and were interviewed in depth by two psychology graduate students. The final 24 applicants, who ranged from 18 to 24, were judged to be the most physically and mentally stable, most mature, and least involved in antisocial behaviours. The participants did not know each other, save for two brothers who became guards but were assigned to different shifts. The applicants were randomly assigned to be either prisoners or guards, although none of the volunteers wanted to be a guard. The nine initial prisoners were then randomly assigned to one of three cells.

Creating a Total Institution

Zimbardo constructed a mock prison out of a hallway, several offices, and a closet in the basement of the psychology building at Stanford. The closet would become solitary confinement or Hole for disobedient prisoners. The prisoners’ uniforms would consist of a smock-like tan muslin dress with numbers on front and back, a woman’s nylon stocking to cover the participants’ hair (a substitute for head shaving), and clogs. He obtained guard uniforms at the local army surplus store. The guards wore reflective sunglasses and had no name tags. He ordered food for the prisoners from the student union, set up videotaping facilities, and bugged the prisoners’ cells. He also contacted Stanford University’s health, legal, fire, and police departments.¹⁹

Zimbardo based his experimental set-up in part on what he’d learned about prisons. One of the students in the spring social psychology class had invited ex-convict Carlo Prescott to talk to the class about his experiences in San Quentin State Prison.²⁰ Prescott described prisoners having bags placed over their heads, inmates being bound together with chains, and buckets being used in place of toilets in cells.²¹ Under Prescott’s mentoring, Zimbardo was able to bring a kind of situational savvy to his experiment.²² Prescott would become the chair of the parole board.

Sociologist Erving Goffman²³ wrote that total institutions are those in which the daily activities of members, such as eating, sleeping, working, and recreation, are collectively regimented, scheduled, and carried out in the immediate presence of many others. Everyone is treated alike and expected to follow a system of explicit formal rules enforced by the staff. Examples of total institutions were prisons, concentration and prisoner of war camps, mental hospitals, military barracks, and monasteries. 

To make his simulation as realistic as possible, Zimbardo created a total institution de novo with a rigid schedule, explicit rules, and strict obedience to authority. In some ways, the start-up of the Stanford County Jail resembled a prisoner of war camp where captured soldiers and military police were brought together for the first time, as compared to the start-up of a prison, where many inmates and guards are transferred in and have already been socialized into a prison environment. Zimbardo held an orientation meeting with the guards, where he discussed the purpose of the experiment, gave them their assignments, and suggested means of keeping the prisoners under control without using physical punishment. Zimbardo explained that he wanted to understand the psychological barriers that prisons create between people. 

The guards were told that the prisoners knew that the guards could not physically abuse the prisoners in any way. The guards, however, could act arbitrarily. Once during each of the three shifts, the guards could line up the prisoners for a “count” using their prison ID number to establish that all prisoners were present and to test them on their knowledge of prison rules. Zimbardo informed the guards that one research question was, “What will the prisoners do to try to gain power, to gain some degree of individuality, to gain some freedom, to gain some privacy?”²⁴

The guards then met with “Warden” David Jaffe, the undergraduate student who had led the prison project in the social psychology class that spring. Jaffe and the guards reviewed the set of rules developed by the class and agreed on a list of 17 that the prisoners would memorize and follow. These rules covered regulation of daily activities, respect for property, privileges controlled by the guards and staff, respect for authority, and taboo or forbidden words.

In Goffman’s view, house rules are a formal set of prescriptions that lay out the main requirements of inmate conduct.²⁵ The rules governing daily activities at the Stanford County Jail. Rule 1 required that prisoners must remain silent during rest periods, after lights out, during meals, and whenever they were outside the prison Yard (the long, narrow basement corridor). They were only to eat at mealtimes, participate in all prison activities and keep their cells clean at all times. The prisoners were not to move, tamper with, deface or damage walls, ceilings, windows, doors, or other prison property; and were never to operate ceiling lights. The rules stated that prisoners had to address each other by number (and the guards would address them by number as well), and address the guards as “Mr Correctional Officer” and the warden as “Mr Chief Correctional Officer,” and, they had to stand whenever the warden, the prison superintendent, or any other visitor arrived and wait for orders to be seated or to resume activities. Rule 9 specified: “Prisoners must never refer to their condition as an experiment or simulation. They are imprisoned until paroled.”²⁶ Ironically, that which must not be said became the safe phrase to get out of prison.

Privileges controlled by the guards, warden and superintendent included the three supervised toilet visits, which were limited to five minutes, and smoking after meals. As in real-world jails, mail would be inspected and censored, visitations would be supervised and terminated by a guard, and failure to obey any of the rules could result in punishment. 

