The Ethics of Digital Data on Convict Lives

© Neo.

The Digital Panopticon project represents a substantial intervention in the nature of the historical record. Simply by reproducing and making newly searchable the records of nineteenth-century criminal justice and transportation it necessarily selects the experience of a distinct group of mainly British nineteenth-century working class, white men (and some women) for particular attention. In doing so, it both defines them by their crimes, and excludes others form our gaze, notably aboriginal Australians and most women. It also arguably reproduces the racist, gendered and imperialist categories of the nineteenth-century British state. There are no easy answers to the ethical issues raised by this project, and by digital history more generally, and no uniform view across the Digital Panopticon team about the issues raised. This page is designed to simply highlight the ethical conundrums raised by this project and our responses to them, in an attempt to stimulate further discussion.

Introduction

Jeremy Bentham's plan for the "Panopticon" prison (1791). From Wikimedia Commons.

The Digital Panopticon project presents three interrelated ethical challenges. The first lies in the drawing of attention to the misdeeds and misfortunes of the dead - are we engaging in a form of Digital Dark Tourism, exploiting the misery of others? The second lies in the selective re-working of the records of the nineteenth-century state. Are we digitising the right records? Are we inadvertently re-enforcing the biases of the archive? The third ethical challenge is created by the choice the project has made to partner with commercial providers of genealogical data. If using this website simply forces researchers to register and pay for access via commercial providers, we need to ask whether this undermines the public curation of memory? In sum, we need to ask what duty of care we owe to descendants? And what duty we have to challenge the contemporary uses made of the lives recorded here?

Uncovering Past Misfortunes

The issues raised by the repurposing of institutions of punishment and suffering, including historical prisons and courtrooms, as popular tourist sites are addressed on the Dark Tourism page. A digital resource containing the details of the crimes and punishments of hundreds of thousands of individuals raises similar issues. Our 'Life Archives' expose to a wider public previously hidden patterns of offending and victimisation, hurtful bereavements, loss of children, bigamous marriages, affairs and liaisons, and so on. What obligations do we have to the (dead) people who appear on this website? What about the rights of their descendants? (Richardson and Godfrey 2003, Godfrey 2011, Godfrey 2016)

One cannot libel the dead. The laws of defamation and libel do not follow beyond the grave. But this does not preclude our obligation to speak the truth about them, as much as we do about the living. Historians have nevertheless felt entitled to speculate about their long-dead subjects more freely than a social scientist, for example, would do about the living. The British Sociological Association recommends that personal information should be kept confidential wherever possible and suggests that in some cases it may not be appropriate to record "certain kinds of sensitive information" at all. In the Digital Panopticon, we only have dead people. Rather than trying the impossible task of guessing which person would or would not want their personal details to be revealed, we sought to reveal as much as we can. Should we therefore try and anonymise the names whilst retaining the other details?

Data deposited with national repositories usually has the names of research subjects anonymised, and depositors are requested to adhere closely to rules on anonymity. Even when dealing with historical data, Godfrey et al (2007) gave pseudonyms to all the criminals in their study of Victorian Crewe. And a similar study, of children who were placed in institutions between 1750 and 1940 because they had committed a crime or because they were vulnerable (Godfrey et al 2017), also anonymised names because of the age of the children involved – their grandchildren and even their children in some cases would still be alive.

This approach was neither practical nor desirable in this instance. The methodology of the Digital Panopticon is record linkage - algorithmically connecting data based on names and other personal characteristics. To allow users to verify these links, names and criminal details need to be available. It is also a principle of scientific research that scholars should be able to repeat and test the research process pursued by others. For the Digital Panopticon this requires that names are revealed. In any case, information about our convicts is also publicly available in archives or in existing digital sources.

If the question is whether the public should be able to access data, then modern criminal cases provides an apt parallel. Modern trials are held in public and reported in newspapers. Yet court records have a closure period of thirty years (for adults) and seventy years (for children). They are closed because the cases continue to directly impact on the lives of the participants (defendant, victim, witnesses). This does not apply here, where the direct participants are long dead. But do living relatives really feel that distance has reduced the potency of the information?

As with a genealogical website, descendants researching their ancestors on the Digital Panopticon may find information they would prefer not to know. A long-cherished family story may suggest that an ancestor was transported to Australia for stealing a handkerchief - a victim of British justice. The story may go that he then re-made his life in a new land, and became a founding father of a new state. The revelation that he was in fact transported for rape or bestiality might not be a welcome addition to the family narrative. Nor would the knowledge that his Australian marriage partner came at the cost of an abandoned spouse and children in England; that rather than leading an honest life in the new colony, he was a recidivist who was prosecuted many times for drunkenness and public urination. Across a century of memory, none of this may not matter at all, or it may affect his descendants greatly.

