Early Uses of Solitary Confinement (1790–1835)

Solitary confinement was first authorized in the United States in the 1790s, when states began building their first prisons. The most significant innovation of these early prisons was sentencing criminals to incarceration as punishment. Previously, during the colonial period, incarceration had not been penal, but administrative, used for the accused awaiting trial, the condemned awaiting their (corporal or capital) punishment, or for those who had been punished but still owed court-imposed fines and jail fees. Penal incarceration was still quite similar to administrative confinement, at least in the sense that prisoners spent much of their time congregated together in large rooms. Distinct from colonial jails, however, these early prisons kept their convicted criminals separate from the untried, witnesses awaiting trial, debtors, vagrants, and others who also remained congregated together (Rubin forthcoming b).

Separation became a major organizing theme of these prisons, as prisoners of different genders and types (recidivist or novice, convicted criminal or unfree citizen) were separated from each other to prevent the spread of disease as well as the mutual moral contamination previously wrought by congregating prisoners in jails (Meranze Reference Meranze1996). Incarceration, combined with discipline-generating hard labor, was designed to create submissive citizens who would respect authority and abstain from crime. In this regime, reformers considered solitary confinement an indispensable tool for prisoner reformation, but not one intended for all prisoners—only for the worst offenders. Solitary offered the advantages of separation—in this case, preventing the worst offenders from influencing the less experienced offenders. However, it went further: as a terrifying experience, it also served as an additional threat to encourage following the prison's rules. Moreover, with no distractions, it would allow prisoners time for reflection and reformation (see, e.g., Lownes [Reference Lownes1793] 1799; Turnbull Reference Turnbull1797).

While solitary confinement was often discussed in the penal reform literature and was included in almost every statute authorizing the early prisons (Rubin Reference Rubin2016, in prep), the place of solitary confinement was more limited in practice. The country's model prison, Philadelphia's Walnut Street Prison, had only sixteen solitary cells—a disproportionately small number for the prison's hundred (and later, more) prisoners (Teeters Reference Teeters1955, 19). Likewise, judges tended to avoid sentencing convicted criminals to any period in solitary confinement (Sellin Reference Sellin1953, 329). For both these reasons, solitary confinement was only used for short periods to punish in-prison infractions; due to the scarcity of cells and the frequent rule violations, however, solitary was often “the last, not the first, resort of discipline,” while punishment at the hands of the prison's administrators was the more common response (Meranze Reference Meranze1996, 196). Similar design flaws and limitations in practice followed in every prison that adopted the Walnut Street model. Ultimately, a flurry of fires, escapes, and riots at Walnut Street, as well as at the nation's other nascent prisons, unavoidably proved that the early experiment had failed to achieve either reformers' stated disciplinary goals or administrators' basic goals of order maintenance (Rubin in prep).

In the 1810s and 1820s, prison officials and reformers alike sought to refine the physical structures and legal rules governing prisons and turned to solitary confinement to achieve their goals better. Referring to past failures, penal reformers and legislators sought to improve their approach and to create a more ordered prison that relied more systematically on solitary confinement—not only for a minority of the most “hardened” prisoners, but for all prisoners. Some reformers, who were increasingly involved with prison administration, argued that solitary confinement would be perceived as more terrifying than the incarceration that (they believed) was no longer a deterrent. Additionally, they believed that solitary confinement would prevent the chaos within the “schools of vice” that the early prisons had become; locked in solitary cells, prisoners would no longer be able to conspire or mutually to corrupt each other's morals. Ultimately, many reformers believed that solitary confinement would enable proper reflection and reformation (Meranze Reference Meranze1996; McLennan Reference McLennan2008; Rubin in prep).

Pennsylvania's reformersFootnote ⁴ were among solitary confinement's strongest supporters; they believed it was the ingredient missing from their first experiment with incarceration at Walnut Street. An 1818 statute authorized solitary confinement for all prisoners with no distractions (Pennsylvania 1818). When Pennsylvania's new prison, Western State Penitentiary, opened just outside of Pittsburgh in 1826, its architectural design proved problematic. The cells were too small and poorly ventilated for long-term solitary confinement; prisoners frequently had to be released from their cells to preserve their health (Barnes Reference Barnes1921; Doll Reference Doll1957).

New York's reformers also expressed frustration with their first prison. Newgate, which was modeled on Walnut Street, had failed to achieve either its disciplinary or order maintenance goals. As in Pennsylvania, New York legislators sought to implement solitary confinement more systematically:

The experiment, begun on Christmas Day 1821 at the new Auburn State Prison, was a disaster. The cells were too small for prisoners to exercise, and their muscles atrophied. The intensive reliance on solitary confinement was even more devastating on prisoners' mental health. Auburn's Warden, Gershom Powers, reported that “one [prisoner] was so desperate that he sprang from his cell, when the door was opened, and threw himself from the gallery upon the pavement. … Another beat and mangled his head against the walls of his cell until he destroyed one of his eyes” (Powers Reference Powers1826, 36). The surviving prisoners—whose “health and constitutions … had become alarmingly impaired”—received pardons, and the experiment was officially discontinued in 1823 (Powers Reference Powers1826, 36). A similar disaster occurred at Maine State Prison several years later (Lewis Reference Lewis1922, 147–48; Rubin in prep).