The prisoners were “arrested” at their homes on Sunday morning by the local police and taken to the Stanford County Jail. Upon arriving at the jail, each prisoner went through dehumanization procedures similar to those in some prisons and concentration camps. They were blindfolded, ordered to strip naked and sprayed with powder to delouse them. They then put on a smock-like dress but no underwear, and a cap made of a woman’s nylon stocking to cover the long hair of many of these prisoners.²⁷

When all the prisoners had arrived, the rules were read slowly and authoritatively by Guard Arnett. The warden told them that a copy of the rules would be posted in each cell and that they were expected to know them and recite them by number. The prisoners were then sent to their cells to memorize the rules. The prisoners were later instructed to sing the rules, and after many repetitions, Zimbardo noted that they had obviously learned them all. Within 24 hours of their incarceration, the prisoners had internalized the rules as part of their prison mindset. The behaviour of both prisoners and guards would be governed by these rules — the guards enforcing them and the prisoners breaking them.

Escalating Confrontations

Zimbardo’s previous research on deindividuation¹³ suggested that sleep deprivation and altered time schedules could lower the threshold of behaviour restraint. He set up three eight-hour shifts for the guards: day from 10 a.m. to 6 p.m., night from 6 p.m. to 2 a.m., and morning from 2 a.m. to 10 a.m. The prisoners did not have a good night sleep Sunday night. They were awakened around 2:30 a.m. Monday by the incoming morning shift for the “count” and then allowed less than four hours of sleep prior to an early morning wakeup call at 6:10 a.m. By that time, they were frustrated with how the guards were treating them.

Before breakfast, the prisoners were tested on the rules and required to do morning exercises. 819 refused to do sit-ups as commanded and was put in the Hole. The guards then conducted a bed inspection with the prisoners standing by their beds. One of the guards told 8612 that his bed was a mess and ripped off the blanket and sheets. 8612 lunged at the guard who called for reinforcements and they put 8612 in the Hole with 819 where the two remained through breakfast.²⁸

After breakfast 5704, who had been denied his after-dinner smoke the night before, convinced his two cellmates (7258 and 3401) to do something to protest their living conditions. 3401 suggested that they push their beds against the door of cell #1, cover the door opening with blankets, and shut off the lights—a clear violation of Rule 6, which forbade operating the ceiling lights. When the guards could not break into cell #1, they rushed into cell #2, which, in Zimbardo’s opinion, held the top of the line troublemakers. One guard grabbed a big fire extinguisher and sprayed it into cell #2. The guards grabbed the three cots and hauled them out into the hall corridor that served as the prison Yard. A struggle ensued, with much pushing, shoving and shouting. 819 screamed, “No, no, no! This is an experiment. Leave me alone!” His cellmate 8612 added, “A fucking simulation. It’s a fucking simulated experiment. It’s no prison.”²⁹

During the struggle, 8612 ended up naked and shouted that in real prisons they don’t take your clothes and bed away. This produced a sudden silence. But then another prisoner tells him that they do. Zimbardo noted that in order for the simulation to work, everyone had to agree to act as if it were a prison and impose communal self-censorship.³⁰ After that everyone avoided mentioning the obvious truth, that it was, in fact, a simulation that was getting out of control.

Following the Monday rebellion, Zimbardo created a grievance committee, with the prisoners electing its members. The committee met with Zimbardo and demanded less physical and verbal abuse and harassment, complained about the food and requested books and more than one visiting night. They asked that medicines be administered, wanted their glasses returned, and requested religious services.

Zimbardo then contacted a Catholic priest to play prison chaplain. As it turned out, Father McDermott had actually served as one in Washington, DC. He agreed to talk with some of the prisoners and then give Zimbardo his honest evaluation of how realistic the prison experience seemed.7 On Wednesday, the priest met with eight of the nine prisoners one at a time, but Zimbardo pulled up a chair and sat next to the priest so he could listen in. Father McDermott asked some of the prisoners whether they have seen the public defender. In response, 7258 gives him his mother’s name and phone number, explaining that his cousin is in the local public defender’s office and might be able to bail him out.

Unexpectedly, 5848, an undemonstrative good prisoner, told the priest that what was going on was an experiment which, in his estimation, was getting out of control. Father McDermott thought that all the prisoners he had met with were naïve and didn’t know anything about prison or what it is for. In his opinion, the study was working as a real prison, and what he saw was the typical first-offender syndrome. The prisoners exhibited confusion, irritability, rage, depression, and over-emotionalization. He assured Zimbardo that such reactions would change after a week or so.³²

Get Out of Jail Free

In the board game Monopoly, players can use the “Get out of Jail Free” card and not have to roll doubles or pay the $50 fine. But the rules for getting out of the Stanford County Jail were ambiguous at best, and some were improvised, including asking prisoners if they were willing to forfeit their pay to be paroled. Exactly what did those prisoners who wanted to leave have to do in order to get out of the study?

