At 2.23 million, the inhabitants of US prisons and jails constitute a population larger than Houston’s and just behind Chicago’s — a larger population than sixteen American states.[1] Nearly one quarter of the earth’s incarcerated human beings, and close to one third of the earth’s incarcerated women, are drawn from the less than 5% of people residing inside the US. [Incarceration Nation] [Kajstura 2018]. A notoriously race-driven system, poor people of color live these numbers [Western 2006][Sakala 2014], while over 6,000-US confinement facilities, as well as the police, courtrooms and prisons stand virtual lightyears away from moneyed white living rooms where they appear as subjects of uncounted television dramas (and regular humor) [Brown 2009]. As long as incarcerated people remain either data points without faces or voices, or the criminal habitués of fearful popular imagination, they are easily dismissed as 2.23M units of a type, rather than being perceived as the nation’s fourth largest city of fathers and mothers, brothers and sisters, grandparents and grandchildren, wives and husbands.

The aim of this essay is to describe one instance of digital humanities practices working not only in contradistinction from “retro-humanism” [Bianco 2013] — the replication of traditional canons on line — but as an irruptive, discursive intervention into the dis-equalizing social practices from which retro-humanism might be counted as one distraction. It describes an instance in which DH’s signature ability to close spaces between people and archives seeks to close the “social distance” between those populations that determine criminal punishment and those upon whom punishments land, a distance that has been shown to mark nations where penal practice is most harsh [Christie 1994, 23–24]. The essay will describe the organic emergence and growth of The American Prison Writing Archive, an archive of first-person non-fiction essays by incarcerated people that continues a 231-year history of American literary resistance to a dis-equalizing institution that over the past forty years has grown into the unprecedented American “penal state” [Wacquant 2009]. While the regular calls for DH to engage with inequality, social injustice, and cultural critique assume DH as an extant practice that needs to reassess its premises and methods [Liu 2013] [Earhart 2013] [Clement 2016] [Gallon 2016] [Bailey et al. 2016] , this paper describes an instance of marginalized voices forcing a break out of prisons and into DH, and of their driving the metadata behind the search facets availed to archive visitors. Roopika Risam notes that “Without attention to the omissions that exist within digital humanities scholarship, the field risks replicating the exclusions of a dominant culture that already relegates difference to its margins” [Risam 2015]. Here I describe a project that grew not from a decision to address such DH omissions but that evolved into the digital in the effort simply to keep up with the insistent will of those omitted by state confinement to press their case for inclusion in, and thus expansion of public discussion of US carceral practice. The APWA evolved from the labor of actors among the very populations that have historically been marginalized by criminal justice involvement — a testimonial action that, like other witness literatures, demands its own manner of reading. Amid debates surrounding traditional close and computation-enabled distant reading [Ramsey 2016] [Moretti 2013], this essay documents the inception and rapid growth of a literary corpus that demands what I will describe as mid-range or cellular reading — a position that not only has a well-established history as a means of intervention in dis-equalizing social practices, but that is the apt response to what incarcerated writers do: write for 2.23 million others while making singular statements. The essay suggests that DH's definitive work of closing material distances is the work to which incarcerated and otherwise confined writers have always aspired: to broach walls (material, social, cultural, of prisons, plantations, death camps) and thus to expose life inside to readers outside, creating public interfaces in resistance to the prison’s founding purpose in legal quarantine from the public sphere’s discursive life [Meranze 1996, 12] [Rush 1787]. My claim is that whenever DH engages directly with social injustice — let alone when those whose inequality has been socially and legally sanctioned demand a place on digital platforms — DH is best assessed by what it does or aspires to enable among social agents; and such enabling cannot come solely through text analysis, text mining, or close reading, but must come, if it seeks to affect change in practice, through a mid-range reading that receives texts and affects readers as these texts are intended by documentary witnesses: as singular statements that assume and offer context-dependent representative testimony for a collective body.[2] When witness-based, text-driven DH projects venture into social justice work, they step onto a well-warn path behind those who have helped to disseminate slave, Holocaust, and other witness literatures, whose reception incurs concrete social effects. These effects have in the past been achieved without the benefits of data mining or enlisting close reading techniques. They have, rather, come about with acknowledgment of the singular human dignity of each witness, at the same time that these witnesses have joined into a testamentary chorus of those on the receiving end of dis-equalizing and dehumanizing practices [Wieviorka 2006]. It is this mid-range reading that DH can aspire to enable and accelerate as a social justice project. Computational methods can help us to decide where to look and help us to look deeper and wider in the archive; they cannot reveal to us how to feel, react, or take action in the face of direct testimony to what the law looks like from the ends of fists and batons inside walled and razor-wired institutions; and it is the moral choice to act that is the beginning of changed practice. Computation can, however, atomize the foundation in human witness that the archive can provide for action. Our primary trust must and can confidently reside in the incarcerated (and the detained, the enslaved…); the voices from inside generously usher us forward into America’s Fourth City. These are teachers who want us to learn everything we can about closed spaces where the law’s violence is always on open display. The archive thus makes possible text analysis that will be of most concrete value when, beyond mining the words of incarcerated people, it is guided by the intentions of those inside. From the start, the APWA has prioritized respectful and non-discriminating posting of incarcerated people’s first-person witness to carceral experience; its faceted search capacities have been built exclusively upon metadata provided by incarcerated people, who complete a questionnaire that welcomes unprogrammed responses. (The questionnaire is linked at the APWA site. See https://apw.dhinitiative.org/collection-description.) Other than a signature granting one-time publication rights and contact information, all information is voluntary. The archive accepts this compromise regarding identifying information in face of the need for contributors to protect themselves and to be given full control over their representation. Writers can publish anonymously or under pen names. As described below, search terms for text analysis are now in development with a consultant who has served thirty-eight years in prison and who is helping to build a team of incarcerated and formerly incarcerated advisors. Every step taken outside prison walls has been and will continue to be guided by those inside.[3]

Enlisting DH platforms for concrete social interventions is now vigorous practice, [4] and such work requires ongoing theoretical revision and questioning. Yet it may also be under-explicated. Martha Nell Smith asks regarding DH projects, “How have these items of knowledge and the organizations and working groups who made them come into being?” [Smith 2014, 409]. This question should invite theoretical consideration; here the question is addressed as literally as possible, as a question about inception, sources, aspirations, outcomes, and potentials; and I suggest that it is in historicizing the trajectory of DH social practice that we set the grounds for the questions that theory might ask and address, thus extracting theory from practice.

