Distantly Studying the Long Eighteenth Century during the Covid-19 Pandemic and Beyond
Some very useful lists of the cornucopia of great 18th-century digital resources are available here:
British Association of Romantic Studies: http://www.bars.ac.uk/blog/?p=2900
Centre for Eighteenth Century Studies, University of York: https://www.york.ac.uk/eighteenth-century-studies/resources/#tab-5
Omohundro Institute of Early American Culture and History:https://oieahc.wm.edu/explore/vastearlyamerica-resources/
We would also recommend Paddy Bullard’s 2013 introduction to eighteenth-century digital resources in Literature Compass, https://onlinelibrary.wiley.com/doi/abs/10.1111/lic3.12085
To help you navigate some of these sources and consider the issues they raise, members of the Centre for Enlightenment Studies at King's offer some more detailed reflections and subject-specific advice below.
- Rowan Boyson, Co-Director for the Centre for Enlightenment Studies
Digital resources are a fantastic tool for the study of the long eighteenth century. There are a huge number of comprehensive repositories to access primary sources, but I am particularly interested in digital resources that focus on helping users analyse connections between sources. These sites are usually based on a discrete source set that users can explore through a number of tools developed by the project team. These sites provide both access to sources as well as a deeper understanding of eighteenth-century societies. These internet resources are particularly useful in understanding more about phenomena that leave only a very small imprint in the archive.
The study of slavery has particularly benefitted from these types of digital interventions. Slave Voyages (https://www.slavevoyages.org/) was a seminal project for uncovering the lived experiences of enslaved people. The database allows users to trace the voyages of slave ships across the Atlantic and within America. It is a valuable tool for contextualising the people and places of the slave trade. Similar projects have focused on the recovering the lives and activities of enslaved people in greater depth. These include Freedom on the Move (https://freedomonthemove.org/), which uses runaway slave ads to understand more about how enslaved people resisted slavery. Likewise, Race and Slavery Petitions (https://library.uncg.edu/slavery/petitions/) collects together legal documents that provide snapshots revealing the names and activities of enslaved people. Meanwhile, Slave Biographies: The Atlantic Database Network (http://slavebiographies.org/) aims to connect primary source collections and the work of contemporary researchers in order to reconstruct as fully as possible the biographies of enslaved people. Finally, the Legacies of British-slave Ownership project (https://www.ucl.ac.uk/lbs/) looks at compensation claims from slave owners to the British government after the 1832 abolition of slavery. The claimants wanted money in compensation for the loss, i.e. the manumission, of their property. The site demonstrates that many more people were involved in slavery than just a wealthy elite.
And it is not just the study of slavery that benefits from digital resources. Locating London’s Past (https://www.locatinglondon.org/) uses a variety of source sets, including poor relief, fire insurance claims, and Old Bailey court records, to map out different perspectives on London. Meanwhile, the Digital Panopticon project (https://www.digitalpanopticon.org/) also uses court records in and other documents to help reconstruct the lives of ordinary Briton. These are only a few highlights among a vast number that showcase how digital tools can be used to connect sources together. In so doing, these tools are useful for overcoming the structural inequalities in how eighteenth-century documents are preserved and help researchers uncover the lives of otherwise marginalised people.
The power of digital tools to connect eighteenth-century sources together is also useful for my work in book history studies. Most notably, these digital resources are fantastic for understanding more about the ephemeral experiences entailed in how audiences read books. For example, the French Book Trade in Enlightenment Europe (http://fbtee.uws.edu.au/stn/interface/) maps the business network that put Enlightenment texts in the hands of readers. On a similar theme, Mapping the Republic of Letters (http://republicofletters.stanford.edu/) maps correspondence between intellectual luminaries, like Voltaire or D’Alembert, to understand more about the social network that underpinned the Enlightenment. Two final projects confront the ephemeral practices of reading and writing. The Reading Experience Database (http://www.open.ac.uk/Arts/reading/UK/) collects together excepts and citations from various sources to connect texts with actual readers. Meanwhile, the Quill project (https://www.quillproject.net/quill) looks at the drafting work involved in writing the 1787 United States Federal Constitution. Again, these are only a short overview of the type of projects made possible by digital resources.
These resources are important not only for the conclusions that can be drawn directly from using their site, but the project teams also designed their project to help users gain a fuller understanding of the eighteenth century through using their tools digital tools. Searching, mapping, and even browsing these resources should help users understand more about the societies that created the eighteenth-century sources. In fact, I would encourage people to engage with as many different digital resources as possible because they all help in researchers looking for ways to build connections within and between other source sets. For more information and suggested readings, the ‘About’ sections on each of the recommended websites have excellent entry points for further study.