The Stanford Prison Experiment informed consent form³³ stated that participants would only be released from the simulation for reasons of health deemed adequate by the medical advisors to the research project or for other reasons deemed appropriate by Zimbardo. Once in the study, the prisoners learned Rule 9 that prisoners must never refer to their condition as an experiment or simulation. To utter the words that must not be spoken would threaten the illusion vital to the success of the study. Participants had to maintain the fiction that supported their assigned roles as prisoners or guards. Rule 9 was broken several times. The first was on Monday morning, in the midst of the “rebellion” when 819 and 8612 said it was an experiment and simulation. The second was on Wednesday when 5486 told the priest that it was an experiment which was getting out of control.

Goffman³⁴ identified four types of adaptation to total institutions: (a) withdrawal, where the inmate drastically curtails his involvement and interactions with both staff and fellow inmates; (b) intransigence or rebellion, where the inmate intentionally refuses to cooperate with the staff in any way; (c) colonization, where the inmate carves out a personal niche within the institution to make life more bearable, and may be accused by fellow inmates of never having it so good, and (d) conversion, where the inmate identifies with and assists the staff in dealing with fellow inmates.

Four of the original nine prisoners were in jail until the end of the experiment on Friday morning. Two were “good” prisoners, or colonizers, from the start, while the other two converted and became “good” prisoners after the Monday rebellion. This suggests that those who adjusted most rapidly to the experimental conditions of prison life were able to minimize the degradation and harassment meted out by the guards.

The remaining five original prisoners used different tactics to get out of jail early. The first to be released was the rebel in 8612. As described above, during the struggle on Monday he screamed it was a simulated experiment. That afternoon, he reported feeling sick and met with Warden Jaffe and complained about the arbitrary and sadistic behaviour of the guards. The warden assured him that he would see to it that the guards eased up. The warden went to Zimbardo and reported that 8612 was really distraught, wanted out, and insisted on seeing him.

When 8612 told Zimbardo that he couldn’t take it anymore, Zimbardo replied that he was the most rebellious, insubordinate prisoner. But 8612 responded that Zimbardo had violated the contract and that he hadn’t expected to be treated so poorly. Zimbardo made 8612 an offer: the guards would not hassle him, and he could stay and earn his money if he cooperated and provided information—in essence, if he became an informer or snitch. 8612 appeared to capitulate by replying, “Well, all right.” Zimbardo later contended that “At that moment, if he had insisted on being released, I would have allowed him that option.”³⁵

Not to be deterred, 8162 managed to unlock his cell door and almost escaped around dinnertime. He was captured and allowed to eat dinner by himself. But he suddenly got up from dinner, raced across the room and ripped down the black scrim hiding the video camera. The guards caught him and put him back in the Hole. He then began complaining about headaches. When he was released from solitary at lights out, 8612 screamed that he was burning up inside. He had a second meeting with the warden and insisted that he wanted to get out and that he couldn’t stand another night. He also asked if he had the right to ask for a lawyer.

When Craig Haney, one of the graduate assistants, returned from his late dinner on Monday, he met with 8612 to determine if he should be released immediately based on severe emotional distress. Haney realized that an early release could compromise the study design and that this was an unexpected turn of events. But 8612 was obviously disturbed. Haney, therefore, decided on his own to release him on ethical/humanitarian considerations over experimental ones.³⁶ Haney contacted 8612’s girlfriend who quickly came and picked him up. Haney told them that if the distress continued, he should visit Student Health on Tuesday morning where arrangements had been made with staff to deal with such reactions. He was released on Monday night approximately 36 hours after the experiment began.

Prisoner 819 was the next to get out of jail. His tactics, like 8612, followed what Goffman termed “rebellion.” He refused to cooperate and, as in the movies The Great Escape(1963) and Cool Hand Luke (1967), got put in the Hole time and was left alone. On Monday morning, after having his sleep interrupted for the 2 a.m. and 6 a.m. counts, he quit the exercises and refused to continue for which he is sent to the Hole. Later that morning, when the guards rushed into his cell and pulled the cots into the Yard, Prisoner 819 screamed, “This is an experiment! Leave me alone!”

When Prisoner 819 started to tell his visiting parents, brother, and sister on Tuesday evening about the Hole and his problems with the guards, one of the guards overheard him and stopped him from saying more. On Wednesday, Prisoner 819 locked himself in his cell and ripped his pillow apart. After another stay in the Hole, he reluctantly consented to meet with Father McDermott. At the meeting, he was not wearing his stocking cap and his hair was a mess. After telling the chaplain what he had been going through, Father McDermott told him that he needed to be less emotional. Later Zimbardo asked 819 if he wanted to be released immediately, but the prisoner insisted he was willing to continue and promised not to try any funny business.