These early episodes with total solitary confinement at Western State Penitentiary, Auburn State Prison, and Maine State Prison became the first set of “historical echoes” that would continue to haunt penal experimentation and innovation. These episodes inspired widespread social critique (an external response to internal failures of implementation in the early prison institutions) and de-legitimized solitary confinement. By the middle of the 1820s, penal reformers and prison administrators alike expressed intense opposition to total solitary confinement (e.g., BPDS [1826] 1827). Most centrally, they concluded that the practice was cruel and inhumane and dangerous to mental and physical health. Moreover, reformers and other commentators believed solitary confinement was too expensive: prisoner labor was increasingly understood to be a central ingredient for prisoner reformation and to offset the cost of prison maintenance, but labor seemed impossible in solitary.

Still, prison officials experimented with different variations on the solitary confinement theme—attempting to legitimize new practices by distinguishing them from old practices, while consistently seeking to maintain order and control. New York invented a new system of confinement in which prisoners silently performed factory-style labor and experienced solitary confinement only at night. Additionally, military discipline—including lockstep marching, ranks for prison staff, and whipping—helped to maintain order. Most states would copy New York's new Auburn System and continue to oppose solitary confinement for anything more than nighttime confinement (Rubin Reference Rubin2015a), but Pennsylvania moved in a different direction to pursue discipline and order maintenance.

Eastern State Penitentiary (1829–1913)

Eastern State Penitentiary opened outside of Philadelphia in late 1829, eight years after its initial statutory authorization. Eastern was Pennsylvania's latest attempt to resurrect the original goals behind its earlier Walnut Street Prison by instituting more order and security while ensuring prisoner reformation. In response to national sentiment, as well as the overt failure at Western State Penitentiary, Pennsylvania reformers and legislators sought to correct their past mistakes. With a different architectural design that included larger, better ventilated cells, provision for in-cell labor, and weekly visits from prison staff and penal reformers, they sought to mitigate the possibility of death, insanity, inefficacy, and impracticality already experienced. Eastern State Penitentiary was the result of these efforts.

The theory underlying Eastern State Penitentiary, articulated in a new statute (Pennsylvania 1829), was referred to as the Pennsylvania System. Under this system, at least theoretically, prisoners lived—worked, slept, ate, prayed—in their solitary cells. Each ground-floor cell was attached to a small, walled-in yard into which prisoners were released at various times. Prisoners would spend the vast majority of their sentence within their cells, passing most of their day (except the Sabbath) performing some form of craft (e.g., weaving, shoemaking, carpentry) or unskilled labor (e.g., oakum picking or bobbin winding).

A combination of prison personnel and official prison visitors would meet with the prisoners at least once a week to mentor, minister, or instruct them. Overseers (guards trained in a craft) would instruct the prisoners in how to perform their assigned work task. A moral instructor (chaplain) would visit the prisoners and offer sermons on Sundays, while a physician would check their physical health weekly. Each week, two members of the prison's five-man Board of Inspectors (the prison's main governing body) and the prison's warden would separately visit the prisoners to mentor them and to inquire into any complaints. Finally, members of the local penal reform society, the Philadelphia Society for Alleviating the Miseries of Public Prisons, would visit the prisoners bimonthly or monthly. The moral instructor, inspectors, warden, and penal reformers would teach prisoners how to read and write, as well as try to convince prisoners to change their ways, while collecting information about what had led a prisoner to commit his (or her) crime in the first place. Apart from these officially authorized visitors, prisoners would never see another person (including family), let alone another prisoner.

This approach to solitary confinement under the Pennsylvania System was different from the kind of total solitary confinement attempted at Western and used briefly at Auburn and Maine in the early 1820s. However, that earlier approach did not disappear completely, as illustrated by administrators' response to refractory prisoners: Eastern's prisoners who refused to work, attempted to speak with other prisoners, attacked their guards, tried to escape, or broke other prison rules were removed to “punishment cells” or “dark cells.” Dark cells were regular cells whose window and skylight were covered with cloth, blocking out the light; prisoners in the dark cells were sustained on a diet of bread and water and given no work or books to distract them from their solitude. A recalcitrant prisoner remained in the dark cell for a day or two (initially) or up to a week (or, occasionally, two), typically until he or she apologized for his or her wrongdoing and pledged good behavior in the future. Although reminiscent of Auburn's early experiment with total solitary confinement, this form of solitary was exceptional—of short duration and only used after administrators warned misbehaving prisoners to obey (Rubin Reference Rubin2017a, in prep).