If you engage in historical research, you must be prepared for whatever information you may encounter. The Digital Panopticon does not hide any information. How that information is interpreted, however, is another question.

Interpretation

Convicts wearing anonymising masks at Pentonville Prison, from Henry Mayhew and John Binny, The Criminal Prisons of London and Scenes of London Life (1862). From Wikimedia Commons.

"Perhaps the most invasive risk for participants in narrative research has to do with the emotional impact of having one's story re-interpreted and filtered through the lenses of social-scientific categories...Narrative research can in this way become intrusive and subtly damaging..." (Smythe and Murray 2000:321).

The subjects of the Digital Panopticon are dead, but their living relatives might not agree with the conclusions academics reach using their ancestors' lives. Because we have made all the data available, users will, of course, be free to draw their own conclusions, and be able to contest the interpretations we have constructed - indeed, they are encouraged to do so.

Using the large amounts of biographical data in the Digital Panopticon, researchers may feel confident that they know enough about a person and the direction their lives took in order to state with a certain amount of authority the connections between offending and other life events (Godfrey et al 2007, 2010). We should remember, however, that the Life Archives in the Digital Panopticon, no matter how detailed, provide only fragmentary details about the lives in question, even with respect to their criminal activities. We are not, and will never be in a position to fully understand the motivations that underlay particular actions. We can correlate, and infer, but can rarely be conclusive. In bringing together so much data about one person and presenting it in the framework of the supposed authority of an academic website, the Digital Panopticon can give the impression that we know more about our subjects' lives and motivations than we actually do. It is an ethical duty to take care with the ambition of our analysis, and to be suitably cautious in our interpretations.

It is also important to recognise the potential for error within the data provided on this website. There are transcription errors, and even more errors in the record linkage (itself a form of interpretation). Given the scale of the records included, record linkage has been primarily an automated process, with an associated risk of error. Some records pertaining to the same individual which should have been linked have not been, while some life archives contain links which are false. Users who spot errors in transcriptions or links are invited to contact us using the Contact Us form.

The Digital Panopticon allows all viewers to interpret the data for themselves. However, researchers, family members and everyone else will need to remember that historical analysis is not straightforward and care should be taken to acknowledge the volatility and complexity of people's lives. This is also the case when selecting the lives to which analysis will be applied.

We must always remember the limitations of the records. The Old Bailey did not see every London criminal pass through its doors, only those who committed an offence within the jurisdiction of the court against a victim who was able (often because of their position or wealth) to secure a conviction. Our database contains more people from the lower strata of society, because the criminal justice system pressed down hardest on that section of society.

Of those who were convicted, some will have broken the law and some were wrongly convicted, but as far as our database is concerned, they were all defined by the courts (and society) as criminals. There is no point in attempting to re-try cases, but it is important to remember that those identified as criminals in the nineteenth century might not be judged that way now.

Online research is rarely the last word in answering historical or criminological questions. There is almost always more evidence available in archives and libraries, particularly in newspaper accounts of arrests, trials and punishments. As explained below, we encourage all users to follow up their online enquiries with further research elsewhere.

Selection Bias

While the Digital Panopticon introduces one form of distortion in calling disproportionate attention to the alleged crimes of its subjects (as opposed to other aspects of their lives), an equally important bias is its failure to say very much about the lives of many other people.

Less than a quarter of the records reproduced here relate to women; and a vanishingly small number relate to aboriginal Australians, or black Londoners. In these records, the contexts of imperial expansion and the massacres of indigenous Australians, of a global system of criminal transportation and forced labour, are obscured. The records themselves focus on young, white, British and Irish male convicts – drawing our attention away from the complex exercise of racial and gendered power, and focussing that attention on the single relationship between white, male convicts and the state.

While the records themselves dictate this selection, we justify it in pursuit of a new "history from below". Although the convicts whose lives are detailed here are, in global terms, privileged by race and gender, they also were at the sharp end of an oppressive system of criminal justice, and represent a class of people generally excluded from history. Many of them are Edward Thompson's "poor stockinger" in need of rescue from 'the enormous condescension of posterity' (Thompson, 1963, p.12).

These records represent a unique body of details about non-elite people in the past. For the vast majority of the world's population in the nineteenth century, there is simply no written record of their experience – and none with the levels of detail available for London's convicts. By reproducing this body of materials, which at the least gives voice and shape to a small percentage of non-elite lives, the Digital Panopticon seeks to contribute to that wider project of recovering the lives and experiences of past generations of working people.

The Main Reading Room, Libary of Congress. Carol M. Highsmith (image ID highsm.11604). Wikimedia Commons.