However, when he heard guard Arnett having the other prisoners shout over and over“  Because of the bad things that Prisoner 819 did, your cells are a mess,³⁷ he began shaking. This is an instance where an authority figure orders or encourages subordinate group members to exert peer pressure on a deviant or cowardly member, who then agrees to conform and behave properly. He told Zimbardo that he still did not want to leave and had to go back in, but Zimbardo called him by his first name, told him it was just an experiment, and that it was time for him to go home.

The third was 1037 who joined the Monday morning rebellion by urging the occupants of cell #3 to barricade themselves in, but later withdrew and refused to come out of his cell. Three guards entered his cell, handcuffed his ankles and dragged him by his feet through the Yard and into the Hole. He was subsequently elected to the Grievance Committee which Zimbardo created on Monday after the rebellion. The committee met with Zimbardo, but 1037 was not convinced that any changes would be made.

1037’s mother wrote a letter to Zimbardo after visiting her son on Tuesday evening. She said her son looked haggard and had not been well sleeping because of the middle of the night “counts.” Her son was sorry he had volunteered, had gone through several moods, and was now resigned. Following the advice of the prison chaplain, FMcDermott, she indicated that she was going to secure legal counsel for her son.

Zimbardo considered the Stanford County Jail as one holding a group of adolescents in pretrial detention following their Sunday-morning arrests. Obviously, no trial date had yet been set for any of them and none of them had legal representation. However, after a full staff meeting, a parole process was quickly improvised.³⁸ Parole was not mentioned in the informed consent form, although Rule 9, which forbade saying experiment or simulation, mentioned the possibility of parole. The parole board consisted of Prescott as chair, Haney, a department secretary and a graduate student. The latter two had little prior knowledge of the experiment.

The research staff met and identified prisoners, including 1037, whom they considered eligible, and invited them to write formal requests explaining why they thought they deserved parole. They also asked the guards to prepare written reasons for denying parole to each of the four prisoners. However, the guards insisted that 416 not be granted such an opportunity because of his persistent violation of Rule 2 that prisoners must eat at mealtimes and only meal times. 416 was brought in to replace 8612 who was released on Monday night. He was horrified by what he saw and believed that quitting was impossible. 416 decided to go on a nonviolent Gandhi hunger strike.³⁹ When he refused to eat, the guards put him in the Hole for three hours, although the rules stated was one hour was the limit. He would remain in the experiment until it ended on Friday.

In his parole petition, 1037 stated he had rebelled, but that evening realized that he was unworthy of better treatment. Since then he did his best to cooperate and no longer cause problems, or, in Goffman’s typology, converted. During his hearing on Wednesday, 1037 told the parole board that he would consider parole even if it meant forfeiting his salary. On Thursday, Zimbardo told him that he would be paroled and would get full pay for the entire experiment after the study and final surveys were completed. His parents picked him up during visiting hours that evening.

The last of the original prisoners to be released was 4325. He had been a good prisoner and had been elected from cell #3 to the Grievance Committee. In Goffman’s terms, he adopted the role of the colonizer. After his appearance before the parole board, he was hopeful that he would be released soon, for the board had agreed that he should be the first of the four to be let out. But when he learned that 1037’s parents had come to pick him up, 4325 became depressed and “broke” as a result.⁴⁰ Zimbardo saw this and had him released.

Following up on his promise, Father McDermott called7258’s mother, who called her nephew, an attorney in the local public defender’s office, who, in turn, called Zimbardo. Zimbardo reluctantly agreed to schedule an official lawyer’s visit for the prisoners on Friday morning as one more realistic element in the study. The public defender was curious and sceptical about the whole situation. Zimbardo briefed him on the study, how serious it had become, and invited the lawyer to treat the matter exactly as if he were working with a group of real prisoners. 

Although Zimbardo had already decided to end the study, he raised the final curtain on Friday morning meetings between the public defender and several prisoners. The public defender told Prisoner 7258 (his cousin) that if he was willing to forfeit that his pay, then the contract would be null and void. Prisoner 7258 replied that he had told the parole board that he was willing to give up his compensation to get out of prison, but it did no good.

The public defender heard complaints from the other prisoners about physical harassment by guards and inadequate supervision by the senior staff.⁴¹ When he told the prisoners he would file a formal report on Monday and try to arrange their bail, his cousin begged to be bailed out immediately because the prisoners couldn’t take another week or even a weekend. The lawyer replied that he could help them but was powerless to do anything right then and there. This brought the remaining prisoners to a new low. As soon as the public defender left, Zimbardo told them that the study was in fact over and that they were free to leave.