James Grande
One application of digital methods has been to liberate texts that no one has ever wanted (or would ever want) to read in codex form. The diary of the political philosopher and novelist William Godwin is one such object. Existing in 32 notebooks in the Bodleian Library in Oxford and spanning some 48 years, it is not a discursive diary, or one that you could sit down and read cover to cover. Instead, it is essentially a set of lists – very meticulous lists, recording everything Godwin read, people he met, topics of conversation, visits to the theatre, and public and personal events, often recorded in oblique or cryptic form. In this sense, it is ideally suited to a digital edition, which can be searched and analysed. I worked as a research assistant on the Leverhulme-funded Godwin Diary Project, part of a team marking up the entries in XML code, assigning unique tags for each person, book or event that appears (11,000 meals, 33,000 meetings, 60,000 names...) and writing annotations. Since the digital edition of the diary was launched, it has been joined by the remarkable Shelley-Godwin Archive, which houses digitized manuscripts of Godwin, Mary Wollstonecraft, their daughter Mary Shelley and Percy Bysshe Shelley. Manuscripts which were once only available to a handful of scholars in major research libraries can now be examined by students all over the world – a true Enlightenment dream.
The book I’m currently writing, on music and the culture of religious dissent, begins with a chapter on William Blake, and this has led me back to the Blake Archive. This was a pioneering digital humanities project, first launched in 1996, and continues to grow and evolve. It’s no exaggeration to say that it’s now hard to imagine teaching or writing on Blake without it. The editors frame their work ‘as an extension of ongoing archival, cataloguing, and editorial enterprises into a new medium in order to exploit its radical advantages.’ As they write, Blake’s works are now ‘highly disparate, widely dispersed, and more and more often severely restricted as a result of their value, rarity, and extreme fragility’. Each copy is unique, and the Blake Archive allows you to compare multiple copies of the same work from different collections around the world and explore the relationship between Blake’s poetry and visual designs. The site affirms Blake’s status as a multimedia artist – although the music he composed to accompany his songs is one element of his art that does not, alas, survive.
Whilst obscure manuscripts and archives are particularly well served by e-resources, more canonical texts may be read and researched remotely. Certainly, Romantic poetry is available abundantly online, and the issue is really choosing which edition to use (assuming an edition is given; general-public-aiming websites such as poets.org and Poetry foundation provide nicely readable text but draw from secondary anthologies). While some of the best modern scholarly editions, with all the invaluable notes, are not yet available online (e.g. The Cornell Wordsworth), others are (e.g. Oxford Scholarly Editions; unfortunately we do not have access via KCL). Many of the out-of-copyright editions - either first or other early editions, or later nineteenth-century and twentieth-century editions – are excellent and using them will often have no problematic impact on your argument; just remember to cite them properly, as well as their online source. But if you are focusing closely on a particular text and / or its reception, you will need to look into the publication history and be aware of any quirks of the edition that you are using, e.g. was a nineteenth-century edition bowdlerized? Literature Online (Proquest) is a great starting point for research, and I’ve often come across unfamiliar primary texts as well as secondary materials this way. Whilst it has an explicit ideological agenda, The Online Library of Liberty contains some good editions of many 18thc thinkers, e.g. of Godwin, Smith, Rousseau. For secondary sources, the best place to start for survey articles is Literature Compass (Wiley). It is not comprehensive but browse the period section, and you might get lucky and find out someone has written a survey on your topic. For an overview of recent work in a period, look at the last couple of issues of The Year’s Work in English Studies (Oxford Academic) and ‘Recent Studies in the Nineteenth Century’ and ‘Recent Studies in the Restoration and Eighteenth Century’ in SEL (Johns Hopkins / Project Muse). I am also very interested in word and concept history, for which online corpora offer enormous potential, albeit with limitations. Google Ngrams are very fun to play with and can confirm hunches, though not adequate for a whole scholarly argument. Excellent 18thC scholarship using digital research for concept history includes: Peter de Bolla, The Architecture of Concepts: The Historical Formation of Human Rights, William Tullett, Smell in Eighteenth-Century England: A Social Sense and Wolfram Schmidgen, Exquisite Mixture; The Virtues of Impurity in Early Modern England.