As Barbara Christian reminds us, theory not based in practice is bound to remain elitist [Christian 1987]. In describing the analog birth and into-digital growth of the APWA, I present an instance of DH serving not simply as a means to expose and study a dis-equalizing institution, but to provide a platform where incarcerated and free people can meet to begin setting the grounds to generate new terms for critical legal and penal theory, and for planning (or conceiving an end to) new penal practice. When theory grows out of practice, it not only avoids elitism; it sets the stage for intervention. In helping broach the walls of prisons (or detention centers, or other state-sponsored lines of confinement), DH practice can constitute a transformative, political action, thus “shift[ing] the focus of DH from technical processes to political ones” [Bailey et al. 2016, 71]. [5] This essay suggests that DH can — as methodology — not only host witness born from the “lived experiences of difference in the U.S.” [Risam 2015] , but can also, if read as the dis-equalized direct us, facilitate moving the incarcerated to the front lines of efforts to uncloak the work of an institution built to enforce all manner of inequalities and to censor the dis-equalized. The distant v. close reading debate in DH [Ross 2015] replicates questions around how to balance addressing mass incarceration as a statistical fact — as a data set — and/or as a human and humanly articulated condition experienced and documented by discrete individuals [Simon 2010]. I want to argue for a plenum of reading proximities by advocating for the neglected middle-range of reading as more than a space of transition between close and distant reading. [6] What I will call cellular reading is the most apt manner of reception of what incarcerated writers — like other literary witnesses — do day by day, decade by decade, over millions of cumulative human years: create paths for readers to follow on the way to precipitating change in practice. Distant reading can aid in mapping the landscape of testament to law’s violence so that we know where to begin in tracking particular, and potentially class-actionable paths through the daily violations of human, constitutional, and civil rights and human decency inside the earth’s largest prison state; it remains in mid-range reading, however, that we will see outside hearts and minds moved to act.

In the spirit of drawing theory from practice, I build my argument upon the back of a biography of the project in point.

In the winter of 2006, I initiated a creative writing workshop inside Attica Correctional Facility. The class was entirely voluntary. It offered no college credit. It would not bolster anyone’s appeals for parole. Only the will to learn to write had brought thirteen men serving twenty to twenty-five-to-life sentences to the room. Over the ensuing ten years, the stories I read and heard painted an unrolling panorama of urban battlegrounds, rural poverty, and the occasional scene of suburban desperation — nearly all marked by some form of addiction or mental illness — as the backdrops to the booby-trapped road that runs from arrest, plea bargains or the very rare occurrence of a jury trial, to incarceration in a place built to produce pain, suffering, and rural employment [Haney 1998] [Huling 2002]. Their stories regularly reinforced and echoed each other, at the same time that each story was seated in an individual biography. In turn, I could not avoid hearing their reports both closely and at some distance: from the details of life in a home on a particular Bronx or Rochester city block, to the “public pretenders” nearly all had hardly spoken with before arriving in court. Over these ten years, dozens of men passed through that workshop. No two were much alike. Yet they held many things in common. [7] Thus began to appear a testimonial corpus seemingly pre-staged for text analysis: denial of freedom denies unlimited randomness in experience. The men who joined the class were Black, Latino, Native-American, Asian, and White, from large and small cities, and came from poor, working-class, and, in one instance, from an affluent background. Together, they offered stories of institutionalized degradation, of coping, and of unimaginable resilience. But these, of course, are not the impressions one gains from outside.

From its inception in 1790, the aim of the American penitentiary was less to cut off lawbreakers from the public than the reverse [Tarter and Bell 2012, 15] [Meranze 1996][Foucault 1995] [Schorb 2014] . In his Enquiry into the Effect of Public Punishment Upon Criminals and Upon Society, Dr. Benjamin Rush addressed the dangerous effects of letting the condemned speak to or even be seen by the public [Rush 1787] [8]; imprisonment would not only stanch such communication; it would deter crime by allowing the images of penal suffering to expand to the limits of public imagination [Rush 1787, 11]. Prisons today have become less a holding place for offenders than one relay in a closed circuit: we do not need prisons in order to contain deviancy, Foucault notes; we need them to mark a deviant class and thereby justify security apparatuses that both model and enforce behavioral norms [Foucault 1995]. Over the past forty years, the abandonment of rehabilitation and embrace of punishment, or the “new penality,” as the prison’s “raison d’etre” [Haney 1998, 58], has been built on the premise that incarcerated people — these (black and brown) “monsters” and “animals” — simply aren’t like “us” [Christie 1994, 42] . “The art of punishing,” Foucault reminds us, “must rest on a whole technology of representation” [Foucault 1995, 104]. Prison walls, today as in 1790, represent law’s monopoly on legitimate violence, they contain and recycle criminalized demographics, they silence the condemned, and in turn they allow fantastic images of the criminalized to be spread without intervention by incarcerated people themselves. The ruse that this is about public safety (incarceration has a marginal effect on crime rates [Study Finds Increased Incarceration Has Marginal-to-Zero Impact on Crime] [Mauer 2006]) can be maintained only as long as the public is not apprised of the full, if often wounded humanity of those inside. At the end of a forty-year US punishment wave, Rush’s fear of incarcerated people speaking to those outside is warranted not only by fear of evoking public sympathy; it would rip the screen upon which an eighty-billion-dollar prison industry and popular media have cast all incarcerated people as not-fully-human [Brown 2009] [Gottschalk 2015, 234] [Alexander 2010, 274]. Their voices might dissolve public moral pretense, revealing an industry built on the backs of caged and fully human beings. [9] Facilitating public access to what goes on inside through the testimony of incarcerated people is about shorting such closed circuits. It’s about revealing “monsters” as the very able, primary witnesses to, and most competent critics of the lie beneath claims of the threat that all imprisoned people pose, as well as the lies of due process, judicial equity, and respect for human rights. It’s about confronting and exposing the final, legally sanctioned enforcer of institutionalized racism, classism, gender-ism, etc.