Ethical Issue: Ability to Leave at Will

The primary research ethics issues of interest are the inability of the volunteers to readily exit the study and the contention that the experiment was continued too long as physical and psychological abuse of the prisoners escalated. On the question of quitting the experiment, Zimbardo⁴² noted that the student volunteers could have elected to quit at any time, but they had promised to do their best to last the full two weeks. The students assumed that they could leave whenever they chose not to continue. But this changed when 8612 told the prisoners after his meeting with Zimbardo that he would not be let out and that they couldn’t get out either. The other prisoners came to believe that if this rebellious leader couldn’t get out, they, too, were helplessly stuck in jail for the duration of the experiment.⁴³

This is an example of how a subordinate group legitimizes the authority and rules of an organization and supports those with power over them. Max Weber⁴⁴ wrote that voluntary submission is a basic criterion of authority, and Peter Blau⁴⁵ noted that authority rests on the acceptance of social norms that a collectivity of subordinates can enforce on its individual members. That is, peer pressure obliges individuals to comply. The failure of the prisoner rebellion and the capitulation of its leader sent a powerful message to the other study participants that resistance was futile, and it was in their best interest to recognize the legitimacy of the prison guards.

When a prisoner did in fact attempt to say, “I quit,” his plea was either ignored or met with delaying tactics. Years later, Prisoner 8612 told the media that the prisoners had only been pretending to go crazy.⁴⁶ It appeared that those participants who wanted out of the study either made a conscious choice to playact medical/psychological symptoms, adopt role models like Gandhi, Steve McQueen in The Great Escape or Paul Newman in Cool Hand Luke, or genuinely succumbed to the stressful conditions. Certainly, for 8612, who was stripped naked and complained about the arbitrary and sadistic behaviour of the guards, the severity of harm had reached a point on Monday evening that Hanley released him from the experiment.

Zimbardo’s study was a simulation rather than a traditional experiment with testable hypotheses and a control group, although participants were randomly assigned to the roles of guard and prisoner. He thought of it as a demonstration of a phenomenon, like Milgram’s obedience study.8A simulation is designed and carried out when one wants to understand the natural dynamics and relationships in a social situation, but it is not feasible to conduct the study in a real-world setting. Good simulations are interactive, and researchers may adjust the ongoing scenario. Zimbardo made the simulation more realistic by bringing in a prison chaplain, creating a parole board and finally allowing the prisoners to meet with a public defender. 

In his application to the Stanford Human Subjects Research Review Committee, Zimbardo had explicitly stated that those playing prisoners would be discouraged from quitting. At the parole board hearings, 7258, admitted he had attempted to escape with a former cellmate but that he had since pretty much followed all the rules. When parole board chair Prescott, asked what he did to get in here, 7258 truthfully answered, “Nothing, sir, but to sign up for an experiment.” On hearing the words that must not be spoken, Prescott quickly responded, “So wise guy, you think this is just an experiment.” But a few minutes later, Prescott noted, “You tell us you’d even be willing to forfeit compensation to get out of prison.”⁴⁸ Upon hearing this, Zimbardo had each of the four prisoners return and told Prescott to ask them if they would forfeit their pay if paroled.

Zimbardo thereby sent a message that only the parole board could give the participants permission to leave, and then only if they were willing to forfeit the money they had earned. Of the four, only 3401, who really needed the money, said he would continue rather than take the offer to leave immediately without pay. He also told the parole board that he had a rash that’s going to break out, but Prescott dismissed his concern. He subsequently developed what appeared to be a full-body rash, perhaps psychosomatically induced, and was released on medical grounds.⁴⁹

Many of the critiques of the Stanford Prison experiment fail to mention that five of the original nine prisoners, either through playacting, rule-breaking, forfeiting pay, or experiencing extreme stress, were able to be released early. Although approximately half were able to get out, this still supports the current requirement that participants should be free to withdraw from a study at any time without penalty or loss of compensation.

Ethical Concern: Lack of Oversight

The second ethical concern was that the experiment continued too long which allowed the physical and psychological abuse of the prisoners to escalate. Haney⁵⁰ reported that after the experiment had ended, an analysis of the data revealed that, in fact, total guard aggression showed a daily escalation even after most prisoners had ceased resisting. This not only happened on camera in the Yard but also when an individual guard was alone with a single prisoner out of range of the cameras on their way to the toilet.

Zimbardo acknowledged that he did not provide sufficient oversight and training for the guards.⁵¹ He told them they could not physically abuse or torture the prisoners but could instil a sense of fear and treat them arbitrarily. Zimbardo and the research team did not respond to prisoner complaints by reprimanding those guards who physically abused the prisoners or kept them in the Hole for over an hour. He admitted “failing to intervene more often, which had thereby given them implicit permission to go to the extremes they did. They might have avoided their abuses had they had better top-down surveillance.”⁵² Apparently, Zimbardo wanted to see how far the guards would go and did not release any prisoners unless they threatened the integrity of the simulation, became extremely uncooperative, or were clearly suffering from extreme stress. 