Eastern's prison officials thus used two versions of solitary confinement for administrative control. The first, central to the Pennsylvania System, was justified primarily as a means of enabling prisoner reformation by allowing reflection and preventing mutual contamination across prisoners. This form of solitary confinement had emerged out of the need to prevent disorder, but was central to reformers' and reformers-turned-administrators' commitment to prisoner rehabilitation. The second form of solitary confinement, however, was more severe (a short-term version of the earlier efforts at total solitary) and punitive: it was a means of regaining control over and demanding deference from unruly prisoners. Moreover, both versions were distinct from the central mode of nightly solitary used elsewhere, under the Auburn System, indicating the variety of forms of solitary confinement used in this single historical period.

Throughout this period, Eastern remained a minority model in the United States. While nearly every state had built at least one prison by the Civil War, only Pennsylvania retained its Pennsylvania System by that time. (New Jersey and Rhode Island had both adopted the system, but quickly abandoned it.) Instead, most prisons followed New York's Auburn System, which restricted solitary confinement to the nighttime and congregated prisoners during the day in large, factory-like rooms for silent labor. Penal reformers in Boston and New York promoted this Auburn System nationwide, arguing that it was more economical, but also that the Pennsylvania System was cruel and inhumane, dangerous to prisoners' mental and physical health, and ultimately ineffective (Rubin Reference Rubin2015a). Foreign visitors who toured America's prisons echoed these sentiments, emphasizing particularly their concern over the Pennsylvania System's severity and the dangerous effects of long-term isolation on individual prisoners (de Beaumont and de Tocqueville Reference DeBeaumont and deTocqueville1833; Dickens [Reference Dickens and Ingham1842] 2000).

Publicly, Eastern's administrators defended their use of solitary confinement, emphasizing its difference from the kind of fatal solitary used earlier at Auburn. Administrators called their Pennsylvania System the “separate system” and referred to their primary use of solitary confinement as “separate confinement.” They reminded their critics that prisoners received approved visitors—the prison staff and the local penal reform society—so the claim that a prisoner at Eastern “never looks upon a human countenance, or hears a human voice” (Dickens [Reference Dickens and Ingham1842] 2000, 113) was untrue. Eastern's administrators further sought to discredit their opponents as profit-mongers who cruelly whipped their prisoners, while presenting themselves as “benevolent,” “humanitarian” gentlemen concerned for their prisoners' welfare. The Pennsylvania System, they argued, was likewise more humane and beneficial for prisoners than the competitor Auburn System (Rubin Reference Rubin2013, in prep).

In practice, however, solitary confinement under the Pennsylvania System was never as severe as its critics—or its supporters—described. Prison administrators and prisoners often violated the Pennsylvania System. Prison officials removed between 10 and 25 percent of the prison population from their cells to perform work around the prison. Although prisoners were supposed to avoid contact with other prisoners and nonprison personnel (e.g., construction workers), records indicate multiple incidents of prisoners mingling; at least two prisoners even carried on an intense romance for two months (Rubin Reference Rubin2013, Reference Rubin2015b). Additionally, prisoners constructed numerous inventive ways of communicating with one another from within their cells, from tapping on pipes, to standing on looms to reach through a skylight and yell to other prisoners, to carving notes on the bobbins passed from winders to spinners (Rubin Reference Rubin2015b, Reference Rubin2017b). Prison administrators also occasionally double-celled prisoners, particularly when one seemed to become mentally ill from the isolation. For most prisoners, however, the Pennsylvania System remained mostly intact until the late 1870s (Rubin Reference Rubin2013, in prep).

Driven by a combination of overcrowding and administrative discretion, in the 1870s, Eastern's administrators quietly abandoned the Pennsylvania System. In its place, they instituted the “Individual Treatment System,” which retained parts of the Pennsylvania System's ideology without its core focus on solitary confinement. For years, the administrators had claimed that, through solitary confinement, they could better observe a prisoner's reformation and customize treatment to the individual's needs. When overcrowding made solitary confinement impossible, the administrators continued to emphasize the customization of treatment but de-emphasized the earlier claim that solitary enabled the special opportunities for observation. During the decades to follow, prisoners were increasingly housed with cellmates, allowed to work in congregate workshops, and released together to exercise in the prison's yards. The administrators maintained their emphasis on the importance of reformation, preventing prisoner communication (to the extent possible within the space constraints imposed by overcrowding), and mentorship, education, and religion. However, solitary confinement was no longer a core component of the formal regime (Rubin Reference Rubin2013, in prep).Footnote ⁵

During the first century of experimentation with solitary confinement, then, cycles of critique, reform, and legitimizing retrenchment were visible. Prison administrators experienced the frustration of internal failures, including overcrowding, imperfect implementation of the “separate system,” and mental health deterioration. Some of these failures were publicly critiqued by the likes of Charles Dickens; others were not. In response to failure and critique, prison administrators and reformers continued to experiment with and refine the practice of solitary confinement. In both New York and Pennsylvania, two of the country's then-most penologically innovative states, legislators and reformers in the early 1820s first sought to correct the incomplete implementation of solitary confinement in their earlier prisons by implementing new programs of total solitary confinement. In both cases, however, they ultimately sought a relaxation of rigid conditions in response to extreme failures like death and insanity. In sum, prison administrators at both Auburn and Eastern (and beyond) responded to critiques by repeatedly refining the quality and quantity of their isolation practices.