The bars to the public serving as secondary witnesses to the primary witnesses of prison conditions are not only made of steel; they have been forged by the courts [Dolovichi 2013], congress [Prison Litigation Reform Act], and over two generations inside American minds. From the myth of the jury trial promulgated by Law and Order while only 2% US criminal cases ever goes to trial [Subramanian et al. 2020], to the carceral soap operas of OZ and Orange is the New Black, to the latest episode of the human zoo carefully staged for Lockup, American media have made fortunes selling misrepresentations of incarcerated people.

The critical literature on writing by incarcerated people is growing but is still in its early phase [Harlow 1987] [Harlow 1992] [Davies 1990] [Hames-Garcia 2004] [Rodriguez 2006] [Zim 2014] [Larson 2017]. There is only one broad historical survey of US prison writing, H. Bruce Franklin’s pioneering Prison Literature in America: The Victim as Criminal and Artist, which offers decade-by-decade readings of this work from 1798 to 1989 [Franklin 1989]. This book and the texts it explicates make evident, however, that for as long as the US has presented itself as a bastion of liberty, equal opportunity, and judicial fairness, prison witnesses have documented the nation’s working state of intersecting penal apartheids. Prisons isolate. They exile. They quarantine. Yet prison witness makes clear that these messages have never been simply from in there. They unearth for us the silent chambers beneath the nation Americans call here.[10]

Franklin’s history ends in 1989, only a decade into the era of true mass-scale incarceration (commonly dated from 1980). Important collections of essays by political and politicized prisoners have appeared since then [James 2005] [James 2003]; and Franklin himself has edited a survey of US prison writing in the twentieth century, reaching to 1995 [Franklin 1998] . But in summer 2009, as I found (again) in searching out new texts for the second iteration of an undergraduate course on US prison writing, there existed no broad national sampling of non-fiction writing by people then living behind bars. In attending national and international conferences for legal scholars and social scientists, I found that even researchers who had conducted surveys or scripted interviews with incarcerated people virtually all crunched this testimony into statistical charts and graphs (which, for their purposes, were certainly of vital importance). I heard no open ended, witness-driven testimony about carceral life. I had not set out to become an editor of prison witness (let alone a digital humanist), but the need — on campus and in the profession — became very clear.

With the help of a research assistant, the first call for essays by imprisoned people went out in fall of 2009. On February 1, 2014, Fourth City: Essays from the Prison in America was published, including seventy-one essays from twenty-seven states [Larson 2014]. The title reflects the total population of US prisons and jails. More importantly, as a first sampling of what a national gathering of contemporary prison witness might reveal, the title indicates a collective experience as cohesive as we might expect from citizens writing of their experience of New York, LA, or Chicago.

From among the first forty-five essays received, eleven stable categories of subject matter emerged and persisted into the final organization of the book. This occurred despite the fact that incarcerated people have virtually no means to communicate between facilities, let alone states, and often even between cells blocks inside any one facility[11]; so the consistency of subjects was striking: autobiographical sketches of the paths to prison, coping with prison life, slice-of-life examples of everyday challenges inside, ways to seek peace, the damage incarceration does to families, the injustices perpetrated inside, critiques of criminal justice policy, the inadequacy of mental and physical health care, activism from inside, prison reentry programs, and messages of warning or hope shouted out to kids on the street or others inside. Take the state names off these essays and it is impossible to tell they are not reporting from one institution committed to vengeance and willful dehumanization, and where the residents struggle mightily, day by day, to make lives worth the name. [12] What the project revealed is that our Fourth City is a metropolis of both willful debilitation and miraculous acts of human resistance, not the least of which is the very act of bearing witness. These are traits that anyone who has worked with incarcerated people might gather; one could do so, however, without gaining the basis for attributing common struggles and victories to a national population. But in gathering this work in a relatively short period of time, from across the nation, the outline of America’s carceral metropolis began to take shape. [13]

In her 2012 essay, “Why are the Humanities so White?” Tara McPherson distinguishes a lenticular view, in which two discrete images can be seen in one frame but never at the same time (her model for the separate-but-equal divide between the histories of computing and social justice activism since the 1960s); and a stereoscopic view, which layers distinct viewpoints on the same subject to create a sense of depth [McPherson 2013, 144]. What I had experienced inside Attica and what emerged from the Fourth City solicitation was stereoscoping of deeply singular voices, each implicitly demanding — like the human face for camp survivor Emmanuel Levinas, like Holocaust witness for critic D.G. Myers [Levinas 2010] [Myers 1999] — an ethical response to the witness borne by individual human beings who are also all exemplars among a mass population; such ethical response, moreover, is asked of a human listener or reader equally implicated in the legal order that concretely and morally marks the distance between the law-breaker and the law-abiding, the condemned and the protected, and those subjected to or spared by the law. Here, stereoscopicism occurs in a mirror: the layered depths of witness’ testimony to law’s violence reveal the degree and manners in which those outside are either spared and protected by the law (especially the white middle and upper classes) or co-objects of criminal justice involvement (especially poor people of color). As in reading witness texts from slavery, the Holocaust, or truth and reconciliation tribunals, it is impossible to hear or read prison witness without hearing echoes across biographies and texts, across razor-wired facility and state (and national) boundaries. It is impossible to ignore abiding tropes generated from spaces that hold those stripped of civil and basic human rights and effective citizenship status — spaces the absence from which marks readers as those in the name of whose security such quarantine exists, or as co-subjects of policing practices outside (much as looking into original stereoscopes impresses us as much with a view of a past, and with location of ourselves in a present, as with the subject at hand). Readers are asked at once to bear secondary witness, and to accept their own positions as either those in the name of whose protection incarcerated people are forced to struggle to regain their discursive existence, or as the similarly hyper-policed. The light cast by prison witness casts silhouettes around readers outside — silhouettes deepened as readers are exposed to each successive text, in contrast to the limited effects of close reading or pixilated reading at distance.