Given Zimbardo’s dual role of Principle Investigator and Prison Superintendent, and his determination to see the simulation play itself out, no one was in an official position of authority to force him to end the study early. The person who persuaded him was Christine Maslach, Zimbardo’s former teaching assistant who had collaborated with him on other research projects. She recently completed her PhD at Stanford and was about to become an assistant professor in the psychology department at the University of California, Berkeley. She was also romantically involved with Zimbardo — a year after the prison study ended, they married. 

Zimbardo had not told her how the study was evolving because she was part of a team that was scheduled to do a thorough evaluation of staff, prisoners, and guards on Friday, near the end of the first full week of the experiment. On Thursday, she unexpectedly was asked to serve on the parole board as a replacement for Haney, who left to deal with a family emergency. That evening, she came down to the prison to meet Zimbardo for a late dinner. On her way in, Maslach stopped to talk with one of the guards, whom she described as being “very pleasant, polite and friendly.”⁵³ Later on, she was told by one of the research staff that she should take a look at the meanest, toughest guard, who had just come on duty. When she looked through the observation point, she was stunned to discover it was the same guard she had chatted with earlier, but now he moved and talked differently, cursing with a southern accent, and swinging his club. This was guard Hellman, who was nicknamed John Wayne.

When Zimbardo told her to look at the prisoners chained together by the ankles and with paper bags over their heads, she said that she had already seen them and averted her gaze. She began to cry and told Zimbardo that she was leaving and to forget about dinner. He ran after her and told her no one else had reacted as she just had.⁵⁴ She replied that she didn’t care if everyone in the world thought that what was going on was okay, she thought it was simply wrong, and that Zimbardo was personally responsible for the prisoners’ suffering. The argument was intense, unlike any they had had before. He acknowledged his responsibility and decided he would end the experiment on Friday morning. 

Someone, who was not part of the research team, should have provided oversight and made decisions about when participants could leave the experiment. Zimbardo⁵⁵ later admitted that he had a serious conflict of interest as Principal Investigator and Prison Superintendent. He thought that if someone else had been acting as superintendent, he would have seen the light and ended the experiment earlier. But, more to the point, he also acknowledged that the study required oversight and that someone should have had authority over him to end it.

The Common Rule §46.111(a)(6) states that “when appropriate, the research plan makes adequate provision for monitoring the data collected to ensure the safety of subjects.” These Data Safety and Monitoring Plans (DSMP) may assign the responsibility to the principal investigator, the sponsoring entity, or a data-safety monitoring committee or board. If the research is not a clinical trial but does involve greater-than-minimal risks to participants, IRB approval of a data-safety monitoring plan is required which may include an observational study monitoring board.⁵⁶

Because the Stanford Prison study was scheduled to last only two weeks and the data would not be available until much later, safety monitoring would depend on ongoing observation. This could have been carried out if the IRB appointed an individual to serve as both safety monitor and Prison Superintendent, thereby eliminating Zimbardo’s conflict of interest. As Prison Superintendent, the safety monitor would have attended research staff meetings, been able to observe from the behind the one-way screen, had the authority to reprimand the guards for excessive abuse, and report adverse events directly to the IRB. The safety monitor could have met with Zimbardo to discuss the possibility of releasing specific study participants, most likely those that, in fact, did leave early. Maslach’s source of authority was her personal relationship with Zimbardo: that is, referent power, as opposed to the organizational and regulatory status of an IRB safety monitor who would have held legitimate power over Zimbardo as Principal Investigator.⁵⁷

To his credit, Zimbardo had scheduled an evaluation of the study on Friday of the first full week and ended the study that morning. He later held group and individual debriefing sessions and regularly collected post-experimental questionnaires over the next months and years. He concluded that the suffering the participants experienced did not extend beyond the confines of the basement prison.⁵⁸

Zimbardo let the simulation evolve until he was finally persuaded to end it. While the original class simulation could be considered exempt from current IRB review since it was designed to enable students to develop experiments to investigate how individuals adapt to a new environment and role, the standalone Stanford Prison Experiment would require IRB review since Zimbardo intended to publish and publicize the results.

Discussion

At the onset of the experiment, Zimbardo had informed the participants that he wanted to study the development of norms that govern behaviour in novel situations and how people who are initially comparable but arbitrarily assigned to play specific roles would differentially perceive the same situation. The Stanford Prison Experiment demonstrated that the behaviours of guards and prisoners were not a result of their personality types but rather of values, norms, and behaviour generated by the prison environment, rules, and role expectations. Individuals will vary in their role adaptations to specific situations. The focus, then, should not be on individual psychological characteristics or moral fortitude, but rather on group dynamics and role socialization which transfer norms, values, beliefs, and behaviours to incoming group members.