Prison administrators also sought to legitimize new solitary confinement practices by distinguishing them from older solitary confinement practices. For instance, in an attempt to distinguish new practices from old, Eastern administrators denied that prisoners were in solitary confinement, insisting first on the term “separate confinement” and later on the term “Individual Treatment System.” They argued consistently for both the practical value (as a tool of control and order maintenance) and the ideological value (as a tool of rehabilitation) of solitary confinement, or “separation.” Over time, solitary confinement continued its trajectory toward becoming a marginal practice in every prison, but it remained ever present, as a technology of control of last, if not first, resort.

Subaltern Models of Solitary Confinement (1867–1978)

Just as the Pennsylvania System was starting to fall apart at Eastern, the debate over solitary confinement cooled. Penal norms had settled: the only advocates of long-term, round-the-clock solitary of the kind used in the Pennsylvania System were limited to Pennsylvania. Moreover, most commentators were content with the Auburn System's use of “isolation at night.” Even so, some commentators who opposed long-term solitary confinement did acknowledge another, limited place for solitary confinement in prison regimes. In their 1867 survey of prisons, Auburn advocates Enoch C. Wines and Theodore W. Dwight argued that the “proper place” of the Pennsylvania System was in jails, for short-term confinement, and in the early portion of one's longer confinement under the Auburn System (Reference Wines and Dwight1867, 55–56). Wines continued to hold and promote this view at national and international congresses on prisons in the coming decades (Rubin in prep).

In 1890, after his passing, Wines's son, Frederick, presented a paper at the national prison congress at Cincinnati affirming his father's earlier view: “We all, I think, have the separate system in mind as the ideal for short-term prisoners, especially for the accused, while waiting [sic] trial” (National Prison Association 1890, 97). This use of pretrial solitary confinement was widely deployed in Europe during the twentieth century, especially in Scandinavian countries, extending well into the twenty-first century (Smith Reference Smith2009).

When it came to convicted criminals' long-term confinement, however, Frederick Wines endorsed a hybrid system that relied on “individual treatment,” a concept that was increasingly in vogue beyond Pennsylvania (National Prison Association 1890, 97; see also Garland Reference Garland1985). Long-term solitary confinement, in the late nineteenth and early twentieth centuries, remained an unacceptable practice, at least in the United States. Though it lost favor in the United States over the course of the nineteenth century, the Pennsylvania model remained popular in Europe—in Britain, France, Germany, Belgium, Holland, Norway, Sweden, and Denmark—and even in other parts of the world, including Chile and New Zealand (Smith Reference Smith2009).

Meanwhile, in the United States, even the Auburn System of old was coming under fire, mainly for its reliance on contract-based labor systems (McLennan Reference McLennan2008). However, as southern penal reformers began to tout their own innovations as progressive, they used northern prisons as a foil (Lichtenstein Reference Lichtenstein1996). An Alabama Inspector of Convicts, reporting on his state's convict-lease system, reported: “The convicts are allowed to talk to one another and to visitors. We have not yet reached that point in refined cruelty where the convict is kept in a solitary cell and not allowed to raise his eyes from his work or to speak to anyone but his keeper, until his mind, from constant communion with its own thoughts, is liable to become a wreck” (National Prison Association 1890, 116). More generally, however, solitary was such a minor practice that it was hardly mentioned outside of discussions of prisoners' punishment for in-prison misconduct.

Reflecting contemporary thought in 1890, the US Supreme Court dismissed solitary confinement as a barbaric and destructive practice no longer used in most of the United States. Recollecting the early experience with solitary confinement, the Court noted:

The Court overturned James Medley's death sentence, finding that his preexecution solitary confinement, under a policy Colorado implemented after Medley committed his crime, constituted a new punishment so severe it violated the ex post facto clause of the constitution (Reiter Reference Reiter2012, 79). Much as Justice Kennedy's concurrence in Davis v. Ayala (2015) represented a judicial “bark” with little legal “bite,” the Medley Court's critique of solitary confinement appeared in dicta, not directly related to the holding of the case, and therefore not legally binding. Still, the decision in Medley represented the most marked legal rejection of solitary confinement in the nineteenth century. Over a period of sixty years, solitary confinement had been theorized as an ideal mode of confinement, implemented inconsistently, resisted by prisoners, refined by prison administrators and legislators, criticized by reformers and yet other prison administrators, and then, nominally, at least, abandoned.