Stereoscopism better approximates but is not, however, fundamentally adequate to representing this situation, in which disparate and discrete biographies intersect at singular angles with the enervating spaces of the prison and with the legal positions of readers outside [Larson 2017, 12]. Prison witness texts — in their non-interchangeability and their clear linkages among themselves and in the minds of readers — are best figured by a term that is at once literal and figurative, at once metaphor and not, and that replicates the enunciatory moment of meeting between witness and reader. This is a meeting that is not so much stereoscopic as “cellular,” at-once literally of the body and of architecture, of vulnerable flesh like the reader’s and of the stone and steel that divide writer from writer and writer from reader, while each text carries traces of the DNA of the next and yet remains distinct, as the condition of identity and of life. Cellular reading is mid-range reading; it is reading at proximal distance. No text is the whole, but the whole gains its status as witness as an aggregate of each with each when each is read and held in the mind as a singular whole, rather than as a mere yet-to-be-aggregated and mined data vessel; authority lies in the “typical” moments across texts, while each text remains absolutely authoritative regarding the experience, anger, aspirations, hope, or hopelessness of its author — an author whose testimony, at the same time, is shaped by widely experienced conditions: conditions implicating the reader as part of the public in the public safety rationale for a city in cages.

Computation can accelerate location of common veins in this corpus; it can reach reading communities at rates unimagined by previous witness writers in their own time. DH can, in this way, facilitate a shift of authority from scholarship (or state agencies) to the communities that any work can garner, as Kathleen Fitzpatrick has observed [Fitzpatrick 2013, 453]. The APWA also makes clear, however, the need to mind the ethical stakes here. As the archive grows and computational methods are applied, the danger rises of reducing back to data points the testimony of writers struggling to assert their human dignity and singular integrity from among the data piles gathered by state agencies and research. Computational analysis should complement and aid cellular reading; my claim is that, between academically inclined close reading, and computationally enabled distant reading, it is from mid-range, cellular reading that human readers will be moved to social action. It is cellular reading that allows us to participate in the communities that can complete the witness transaction, recognizing voices that never stand alone or outside the legal matrix upon which we are all located; voices that are never effective until embraced by others, and that work directly in the face of organized authority that is often invisible to those outside.

Cellular reading is an irreversible process of revelation. Close reading one piece of prison witness (or slave or Holocaust narrative) is qualitatively distinct from reading five, and twenty, and a hundred; and while few readers can commit to reading thousands of essays, cellular reading can set the ethical context for data mining among much greater numbers.We need simply read enough to see how our ability to doubt, to find excuses for testimony that reveals the collective damage of collectively sanctioned and supported institutions thins and vanishes as we read repeated reports on, for example, lethal medical neglect, or feel the human toll paid by female victims of staff rape that research shows occurs in numbers that rival reported instances of rape outside [Gottschalk 2015, 137]. Amid multiple interfaces, some readers will discover cellular links that surround and mark them as the virtually law-immune or legally protected, while others will feel solidarity as those living outside prison yet in subjection to the law’s pathologies. [14] Such cellular reading sets the grounds for what Jamie “Skye” Bianco calls a “relational ethics” [Bianco 2013, 109] both between writers and between writers and readers, where “context, affect, and embodiment …remain viably dynamic and collaborative” [Bianco 2013, 108]. Johanna Drucker notes that “situatedness and enunciation are intimately bound to each other” [Drucker 2013, 92]. Reading prison witness reveals a causal exchange relation between these moments: when incarcerated writers write of their situation, the reader — as a co-subject of policing systems, or their beneficiary — sees revealed their own legal situation within the current order. If cultural influence is indeed at play in every act of interpretation [Clement 2016, 164], the interpretation of prison witness is irrevocably influenced by the position of the reader within legal culture. Unveiling the complexities of a city in cages, each text retains its singularity even as its uniqueness is dispersed across the whole. Close reading can acknowledge any one text’s ethical demand; but as the numbers of exposures grows an osmotic leak occurs; we hear the reverberations of beating on one steel wall, pouring and echoing into the next across facilities, and state borders, and across the affective interface of writers and readers. While we may then find bracing patterns in data mining the archive, it is the base in cellular reading that reveals that such witness is not like reading fiction or other literary texts bound by time or subject. Here, writing is not representation; it is human and (re)humanizing witness to the legal and extra-legal conditions that surround and demark citizenship and fully human status inside a carceral regime that, as Colin Dayan notes, metes out “state-sanctioned degradation … propelled by a focus on personal identity [and] the terms by which personality is … threatened, or removed” [Dayan 2011, xii]. Or, as a man from Texas writes, “Ah to be a human being again! Worthy of respect and affection, capable of dignity and contribution, to have a valuable life...” [Vitality 2019].

The following excerpts are offered as examples of how closely prison witnesses speak to one another’s experience, of cellular testimony to “state-sanctioned degradation,” and as practicum in readerly response:

Some readers may feel in these passages a seamless continuum with the accepted legal and public safety aims of incarceration. For others, such witness resituates the reader vis-a-vis the law’s magnetic center in state-sponsored violence [Benjamin 1996, 239–242]. As revealing as distant reading might prove, it also risks atomizing such an existential moment. This is a case in which, as Cuban digital activist Ernesto Orozo remarks, “answers ooze from the circumstances,” and “Renewed answers will always come from the resistance” [Gil 2016, 191–192].