The simulation gave Zimbardo valuable insights into prison behaviour, riots, and abuses, specifically the retaking of the Attica New York Prison after the September 1971 riot and abuses by the 372nd Military Police Company of the US Army Reserves at the Abu Ghraib detention centre in Iraq in 2003-2004. Ashley Rubin, who studies penal change, thought that Zimbardo got a lot right. Although rare, Attica Prison and Abu Ghraib are real-world examples of “how prison can unleash the worst of human nature with terrible consequences.”⁵⁹ The Stanford Prison Experiment showed how quickly a prison-like an environment can morally degenerate when guards are left to their own devices.

For the most part, Zimbardo followed the ethical procedures in place at the time. The consent form stated that participation in the research project will involve a loss of privacy, that release from participation would only be for reasons for health determined by the medical advisers or by Zimbardo, and that participants were expected to follow directions from staff members of the project or from other participants. They have not informed the type of prison clothing they would be wearing, or the degree of harassment and humiliation they might experience. Because it was a simulation, Zimbardo and his research team could not anticipate many of the events that emerged over the five days. He did not have a safety monitor who might have spotted the escalating violence, but relied on and followed the advice of his graduate research assistants, ex-convict Prescott, and Father McDermott who believed the simulation was realistic and the students slowly becoming socialized to their new roles as prisoners.

Lessons Learned

Fyodor Dostoyevsky, who spent who spent four years of hard labour in a Siberian prison camp, wrote that “The degree of civilization in a society can be judged by entering its prisons.”⁶⁰ To be sure, Zimbardo’s prison simulation was not normal paradigm science in which formal theories generate hypotheses and variables that can be tested using established experimental methods. Zimbardo’s controversial methods and findings triggered an ongoing debate on human nature, the social sciences, and research ethics that has lasted almost 50 years.

The Stanford Prison Experiment showed how good people can do bad things in a prison environment if oversight and safeguards are not in place. It justified the need for social science research ethics and practices that protect participants and, when necessary, closely monitor studies in which the IRB determines that the probability and the magnitude of possible harm or discomfort anticipated is greater than those ordinarily encountered in daily life or during the performance of routine physical or psychological examination or tests. These are lessons that should be included in textbooks with a discussion of what happened, how the participants behaved and reacted to the situation, and the implications for real-world prisons and detention centers.