The very fact that Colorado initiated the practice of keeping at least some prisoners in solitary confinement in the 1880s, however, reveals solitary's persistence. In 1891, just one year after the Medley decision, the US Supreme Court upheld the use of solitary confinement in a similar context: a New York facility where prisoners were awaiting execution (McElvaine v. Brush  1891). Over the next half-century, the US Supreme Court only mentioned solitary confinement two more times; each time the Court presumed both the severity and the constitutionality of the practice (United States v. Moreland  1922, 433, 449 [Brandeis, J., dissenting]; Chambers v. Florida  1940, 227, 237–38; Haney and Lynch Reference Haney and Lynch1997, 540). Meanwhile, Pennsylvania prison administrators systematically relaxed the use of solitary confinement over the nineteenth century, as the Auburn model of congregate labor gained popularity. Most of the burgeoning US prison population in the early twentieth century continued to live and work in congregate settings, although a few troublemakers were perpetually excluded in solitary confinement.

Prison records from infamous facilities across the United States also confirm that solitary confinement remained in use as punishment. For instance, records from Alcatraz, a federal penitentiary operating from 1934 to 1963, describe the “Spanish dungeons” used to isolate troublesome prisoners (Odier Reference Odier1982, 117). Of course, solitary confinement had also been used in this way throughout the nineteenth century, such as in Eastern's “dark cells” for refractory prisoners. Like those earlier dark cells, the dungeons at Alcatraz functioned primarily as a tool of control and order maintenance. Alcatraz's warden even attempted to impose an Auburn-style model of congregate, but silent labor. This attempt at reviving the solitary-like conditions of early US prisons, however, faced immediate and sharp criticism, much as Auburn and Eastern had faced in the nineteenth century. In 1938, the Saturday Evening Post published an exposé alleging that more than one dozen prisoners had gone insane under these restrictive, silent conditions (Haney and Lynch Reference Haney and Lynch1997, 488).

As the twentieth century wore on, prisons lost their experimental sheen and became an integral part of US democracy. Solitary confinement, on the other hand, continued to inspire criticism, and critics from the Supreme Court to the Saturday Evening Post continued to presume that the practice of solitary confinement, unlike incarceration, was far from integral to American democracy.

Still, prison administrators continued to use solitary confinement, whether in the form of a few dark cells and dungeons, or in the form of more systematic attempts to impose silence and other restrictions. These internal institutional practices continued to generate external critiques. In the late 1960s and early 1970s, in particular, prisoners began organizing for and asserting their rights to better conditions of confinement (see, e.g., Reiter Reference Reiter2016; Thompson Reference Thompson2016), while courts became increasingly receptive to prisoner litigation (Feeley and Rubin Reference Feeley and Rubin2000).

Due to a series of legal changes granting individual plaintiffs broader rights to sue state officials, litigation over conditions of confinement in prisons flourished across the United States by the late 1960s (Reiter Reference Reiter2012). Horrifying prison conditions—including practices that resembled slavery and constituted torture—became a focal point of social critique. By the early 1970s, more than thirty state prison systems faced challenges to conditions of confinement in at least one facility, and sometimes the entire state system (Feeley and Rubin Reference Feeley and Rubin2000). Every major case in this litigation movement for better prison conditions addressed the question of conditions in solitary confinement, or isolation. In one of the few of these cases to reach the US Supreme Court, Hutto v. Finney (1978), the Court noted that even short-term uses of solitary confinement in Arkansas prisons were “punitive” and “serve[] no rehabilitative purpose” (688). The Hutto Court, like many lower courts considering similar challenges to solitary confinement in state prisons across the United States, upheld caps of fifteen to thirty days on durations of solitary confinement, seeking to avoid lengthier stays (Reiter Reference Reiter2012).

Over the course of the twentieth century, then, experiments with solitary confinement continued, and these experiments continued to inspire critique. Although there were a few media exposés of solitary, like the Saturday Evening Post piece about the attempt at imposing a regime of silence at Alcatraz, the primary critics in the twentieth century were not public reformers or legislators or political theorists, but courts. The critics were new, but the responses were old.

First, prison officials defended themselves against these critiques, resisting court orders to reform conditions in solitary confinement and to cap the amount of time prisoners could spend there. In the Hutto case out of Arkansas, for instance, prison officials resisted over the course of thirteen years, appealing these orders all the way to the US Supreme Court. In particular, prison officials objected to any total ban on the use of solitary confinement and to any requirement that solitary confinement be “rehabilitative” (Hutto 1978, n8). In a way, this objection to the requirement of rehabilitation represented an implicit distinction between solitary confinement in Arkansas in the 1960s and solitary confinement in Pennsylvania in the 1830s: history had taught prison officials that solitary confinement was unlikely to be rehabilitative, so they resisted any requirement to make it so. Instead, they insisted that it was necessary to manage and control prisoners.Footnote ⁶

Second, in addition to defending solitary confinement, prison officials continued to impose the practice—despite the Medley Court's condemnation in 1890, the Saturday Evening Post's condemnation in 1938, and even the Hutto Court's condemnation in 1978 (stemming from a case initiated in 1965). Prison administrators, like the warden at Alcatraz, and the guards in Arkansas whose solitary confinement policies were challenged in the case of Hutto v. Finney, experimented with old regimes of silence and isolation in dark cells, modifying in response to criticism, but often continuing, their experiments with technologies of power.