The final deadline for Fourth City passed in the fall of 2012. But work never stopped coming. Cover letters apologized for arriving late and offered thanks for our willingness to read pages the authors assumed would never see print. The initial pool of 154 essays grew to two hundred, three hundred, then five hundred… Incarcerated people, given the slimmest chance, were willing to risk their safety in order to get word out. [15] Each day that I walked from the college mail center to my office, I carried yet more envelopes from California and Texas, from Alabama, Oregon and Maine. [16] But it was not easy finding a publisher for Fourth City (which one agent called a “book for people who can’t afford to buy it” [Sobel 2010]); in fact, I was still seeking that publisher. I had tapped a vein that would not be stanched. Happily, colleagues had secured a Mellon grant to start up and support digital humanities projects through collaborations between Hamilton Library and Information Technology Services (LITS), a newly minted Digital Humanities Initiative (DHi), and national collaborators [McFall 2016]. At the end of 2012, we began our conversations.

Credit for building the platform, storage capacities, and the sustainability of DHi-APWA lies with the funders and the DHi-APWA team and LITS in an ongoing, layered Venn relationship; but its inception, the populating of its metadata contents and resulting search facets, and the growth of its content are due to the unrelenting will of those inside to bear witness. [17] The APWA currently hosts 3,307 essays, the equivalent of over fourty-six volumes the size of Fourth City. It grows by approximately seven additional volumes each year (500 essays); and that rate is largely the result of a single, quarter-page solicitation in the pages of Prison Legal News, a monthly magazine dedicated to legal cases, legislation, and investigative reporting of interest to incarcerated people. That we receive this number of essays from a small ad is testament to the sheer size of the prison population; it is that size and scale that has allowed the statistical skills of the social sciences to lead the national discussion of mass incarceration [Simon 2010]; it is scale that also creates the ethical dilemma that the APWA itself now poses.

Incarcerated people and the writers among them are the subjects of intersecting statistical and documentary determinations: from so-called hot-spot policing, to sentencing guidelines that determine years of punishment from a grid; from the recording of their every demographic trait upon entering jail and prison, to psychological profiling, to the marking of every violation they incur inside, to prison, block, and cell assignment, to whether they recidivate upon release — release decisions increasingly influenced by risk assessment algorithms [Stevenson and Slobogin 2018] [Grann et al. 2007]. Such mapping and examination, Foucault writes, “engages [penal subjects] in a whole mass of documents that capture and fix them” [Foucault 1995, 189]. Incarcerated people are rightly wary of becoming the subjects of yet further datafication.

At the same time that we can use computation to help map the carceral landscape, socially responsible and responsive text analysis must align itself with, and serve to further the struggle of incarcerated people to be heard and read on their own terms, thus avoiding the danger of refracting and deflecting the work of prison writers — who are themselves the select representatives for other incarcerated people — through our own presumptions. In this vein, we might use computational tools primarily to locate, amid a mass of essays, those subject clusters that incarcerated writers make salient: such as the prison industry’s financial exploitation of incarcerated people, staff assault, damage to family ties, etc. Such indexing can help readers to see the pervasiveness of such concerns across the nation, respect the integrity of individual writers, and create the basis for focused advocacy. Text analysis can thus work for the kind of social justice action that incarcerated writers seek to forward. Socially conscientious data mining must recognize that witness texts are not mere data, they are not simulacra. They are the literary corpus of bodies and minds struggling amid the “penal harm” that the carceral state and the American public has embraced for nearly have a century as an explicit aim [Haney 1998] [Travis and Western 2017, 307]. Text analysis can serve this corpus and amplify its voices by maintaining to it less an analytical than an indexical relationship.

At the other extreme from distant reading, meticulously close reading will also not bring us to the understanding of prison texts and practice that prison writers overwhelmingly intend. Incarcerated people know they are held inside a mass-scale system. Even when they focus on their own situations, they clearly understand and articulate those situations as examples, not isolated instances. The image of a city in cages — while it reflects the cohesiveness of subjects and themes among prison texts — moves in the direction of better focusing our views of that mass into the diversity and distinctiveness of a major metropolis. The APWA’s facetted and key word searches allow further granulation — by demographic categories, locations, keywords, and layering of these factors. Computational work, as Moretti observes, offers “not an obstacle, but a specific form of knowledge: fewer elements, hence a sharper sense of their overall interconnection. Shapes, relations, structures” [Moretti 2005, 1]. From their beginning in 1790, as Michael Meranze points out, prisons have been spatial answers to social problems [Meranze 1996, 9] [Meranze 1996, 250]. They are shapes and structures that, as Dr. Rush intended, alienate human relations. What we should be seeking from computer-aided text analysis are mappings of these shapes, structures, and relations as they are reflected in their counter-discourse, in humanizing, mass-scale literary resistance. This is a literature that turns the prison’s shapes and structures inside out: a literary corpus that exposes the denaturalizing angles, shapes, and structures of penal confinement; a corpus of textualized bodies and minds and hearts, presenting for mutual exchange and transformation, the shared, fluid, and ineradicably interactive humanity of those whose freedom outside and unfreedom inside are built upon, and have been warped by, over two hundred years of broken relationships.

My conviction is that this damage will not be repaired by analysis alone. Instead, we might approach our work as that of curation, what I would term, “thick curation” — or, in a more text friendly term “thick indexing.” The attempt here is not to follow our interests in the archive, but to learn to see, and engage computation to aid our learning to see law and justice as imprisoned people see them — to see, for example, the wreckage that furnishes the path to so-called justice, and the ways in which the law and a broken civil society have created that wreckage. Thick curation and indexing will us to follow the branches of subjects and concerns among incarcerated people. One model for this work is Darwin’s sketch of speciation, reproduced in Moretti’s Graphs, Maps, Trees (Figure 1; [Moretti 2005, 63]).

Translated into a register of texts, this graph becomes an index. The following lines are excerpted from a list — based on the experience of the archive’s front-line readers of incoming essays, with the consultation of a formerly incarcerated person — that will serve to seed further work in text analysis.

Text analysis can help to populate these tiers with whole essays. Rather than atomizing prison witness, this indexing model recognizes each essay, as a whole, as an organic element of any map of life inside, and capable of presenting its own purposes. It is important to note that nowhere does the archive suggest that its makers presume to speak for incarcerated people, or presume to suggest what readers will or should find there. This may be why, as the archive grows, the rate of submissions increases [Growth History]. The APWA’s credibility deepens in the eyes of incarcerated people (with the aid of internet-enabled allies outside) simply because it has been trusted by other incarcerated people to treat their work in whole, without editing, and without secondary purposes.