This also justifies what might be seen as an undue London focus in the Digital Panopticon project. We start with London because that is where the Old Bailey court was located, and the published accounts of its trials, the Old Bailey Proceedings, constitute the largest body of historical texts detailing the lives of non-elite people ever published. The unique details provided in the "Proceedings" form a cornerstone for all the "Life Archives" which can be created in the Digital Panopticon. This creates the danger that future research on crime and criminal justice, facilitated by the existence of digital sources, will be disproportionately focused on London. But it does not have to be: we strongly encourage researchers to investigate other criminal jurisdictions, and, where possible, link their cases to records in the Digital Panopticon, particularly since many of the Punishment records on this website include convicts from other parts of the country.

Moreover, since only a minority of the archives, even the criminal justice archives, have been digitised, anyone seeking to tell the full story of a convict life will need to get up from their chairs and go to the archives. They need to get their hands dirty by consulting the manuscript archival materials. The Digital Panopticon should only be the starting point for further research and debate; the "Further Information" sections of all the information pages, and particularly the research guides, provide suggestions for how to go about this.

A more subtle bias also informs these records. We have sought to reproduce the precise language and categories of data recorded in the originals. But this effectively reproduces nineteenth-century assumptions about race and gender which shape how these records are read. Again, there is a fine balance between the value of making historical materials available to all, free of restriction, and the extent to which these materials actively encourage misinterpretation. We believe that the balance falls in favour of access. Users are reminded that, despite their appearance on a digital interface and in modern fonts, the records reproduced on this website reflect the language of the original sources, and need to be interpreted as such. To assist with interpretation, in our Historical Background and Records background pages we have sought to provide detailed context for the records included in the hope that researchers using this site will take seriously the historians' duty to represent the past accurately, and to contextualise their findings fully.

We are painfully aware of the extent to which the lives made available through the Digital Panopticon have been used in the construction of both a specific "white" national identity in Australia; and in a British context, a narrative of national evolution that largely excludes the impact of the Empire. Ironically, given the substantial over-representation of Aboriginal peoples in Australia's prisons, and of minorities among Britain's, dead convicts have been marshalled to create a distorted and incomplete picture of past that excludes just these people. The Digital Panopticon believes that the best antidote to selective and self-serving history is fuller evidence of more lives.

Partnerships

Wandsworth Prison Prisoners Photograph Album record for Sarah Durrant (1872). The National Archives, reference number PCOM 2/290, f. 46. © The National Archives.

The Digital Panopticon aggregates data from a large number of datasets, most of which were created by others. We are grateful to all our partners for giving us permission to include their data. Where that data is freely available, normally when it has been created as part of an academic project, we include all of it.

But much crucial data is only available from commercial organisations, which have only given us permission to include basic details of the underlying data, leading users to a paywall before they can access the full record. We recognise that this can be frustrating. And for a public resource funded by the Arts and Humanities Research Council, it raises a serious issue. Ideally all public records would be freely available to the public, but this is not possible. The Digital Panopticon sees its arrangements with its commercial suppliers as a necessary and justified compromise. Undigitised records, sitting in government archives, are not free. The charge for their use comes in travel costs and time in a physical search room. The documents made available here form a real advance on the privileged nature of access created by the paper archive. But the digitisation of those records costs money, and the companies which paid for that understandably seek a return on their investment. The inclusion of commercially digitised materials is predicated on the belief that in the continuum from "free" to "charged" – in which accessing the physical archive is perhaps the most expensive option – including pay-walled materials is justified.

Our strategy has allowed the project to incorporate numerous records that add substantial value to those made available for free - whether or not users choose to pay for access to the full records, they have been alerted to the existence of such records, which are available both physically at The National Archives and online. We are grateful to our commercial partners for agreeing to this mutually beneficial arrangement.

In Pursuit of a Balance of Usefulness

The Digital Panopticon builds on twenty years of effort by teams across the globe to digitise and make available the sources of history. Most obviously, it incorporates and builds upon the Old Bailey Online and Founders and Survivors digital resources. As a generation of historians, we have been uniquely privileged in having the opportunity to shepherd the records of the past from their physical form in the archives to new digital formats. In the process, we have been challenged to determine what is worth digitising and what is not. We have also been confronted by the clear biases of the archival selections made by past generations. One need think no further than the wilful destruction of the records of decolonisation by the British state in the 1950s to recognise that the selection, preservation and publishing of historical records forms a powerful political intervention.

The Digital Panopticon represents an active choice in which the depth and quality of information available for the subset of working lives found in the data is balanced against the undoubted bias of the archive itself. Through the way we present the data we have attempted to ensure that those biases are rendered clear to the user.

And by explaining the rationale for our choices, we hope to stimulate further debate on the newly emerging topic of the ethics of digital data, not only on the history of crime, but for history more generally. We fully recognise the space for alternative points of view, and encourage further debate.

Further Information

Author Credits

This page was written by Barry Godfrey, Tim Hitchcock, and Robert Shoemaker, with additional contributions by other members of the Digital Panopticon project team.