Experiments with Longer-Term Solitary Confinement and the Modern Supermax (1974–Present)

Even after the barrage of litigation in the 1970s and 1980s, the use of solitary confinement expanded, albeit in new “experimental” iterations. In the 1970s and 1980s, prison systems across the United States began testing out new technologies of longer-term, more systematic isolation. These experiments took three forms: behavior modification through sensory deprivation, semi-permanent “lockdowns,” and supermaximum security prisons. As with earlier forms of solitary confinement use, prison officials faced critique, but they defended themselves against these critiques, often with reference to historical echoes of earlier regimes.

Behavior modification programs were common across US prisons in the 1970s (see, e.g., Cummins Reference Cummins1994), but the Control and Rehabilitation Effort in the federal system and Special Treatment and Rehabilitation Training (START) in Missouri particularly attracted attention—and critique. These programs experimented with techniques of brainwashing piloted in military training exercises: placing prisoners in extreme isolation, without reading materials or human contact, and then gradually relaxing the conditions of confinement as prisoners met behavioral benchmarks. In justifying the programs, prison officials focused on the dual goals of order maintenance and discipline (with at least an implication of the potential for rehabilitation). For example, prison officials defended Missouri's START program by claiming, “the primary objective … was an attempt to work with these offenders to control their behavior so that they could participate in regular institution programs” (Clonce v. Richardson  1974, 340). And prison officials distinguished the new behavioral modification regimes as novel, using “new techniques” that differentiated them from earlier (discredited) forms of solitary confinement. Such justifications both internalized lessons from earlier critiques—solitary confinement with no end or alternative goal had been abandoned, for instance—and sought to distinguish modern penal technologies from earlier penal technologies, like silent regimes or dark cells.

Nonetheless, the inherent abusiveness of these regimes was documented both in a journalist's book-length exposé about cruel prison conditions (Mitford Reference Mitford1974, 134–35) and in federal litigation (Adams v. Carlson  1973, 621–22; Clonce 1974, 345). In two federal cases, Adams and Clonce, courts ordered modifications to the harsh behavioral management regimes, and insisted that prisoners be granted at least minimal due process protections, such as notice of why they were being placed in the restrictive behavioral management programs, prior to being placed in them.

Still, experiments with isolation continued. Violent altercations between prisoners and staff, along with full-blown revolts in prisons across the United States in the 1970s—in California, New York, Illinois, and Massachusetts, to name just a few examples—precipitated long-term lockdowns following each riot (see Reiter Reference Reiter2016; Thompson Reference Thompson2016). During these lockdowns, prisoners were stuck in their cells all day every day for weeks, months, and then years at a time (Toussaint v. McCarthy  1984, 1397, 1410; Ward and Breed Reference Ward and Breed1984; Cummins Reference Cummins1994, 232; Bissonnette et al. Reference Bissonnette, Hamm, Dellelo and Rodman2008). More prisoner litigation challenged these long-term lockdown conditions (Reiter Reference Reiter2012). And again, prison officials justified the lockdowns as critical for maintenance of institutional order, especially for managing “those portions of the prison population composed of prisoners who are dangerous and violent” (Toussaint 1984, 1367). This time, instead of seeking to distinguish long-term lockdowns from earlier uses of solitary confinement, prison officials sought to distinguish 1980s prison culture from earlier prison cultures, pointing to increasing problems with gang leaders and gang affiliates threatening to control the prison system (1375), to justify more extended uses of isolation. Courts resisted these arguments, again ordering administrators to relax the conditions of confinement—to provide more time out of cell, more opportunities to visit with nonprisoners, and more procedural protections to govern who was placed on lockdown.

But instead of designing alternatives to solitary confinement, prison administrators worked with architects to design the first “supermax” prisons—technologically advanced facilities that institutionalized lockdown practices. In 1986, Arizona opened Cell Block 6 (CB 6) at its Eyman Complex, the first supermax facility in the United States (Lynch Reference Lynch2010). A supermax, or supermaximum security prison, is a unit, or an entire prison, designated for prisoners' solitary confinement. Prisoners spend an average of twenty-two to twenty-three hours per day locked in their cells, in conditions that keep any kind of human contact at an absolute minimum. The first supermax units in Arizona, opened in 1986, were stark. Each cell, made of poured concrete, had one hard concrete ledge for a bed, and an angular cube-and-shelf combination for a desk. These “fixtures” were solid concrete, continuous with the cell walls. There were no windows.Footnote ⁷ The perforated steel doors were equipped with a small locking hatch at about waist level, allowing for both a food tray to slide in and for prisoners to reach out their hands for cuffing before being removed from the cell—a rare occurrence. Prison officers in a central, Panopticon-like control booth could look out over multiple blocks of cells at one time, and operate the doors through a computer system, requiring no face-to-face human contact to release a prisoner from his cell for his daily one hour of exercise in a “yard”—usually another concrete enclosure with a skylight or some other access to natural light. In 1989, California opened the 1,056-bed Pelican Bay State Prison Security Housing Unit, explicitly modeled on the Arizona Security Management Unit. And in 1994, the federal prison system opened the Administrative Maximum, modeled on the Arizona SMU and the Pelican Bay SHU (Reiter Reference Reiter2016).