Amid a highly politicized, national debate on mass incarceration,[18] the archive is built on trust in incarcerated people and archive visitors; it is not intended to make new arguments — as Tom Scheinfeldt once claimed it may not yet be time for DH to do [Scheinfeldt 2013, 56] — but to provide a meeting place where new questions can be asked about the human effects and costs of the current legal order. Incarcerated people have been spoken for (and objectified, medicalized, and demonized) for too long, including by academics. A thick indexing method seeks to disentangle branches of consensus that appear among prison witnesses even while imprisoned writers are denied communication with their peers. The APWA’s sole presumption is that it can present the cells of those branches and, in future, using computational aids, offer readers places to start on the path of extended cellular reading.

While DH work shrinks physical spaces, it does not necessarily change the direction of Supply (campus scholarship) to Reception (by other scholars, students, and the broader public). The APWA attempts to serve as the mediating facilitator of an ambi-valent, carceral margins-mainstream meeting point and conversation that, among other effects, sidesteps the “hack v. yack”[19] dyad at the moment theory assumes a secondary role to — and indeed grows from — an extra-academic dialogic. Sociologist Bruce Western observes that the public safety returns on incarceration are now badly outweighed by the social damage it imposes [Western 2018, 178]. The APWA offers visitors (including voters) the opportunity to measure the human breadth and depth of carceral damage. The project does not pose arguments but allows us to ask by implication what new “materialist epistemologies” [Ramsey and Rockwell 2013, 77] might be possible if we resituate incarcerated people from their current place in public discourse as a mass deficit, to positions as our foremost authorities on what imprisoned people are, what prisons actually do, the real effects of the politics and economics of the criminal justice industry as a whole, and how we might proceed on the path toward greater humanity in addressing all the social, civil, economic, political, and ethical failures that surround crime and racialized punishment. The archive allows us to consider what happens when we shift authority from prison administrators, law-makers, judges, prosecutors, police, and outside scholars, into the hands of the community of organic intellectuals among incarcerated people. It suggests that DH social justice work is what DH/SJW enables: demystification of historically dis-equalizing institutions, laying new grounds for newly goaled public policy and growing consensus about the practicality and moral imperative for a range of abolitions: from judicial misconduct, to police violence, the knowing creation of employment deserts, domestic violence, poverty, failing schools, to prisons themselves.

Continuing the literary legacy documented by Franklin, these writers are creating what may well become one of the largest bodies of witness literature in human history [20] — a literary corpus only possible on a platform that can sidestep the censors (market-sensitive agents, editors, and sales reps) of commercial and academic analog publishing. Unlike the most salient on-line witness archives — at the US Holocaust Memorial Museum [USHMM] and the University of North Carolina’s Documenting the American South’s archive of slave narrative [Documenting the American South] — the APWA is a repository of primary witness that expands as you read this sentence, as incarcerated people bear testimony to their experience. Though already the largest such archive, it is clearly in its early growth. Yet even now, its promise is clear: These writers are not only getting word out about what life is like inside; many write about their lives from birth to arrest, as well as through cumulative millennia of incarceration. What DH work has done in this case is to build a platform not only for those outside to look inside, but for those outside to see, from inside, the contours of US policing, judicial, and carceral practices, as well as from the interiors of the homes and communities affected by poverty, untreated addiction and mental illness, unemployment, crime, and the damaging feedback loop of mass-scale punishment that criminalizes poor people of color, the gender non-conforming, and the mentally ill — punishment that can exacerbate the very problems that precipitate crime [Clear et al. 2014]. That closed loop is only possible as long as the prison’s walls are reinforced by political and cultural circuits that remain unbroken, as long as the public sees its social distance from the condemned as based in moral virtue (rather than the chances of birth) and thus lives complacently with walls of concrete while consuming the salacious media misrepresentations used by “tough on crime” prosecutors, judges, and law makers to garner votes despite a steady, twenty-six-year decline in rates of crime [Golash-Boza 2017]. Breaking such circuits and dismantling such walls will not be achieved by close reading of a handful of essays, nor by reading data hit lists and clusters (however useful such reading methods clearly are to particular purposes and, again, to helping us determine where to read at greater depth). These simply are not how incarcerated writers conceive of their work or intend it to be read; and if incarcerated lives matter, so should the intentions among their representatives.

Legal scholar Robert Cover famously noted that “Legal interpretation takes place in a field of pain and death” [Cover 1992, 203]: the law suppresses violence by casting legal sanction over the acts of violence carried out by the police, courts, and prisons. Since the deaths of Michael Brown, Tamir Rice, Eric Garner, Sandra Brown, George Floyd, Breonna Taylor, and too many others, and the rise of the Black Lives Matter movement, the cost-benefit balance of this sanctioning (Black costs v. white benefits) have been called into question. Cover did not live to see his claims brought home by cell-phone videos, and he only tentatively pointed toward the corollary to his own claim: Without understanding both the breadth and the depth of human suffering experienced on the receiving end of legalized violence, we can never understand the full human cost of, let alone understand our location within the current legal order. This is a cost and location that prison witness, on a scale that only a digital platform can provide, makes clear.

As a social justice enterprise, new here are the technologies that offer rapid dissemination, facetted searching, and the rich rewards (and temptations) of distant reading that create the close v. distant, either/or dilemma. What is not new is the cellular or mid-range reading that the APWA helps to make possible, at an accelerated scale, for prison witness.