Acknowledgments

No Funding source

Conflict of Interest

None

References

1. Bankert, E. Cohen, J.M, Cooper J.A, Davis, B, D., and Hicks, L. (2009).History & Ethics Collaborative Institutional Training Initiative.
2. Zimbardo, P. (2008). The Lucifer Effect: Understanding How Good People Turn Evil. NY Random House. p 185.
3. Le Texier, T. (2018). Historie d’un mensonge Enquête sur l’expérience de Stanford. Paris: Éditions La Décourverte. ISBN: 978-2-3552-120-0.
4. Blum, B., (2018). The Lifespan of a Lie: The most famous psychology study of all time was a sham. Why can’t we escape the Stanford Prison Experiment? (June 7, 2018). Available at https://medium.com/s/trustissues/the-lifespan-of-a-lie-d869212b1f62.
5. Toppo, G. (2018). Time to Dismiss the Stanford Prison Experiment? Inside Higher Ed. (June 20, 2018). Available at https://www.insidehighered.com/news/2018/06/20/new-stanford-prison-experiment-revelations-question-findings.
6. Resnick, B. (2018). Philip Zimbardo defends the Stanford Prison Experiment, his most famous work. Vox,(Jun 28, 2018). Available at: https://www.vox.com/science-and-health/2018/6/28/17509470/stanford-prison-experiment-zimbardo-interview.
7. Department of Health Education and Welfare. (1971). Institutional guide to DHEW policy on protection of human subjects (NIH 72-102) Washington, D.C. 20402. Available at: https://archive.org/stream/institutionalgui00nati/institutionalgui00nati_djvu.txt.
8. Zimbardo, P. (2008). Ibid. pp 235-236.
9. National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research. (1978). The Belmont report: Ethical principles and guidelines for the protection of human subjects of research. Bethesda, MD: The Commission.
10. Milgram, S. (1974). Obedience to Authority: An Experimental View. New York: Harper and Row.
11. Zimbardo, P.G, Maslach, C., and Haney, C. (2000). Reflections on the Stanford Prison Experiment: Genesis, Transformations, p 3. Available at:zimbardo.com/downloads/Publications.pdf.
12. Zimbardo, P.G.(2008). Ibid, pp 32, 494, 495.
13. Festinger, L., Pepitone, A., and Newcombe, T. (1952). Some Consequences of Deindividuation in a Group. Journal of Abnormal and Social Psychology 47:382–89.
    CrossRef
14. Zimbardo, P. (1969). The Human Choice: Individuation, Reason, and Order versus Deindividuation, Impulse, and Chaos.In W. J. Arnold, & D. Levine, D (Eds). Nebraska Symposium on Motivation (vol 17:237-307). Lincoln, NB: University of Nebraska Press,
15. Haney, C., Banks, C., & Zimbardo, P. (1973) A Study of Prisoners and Guards in a Simulated Prison. Naval Research Reviews.p1. Available at: www.zimbardo.com/downloads/1973%20A%20Study%20of%20Prisoners%20and%20Guards,%20Naval%20Research%20Reviews.pdf.
16. Federal Policy for the Protection of Human Subjects 45 C.F.R. 46 §306. (1991).
17. Zimbardo, P. G. 1971a). Application for Institutional Approval of Research Involving Human Subjects. Available at: http://www.prisonexp.org/pdf/humansubjects.pdf.
18. Zimbardo, P.G. (1971b). Participant information sheet. Available at: http://www.prisonexp.org/pdf/geninfo.pdf.
19. Zimbardo, P.G. (2008). Ibid. pp. 28, 40.
20. Zimbardo, P.G, Maslach, C., and Haney, C. (2000). Ibid. p 23.
21. Prescott, C. (2005). The Lie of the Stanford Prison Experiment, The Stanford Daily. (April 28, 2005).
22. Zimbardo, P.G. (2008). Ibid. p 68.
23. Goffman, E. (1961). Asylums. Essays on the Social Situation of Mental Patients and Other Inmates. New York: Doubleday Anchor. p. 5-6.
24. Zimbardo, P.G. (2008). Ibid. p 55.
25. Goffman, E. (1961.). Ibid. p.48.
26. Zimbardo, P.G. (2008). Ibid. p 44-45.
27. Zimbardo, P.G. (2008). Ibid. p 40.
28. Zimbardo, P.G. (2008). Ibid. p58-59.
29. Zimbardo, P.G. (2008). Ibid. p 61.
30. Zimbardo, P.G. (2008). Ibid. p 496.
31. Zimbardo, P.G. (2008). Ibid. p 100-104.
32. Zimbardo, P.G. (2008). Ibid. p 104-106.
33. Zimbardo, P.G. (1971c) Informed Consent Form. Available at: http://pdf.prisonexp.org/consent.pdf.
34. Goffman, E. (1961). Ibid. p.61-63.
35. Zimbardo, P.G. (2008). Ibid. p 71.
36. Zimbardo, P.G. (2008). Ibid. p 77.
37. Zimbardo, P.G. (2008). Ibid. p 105-107.
38. Zimbardo, P.G. (2008). Ibid. p 130-131.
39. Zimbardo, P.G. (2008). Ibid. pp.115,116, 131.
40. Zimbardo, P.G. (2008). Ibid. p 161-162.
41. Zimbardo, P.G. (2008). Ibid. p. 177-178.
42. Zimbardo, P.G. (2008). Ibid. p. 222.
43. Zimbardo, P.G. (2008). Ibid. p.70-71.
44. Weber, M. (1947). The Theory of Social and Economic Organization, NY: Oxford University Press. p.324.
45. Blau, P. (1964). Exchange and Power in Social Life. New York: John Wiley and Sons. p. 200.
46. Zimbardo, P.G. (2008). Ibid. p.361.
47. Zimbardo, P.G. (2008). Ibid. p.197.
48. Zimbardo, P.G. (2008). Ibid. p.139-140.
49. Zimbardo, P.G. (2008). Ibid. p.135, 143.
50. Haney, C., Banks, C., & Zimbardo, P., (1973). Ibid. p.12.
51. Zimbardo, P.G. (2008). Ibid. pp. 4, 55, 166.
52. Zimbardo, P.G. (2008). Ibid. p. 183.
53. Zimbardo, P.G. (2008). Ibid. p. 169.
54. Zimbardo, P.G. (2008). Ibid. p. 170.
55. Zimbardo, P.G. (2008). Ibid. p. 234-235.
56. National Heart Lung and Blood Institute. (2008). Data and Safety Monitoring Policy. Available at: http://www.nhlbi.nih.gov/funding/policies/dsmpolicy.htm.
57. French, J.R.P., & Raven, B. (1959). The Bases of Social Power. In D. Cartwright (Ed.), Studies in Social Power, (pp:150-167). Ann Arbor: MI: Institute for Social Research.
58. Zimbardo, P. G. (1973). On the ethics of intervention in human psychological research: With special reference to the Stanford prison experiment. Cognition 2(2):243-256.
    CrossRef
59. Rubin, A. (2018). What a widely attacked experiment got right on the harmful effects of prisons. The Conversation. (October 2, 2018). Available at: http://theconversation.com/what-a-widely-attacked-experiment-got-right-on-the-harmful-effects-of-prisons-103967.
60. Dostoyevsky, F. (1862). The House of the Dead. C. Garnett (trans). New York: Macmillan Co., 1950. In F. R. Shapiro (Ed.). (2006). The Yale Book of Quotations(p. 210). New Haven, CT: Yale University Press.