By the late 1990s, nearly every state had a supermax (Riveland Reference Riveland1999). In many ways, supermaxes are the archetypal punishment in the late-modern era. They emerged across the United States roughly simultaneously with the largest period of prison building (see Eason Reference Eason2017, 6). They epitomize the “warehouse” style of these newer prisons: geometric concrete boxes with no landscaping, combined with layers of barbed wire fencing, located in isolated rural settings (Feeley and Simon Reference Feeley and Simon1992, 460; Irwin Reference Irwin2005). They are designed for the so-called worst of the worst, the problem prisoners who are too dangerous to interact with or live among prison personnel or other prisoners. Supermax prisoners fit the cultural-political imagination of prisoners as monstrous (Kennedy Reference Kennedy2000) and are the physical embodiment of the metaphorical toxic waste that Feeley and Simon (Reference Feeley and Simon1992) argue is targeted by modern incapacitation policies. Supermaxes thus represent the latest experiment with solitary confinement as a technology of power, deployed by prison administrators for labeling, threatening, and controlling prisoners' bodies.

Not only do supermaxes redeploy solitary confinement as a tool of institutional control, experimenting with new structures and procedures to achieve the same outcome, but supermaxes, like each of the preceding iterations of solitary confinement, have been the subject of both public and legal critique. As with every prior experiment with isolation in the United States, the new supermaxes in Arizona and California quickly inspired public critique, including a 60 Minutes exposé of the conditions at California's Pelican Bay in 1993 (Reiter Reference Reiter2016). As with many critiques of experiments with solitary confinement, at least since prisoners gained the right to make legal challenges to the conditions of their confinement in the 1960s, litigation also ensued, most notably in the case of Madrid v. Gomez. Almost as soon as the Pelican Bay Security Housing Unit opened, prisoners there started complaining about the harsh conditions of confinement. The critiques began within the prison system, internally, but they soon inspired external oversight. A federal judge in northern California quickly certified a class of prisoners and ultimately held a three-month trial in 1993 evaluating the constitutionality of conditions in the Pelican Bay SHU.

As in earlier challenges to solitary confinement, prison officials quibbled over the labels applied to the facilities. The supermax at Pelican Bay was not called solitary confinement, or a separate system, or a dark cell, but instead a “Security Housing Unit” or “SHU.” When critics said the SHU imposed harsh conditions of solitary confinement, prison administrators objected, saying “[i]t's not Draconian, it's Spartan” (Reiter Reference Reiter2016, 115) and pointed out that prisoners have modern privileges in isolation, like the ability to take correspondence courses and watch television (Dolan Reference Dolan2011). Such attempts to legitimize a new penal technology by distinguishing it from an old one, however, ignore the fact that prisoners at Eastern learned to read and write and had access to books—nineteenth-century equivalents of correspondence courses and television.

Nonetheless, prison officials' work to distinguish supermaxes from earlier, discredited forms of solitary confinement, legitimizing the supermax as new and different, were ultimately fairly successful. The final holding in the Madrid case: prisoners needed better mental and physical health care, and mentally ill prisoners should never have been sent to isolation in the first place, but the overall practice of keeping prisoners in solitary confinement indefinitely was constitutional (Reiter Reference Reiter2016). Solitary confinement, in its newest iteration, would persist. As with the Pennsylvania model, the Auburn model, and other early US experiments with solitary confinement, supermaxes would also have global appeal, appearing in countries as diverse as Ireland, Denmark, and Turkey, to name just a few examples (O'Hearn Reference O'Hearn2013; Ross Reference Ross2013; Reiter Reference Reiter and Sullivan2014b).

The relationship between supermaxes and their many historical antecedents—whether the Eastern or Auburn penitentiaries, dark cells or silent dining halls, lockdowns or behavioral management programs—is contested, with scholars pointing to multiple sources and explanations for supermaxes. On the one hand, scholars explain supermaxes as a brand-new technology of power, a product of the penal politics and policies of this latest penal generation—the period after the 1970s that generated mass incarceration and the whole host of policies and practices that constitute the modern penal assemblage (Shalev Reference Shalev2009; Ross Reference Ross2013; Simon Reference Simon2014). On the other hand, scholars explain supermaxes as an old technology of power repackaged in a modern architectural form, but originating at Alcatraz in the early twentieth century, or even at Eastern State Penitentiary in the early nineteenth century (Mears and Reisig Reference Mears and Reisig2006, 34; Mears and Bales Reference Mears and Bales2009, 1132; Reference Mears and Bales2010, 545). Supermaxes, however, can be better understood as a product of these contested origin stories: a reinvention and reinterpretation of solitary confinement, with multiple eras of critique integrated into the institution, as encrusted layers of both justification and practice.