A first major proof of the power of a peculiarly American literary corpus, subjected to cellular reading, to foment social change came with the rise of slave narratives. First-person testimony by former slaves would prove one of the most effective weapons in the abolitionist arsenal [Johnson 1972]. While documenting previously unimaginable conditions of degradation, and thus confronting southern propaganda, the witness borne by former slaves presented depths of moral and intellectual acuity, emotional allegiance and social alliances that few white readers had to practice in their own lives. [21] Like incarcerated people today, former-slaves countered popular misrepresentation of a dehumanizing institution, protesting their condition by bearing witness to the dignity of the objects of popular invective. This was not achieved by close reading of a handful of texts, nor, of course, with data sets, but by cellular reading, in which every whipping taken by one author stood proxy for a thousand, including those whose experience never saw print, and thus exposing the moral culpability of passive compliance. All witnesses speak for the muted, for the musselmanner, as the living called the walking dead inside Auschwitz [Levi 1995] [Wieviorka 2006]; and all witness recipients are (at least potentially) morally repositioned by such reception. The aim of the APWA, after 231 years, is to open precisely that public, “affective exchange” that Benjamin Rush sought to block in fear that exposure of human suffering would evoke sympathy and desire to relieve that suffering [Schorb 2012, 167]; an exchange much like that which slave narrative helped to precipitate war and the abolition of slavery. With this aim, the argument here is, again, not against data analysis. It is for a plenum and balance of reading proximities, from close, into cellular reading (with its tried historical precedents), to those generated by computation, and for an ethically responsible complementarity of reading proximities that never loses sight of the primacy of whole statements by whole individuals. Stephen Ramsey remarks that DH is strongest when it resists seeing data as “more valid” than humanistic inquiry [Ramsey 2016, 528]. My claim is that such a balance — and embrace of the broad, deep, and morally engaging middle ground between the close and the distant — is best suited to affect change in social practice. History and prison witnesses themselves have rushed the gates of analog publication and into DH; we must take care to avoid atomizing their testimony.

Witness literatures are resistance literatures, and “The literature of resistance sees itself…as immediately and directly involved in a struggle against ascendant or dominant forms of ideological and cultural production” [Harlow 1987, 28–9]; they are grounded in concrete and whole human experiences of the effects of these forms of productions — including the production forms of digital computation used by state agencies — and they are collectively dialogic: duty, propriety, right and wrong, justice, and equity bear different denotations to a slave and to an incarcerated person than they do to slave masters and prison guards. Acts of prison testimony document not only how failures in public policy, civil services, mental and physical health care, foster care, public education, policing and court and prison practices are translated into broken lives; they instruct us in how to retain our humanity in the face of legally organized suffering, which also includes retaining human status in a world beset by datafication. As Jill Stauffer writes, “…understanding what dehumanizes also helps us determine what we take a human being to be” [Stauffer 2015, 15]. For even while prisons strip incarcerated people of other means of attempting to make good of their pasts, they are doing the work of restorative justice, offering up what they have learned from their lives, their crimes, and from our punishments. This is the reason why we must pick our steps carefully between the ironies that line the path of (re)datafication of voices struggling to break free from the data piles gathered by the administrators of state violence, and by researchers outside. As noted, APWA writers provide the contents of the archive’s metadata; they can also choose to withhold that information, not only for their own protection, but as an exercise of freedom.

Commitment to individual witnesses starts with granting full discretion over the metadata they provide. It continues with the hard-copy documents we receive, scanning pages so that visitors can see (as far as flat screens are able to convey) the material condition of individual texts. These include hand writing (in about half of all submissions; Fig. 2); texts typed on manual typewriters and word processors — the latter available to those few with access to computer stations in classrooms or who have white collar jobs inside (Fig. 3 and 4); and, less frequently, texts previously published in prison newsletters and newspapers (the latter by third parties outside; Fig. 5, 6).

While the value of all prison witness is in principle equal, the conditions of composition clearly are not. Prison-composed texts show evidence of state-imposed “resistance in the materials”; but rather than “reveal[ing] opportunity,” as Bethany Nowviskie claims [Nowviskie 2016, 178] , this manner of resistance shows the unequal distribution of opportunities inside. A text composed with a ballpoint pen filler (the plastic cylinders prohibited since they can be turned into tattoo guns, syringes, or weapons) and a library staff clerk’s word-processor indicate different opportunities in the very act of composition. A newspaper article necessarily indicates outside support. The graphic state of texts offers one consistently distinguishing material indicator of what it is for the writer at hand to write from inside. As Paul Fyfe claims (in another context) “the accidents are often essential” [Fyfe 2013, 274]. The accidents of prison-witness documents announce their emergence from the unequal conditions of writers in willfully low-to-no-tech environments.

This, of course, is lost amid big data, along with the voice of the individual, and the distinct conditions inside each prison, prison block, and prison cell; and such documentation is, after all, largely the point for these writers. Life with an aggressive, schizophrenic cellmate in a grossly understaffed prison is quite different from a tenth year in solitary confinement, and from life with a cellmate with whom one shares interests in, say, Scrabble and historical fiction. Only mid-range reading can account for such differences in the way that these writes intend, and in the ways that can affect the ethical agents lighted up inside witness readers — however well text analysis might supplement any such effect.

The next level of individuation of these writers is, of course, the voices of the texts, which range from desperate, semi-literate cries for help, fact-based policy critiques, and hip-hop inflected shout outs, to structured legal briefs, angry diatribes, wistful pleas, clinical treatment proposals, and everything between.[22] Beyond single voices, there is the subject matter of these essays, which, as noted, narrows the range from all included voices to a recognizable set of issues (though often multiple and mixed inside individual essays) inside facilities whose basic mission is the same: to confine human beings against their will and do them damage. While we mine for broad sets of words, sentiments, syntactical patterns, etc., cellular reading reveals that even when general subject matter may be similar, and the language similar enough to fall together in text analysis, each writer brings their individual biographies, conditions, and experience to these issues in distinct settings; each writes at once to document and protest their particular conditions, and to represent US incarceration en masse. (Even in the small sampling above in figures 2 through 6, note how often writers assume the authority to address the prison system as a whole.) Mass incarceration dehumanizes by treating individuals as mere units of a mass. Data mining can no doubt reveal important and telling patterns in such statements, and it can lead us to rich sites for closer reading; but it can also — like state actors — reduce imprisoned people and their discourse to types. We cannot undo mass incarceration by the massification of prison witness; nor, even as it helps us map, can we expect data points to evoke the affective responses needed to motivate advocacy for changes in penal conditions, thinking, or policy. These writers write at their risk, in order both to distinguish themselves from and to witness for a mass population. The mid-range reading for which I’m advocating is a practice that contributors assume. Lexical patterns and word frequencies can reveal broad trends within a corpus; they are not what prison witnesses intend to create, and it is with respect for and to seek guidance from the incarcerated that the APWA was built, as a basis for policy critique and activism aimed at concrete changes in practice. The other side of this project is, of course, in the hands of readers using the search capacities availed to them by the site and populated by information provided by incarcerated people.

Visitors will come to the archive with a range of motives and curiosities. It is in the aggregate of visitor reading patterns and purposes and outcomes where there unfolds the critical mirror effect, the greater-than-individual/less-than-mass-scale balance point (> <) that serves as a fulcrum for the demassifying of mass-scale witness while fully cognizant (as are prison writers) of a mass-scale backdrop. One can, for example, select out essays by writers who self-identify as African American, as male, and as confined in Michigan, thus curating, for the first time, a cellular network of men who can speak to racial and legal conditions in that state. This is the point (both critical moment and purpose) of cellular or proximal reading aided by search capacities. The many ways in and subjects about which incarcerated people write become animated by the many ways of searching, reading, and curating this work. Daniel J. Cohen notes that “Curation becomes more important than publication once publication ceases to be limited” [Cohen 2013, 321]. Working with prison witness, curation becomes inextricably political; it expands, into social and legal practice, Alan Liu’s claim that DH can establish new communities between the humanities and the public [Liu 2013, 498]. One community that has been created of necessity around the APWA — creating a new opportunity from the “resistance in the materials” — is that of transcribers of handwritten essays into internally searchable texts. “Transcribe-athons” held at libraries, courses that offer transcription as a requirement option, and people outside of academia volunteering their time have joined in the work of making carceral witness available to search. At their best, such projects refocus DH on the “intersection of digital production and social transformation” [Bailey et al. 2016, 72]. On the other, big-data side of this meeting point, computational analysis is, as noted, now in progress, and it will open a range of reading, research, and secondary witness methods to view and to multiplication.

In the final chapter of Michelle Alexander’s 2010 best-selling The New Jim Crow — a book that helped raise the conversation about mass incarceration into a national debate [Gottschalk 2015] [Weaver 2017] — the author warns: “Isolated victories can be won — even a string of victories — but in the absence of a fundamental shift in public consciousness, the system as a whole will remain intact” [Alexander 2010, 234]. Political scientist Marie Gottschalk echoes Alexander’s conclusion and (inadvertently) suggests where such a shift might begin, modelled on an earlier literary revolution:

The slave narratives of the antebellum period, which graphically rendered the physical pain that slaves suffered and made it widely visible, helped to propel the abolitionist cause. Today, what happens in prison stays mostly in prison, making it harder to draw connections in the public mind between justice on the inside and justice on the outside. The ability to identify with an offender — or not — is a key predictor of why people differ in their levels of punitiveness. The invisibility of the millions of people behind bars has made it extremely difficult to alter the negative portrait that members of the general public have in their heads of people who have been convicted of a crime. They are simply prisoners and criminals. As such, they often are denied their humanity and any right to democratic accountability, much as slaves were in the United States.

[Gottschalk 2015, 274]

The aim of the APWA is to unveil the prison and imprisoned people from centuries of intentional invisibility and willful misrepresentation, as well as to invite humanists of all stripes — with their highly nuanced, diverse, and diversifying manners of reading — back into the work of helping to understand what US prisons actually do, what the full, rich range of imprisoned people encompasses, and to allow and invite the general public, as well as lawmakers and the families of incarcerated people, to meet those who are arguably our most incisive guides to a world beyond prisons. My hope is that the APWA, as well as other DH workers and especially those working with first-person witness corpora, will use the powerful tools now at hand to amplify and arrange, as witness voices dictate, rather than merely to datify the testimony of those on the receiving end of collectively organized, silencing, and too often silently suffered institutions of legal quarantine.

For every man, woman, or trans person recorded on video while dying at police hands, thousands of Americans are quietly disappeared into US prisons. These people are now speaking in mass because DH/SJW has allowed them to speak from across state boundaries, from city and country, raising a chorus that sounds out every pothole on the road of opportunity supposedly open to every American. Reading of the suffering demanded for stumbling upon that road, we are forced to reassess whether any concrete facts lie behind the language we use to discuss rule of law, due process, and justice. No political representative, no social scientist, and no one else speaking from outside can provide what direct witness to legally sanctioned suffering offers. No other platform can open that witness to a world that today needs to learn humility before the voices of those on the receiving end of populist and demagogic rhetoric, which saw its recent inception in the rhetoric of “law and order.” [23] Countering the dehumanizing, tribal nationalism we see rising today, witness literatures offer the chance to strengthen as we test the common terms we use to understand our humanity; and this is understanding that we need in order to underwrite democratic institutions.

This is what empathic revolutions do: they open one lexicon in order to recalibrate another, humbling unquestioned presumptions about what constitutes the human. Prison witness has the potential to dismantle and reset the terms of discussion of a nation that regularly claims itself a model to the world of humane, democratic, and egalitarian practices, offering a new vision of how we might proceed. It illuminates, as the testimony of former slaves once did, just where and how civil life, its capital means, cultural and social assumptions, and majoritarian tyrannies — once translated into legal enforcement — have worked to sustain an American underclass. On its basis — gathered, archived, disseminated, read, curated, as well as analyzed, and mined at scale — we might be positioned to recalibrate how we theorize and practice law, punishment, and even nationhood. [24]

My conviction is that readers willing to read prison witness will find spread before them not only a richly illustrated map of our failures to care for each other — civic, political, ethical, economic, and social — but a map as well for building a nation empathic at last to all of its members: citizens and non-citizens of all races and ethnicities, all classes, genders, creeds, abilities, and sexualities… In response to calls for DH to address inequality and social injustice, digital humanists can seek to build platforms for activism modeled on proven reading methods of the past while keeping their eyes on the yet untouched horizon of a more humane and equitable future.