Administrative Goals versus Social Change

One of the basic contributions of the punishment and society literature has been to illustrate the interconnectedness between punishment and society. Rather than a simple response to crime, punishment is a function of society—its religion, its economy, its politics, its culture, its cohesion, and, especially, its diversity. This lesson has been repeatedly illustrated by scholars for more than a century (Durkheim [Reference Durkheim1893] 1984; Rusche and Kirchheimer Reference Rusche and Kirchheimer1939), although it was most popularly propagated by Michel Foucault (Reference Foucault1977) and David Garland (Reference Garland1985, Reference Garland1990, Reference Garland2001), and the many scholars in dialogue with them (Beckett Reference Beckett1997; Wacquant Reference Wacquant2002; Zimring Reference Zimring2003; Simon Reference Simon2007; Barker Reference Barker2009). By emphasizing the important role society plays in shaping punishment, however, these macro-level, context-focused accounts downplay factors internal to the penal system that might also contribute to changes in penal policy.

This study, by contrast, has illustrated the importance of these internal factors. First, we have illustrated the way in which prison administrators, driven by their perception of internal needs like prisoner control, have offered a counterpoint to changing norms in the larger society. Administrators' continued use of solitary confinement has frequently proceeded even after the technology has been publicly discredited. Second, we have also highlighted the importance of historical legacies. We have demonstrated that pivotal events within penal institutions—such as Auburn's fatal experiment with solitary confinement in New York; prisoner-initiated complaints about harsh conditions of confinement as in the Hutto case out of Arkansas; or moments of significant unrest, including revolts in the 1970s California, New York, and Massachusetts—have significant consequences for the shape of punishment to come. These historical legacies, and prison administrators' reactions to them, in turn become part of the penal context, or field, in which prison administrators operate (see Bourdieu and Wacqant Reference Bourdieu and Wacquant1992; Page Reference Page2011, 10–14). They act as “feedback effects” (Schoenfeld Reference Schoenfeld2010, 737) that structure the range of choices—scripts and actions—open to actors within the criminal justice system. Such important internal factors are overlooked in analyses that only focus on external or macro-level changes at the level of politics or economics.

Implications for Solitary Confinement and Other Extreme Punishments

By examining the persistence of a particular penal technology, solitary confinement, we not only illustrate an important influence on penal policy and practice, but also highlight one significant contradiction of contemporary penal rhetoric. Modern penal officials insist that they are not using solitary confinement, seeking to differentiate themselves from widely condemned practices abandoned at Eastern, and dismissed as barbaric by the Supreme Court in Medley in 1890 (albeit in dicta). In other words, modern prison officials want to distance themselves from something that was perceived as inordinately harsh and to argue they are imposing something more civilized. In fact, what exists today is, in many ways, calculated to be significantly harsher: modern supermax prisoners do not have the private yards attached to their cells, the expectation of being provided with work, or the weekly visits from staff and volunteers, which Eastern State Penitentiary prisoners did have, for instance.

Better understanding the similarities and differences between these two innovations thus illustrates precisely what is and is not new about supermaxes and their use of solitary confinement. In turn, understanding the ways in which solitary confinement use has persisted over time reveals two mechanisms of this persistence: administrative control and legitimization in relation to historical legacies.

Recent calls for solitary confinement reform or abolition, especially those made by Supreme Court Justice Kennedy in the 2015 case of Davis v. Ayala, and echoed by his colleague Justice Breyer in the 2017 case of Ruiz v. Texas, tend to assume that litigation and critical attention might result in significant changes to solitary confinement practices. Such assumptions, however, reveal a lack of understanding of both the persistence of solitary confinement and the mechanisms of this persistence. The history of solitary confinement use in US prisons suggests that significant changes to the practice have been both initiated and coordinated internally, by prison administrators, rather than externally by critics.

In light of this history, the total abolition of solitary confinement, as imagined by Justices Kennedy or Breyer, as well as by other current reformers such as the national “Stop Solitary” campaign currently coordinated by the American Civil Liberties Union National Prison Project, seems unlikely, absent broader changes to the structure of incarceration, and especially the control that prison administrators wield. Indeed, Reiter (forthcoming) argues that a more global perspective on solitary confinement reveals that solitary confinement has been persistent and resistant to reform in Europe, as well as in the United States.

While our analysis has focused on solitary confinement, and the implications of historical practices for current reforms, this analysis also has broader application to other “extreme punishments” (Reiter and Koenig Reference Reiter and Koenig2015, xi), as well as to other penal and legal technologies. Like long-term solitary confinement, extreme punishments such as capital punishment and deportation have undergone similar processes of superficial transformation while fundamentally persisting (Reiter and Coutin Reference Reiter and Coutin2017). In each case, the penal or legal technology is always in the background, used at society's margins in ways that are useful to administrators or, more generally, people in power. The administrative and historical mechanisms through which such practices persist deserve more attention and exploration.