For French online materials, Gallica, the online collection at the Bibliothèque nationale de France, remains a first port of call. You can download, read and search the world’s greatest collection of French-language online materials. Another superlative resource is the ARTL project, https://artfl-project.uchicago.edu/. Here you can find for free a fully searchable digital version of Diderot and D’Alembert’s Encyclopédie; the Dictionnaires d’autrefois, a great resource for seeing the evolution of words, or tracing the history of concepts over time. You can also find fully searchable online editions of Rousseau, Voltaire, Abbé Raynal’s Histoire des deux indes; the journals of Marat; digital editions of 576 French plays and other theatrical texts between 1617 and 1792. In the resource, there is also the Bibliothèque Bleue de Troyes, a collection of mass-produced, inexpensive books, spanning 250 years from the 17th to the 19th century that were sold for pennies by colporteurs. This is a great resource for studying popular culture. Last but not least is the digital edition of Grimm’s Correspondence littéraire. This biweekly newspaper, copied by hand, is indispensable for the study of 18th century cultural affairs, especially art and literary criticism (Diderot’s Salons, essays on art, first appeared here).
The ARTFL project has enough resources to last several lifetimes. Other online resources I also use include:
If you are looking for online versions of great political thinkers of the 18th-century and beyond, there is always the Online Library of Liberty. This is handy for quick access to English translations, although some of the texts are truncated.
Finally if you want to trace the history of a book, in various translations, across time, the Virtual International Authority File is fantastic.
I’m quite haphazard in my methods for locating e-resources. I quite often find that the ‘external links’ section at the end of a Wikipedia entry on an individual thinker leads me to obscure online editions and other online resources that I might otherwise not have found (don’t forget foreign-language Wikipedia: German Wikipedia is often particularly thorough). A lot of publishers have made additional books freely available online during the COVD-19 pandemic: I’ve found this list at publicbooks.org extremely helpful. For stimulating discussions of secondary literature, particularly highlighting recent work and themes of current interest, the review essays in the Historical Journal are often particularly useful, and worth a browse when embarking on a topic; Reviews in History, curated by the Institute of Historical Research, is also worth a look. Having recently completed co-editing a volume in the Cambridge University Press ‘Cambridge Histories’ series, I’m aware that these door-stopper volumes at their best offer an excellent interpretive introduction to a topic: they are mostly available online via the Cambridge Core collection. A lot of my current work focuses on the nexus between religion and political thinking in the long eighteenth century. Some edited volumes that provide a stimulating entry-point into these debates are: Alister Chapman, John Coffey, and Brad S. Gregory, Seeing Things Their Way: Intellectual History and the Return of Religion; James E. Bradley and Dale K. van Kley, Religion and Politics in Enlightenment Europe, and Ira Katznelson and Gareth Stedman Jones, Religion and the Political Imagination. Barbara Taylor’s Eve and the New Jerusalem remains for me a path-breaking classic.
We all love digital resources, some of us so much that we have made our own! In the late 1990s with two other Early Career Researchers I had the idea of a database of all clergy from the Reformation to the mid-19th century in England and Wales based on the records of 50 archives or more to be released on CD-ROM – we wrote a bid, got the money and a few years (I originally typed ‘tears’, which was also true) later launched an online Clergy of the Church of England Database which is still available now some 20 years on. More recently, I’ve been helping lead the Georgian Papers Programme to digitize all the papers of the monarchs from 1714-1837 and make them available online. In between, I’ve served on numerous advisory boards for other digital projects, made frequent use of the resources available to complete articles in 6 months that would have taken three years when I started out in the pre-digital world, and worked to make sure all my PhD students and MA students are fully aware of the extraordinary possibilities such resources have opened up for research.
Working so closely with something, however, and watching other people use the resources you have created in ways you had not anticipated, means that you become as well if not more aware of the shortcomings and problems with sources and how people use them than of the advantages. This is not an anti-digital reflection. Nor is it a comprehensive or scholarly discussion of the issues presented by digital resources, which range so widely in character and quality that each will present its own challenges. Rather, it is a brief reflection on some issues which underpins some advice based on my experiences working in the field for a quarter of a century. I am not a technician, or even a digital humanist. These are the reflections of someone who remains at heart an archival historian, but who is happy to access those archives in any way possible.
My first thought relates to free publicly available resources, such as my own Clergy Database, or Old Bailey Online.
- EVERY PUBLIC DATABASE IS DESIGNED TO ENCOURAGE USERS TO USE IT RATHER THAN SCARE THEM OFF. REMEMBER THIS! AND YOU ARE NOT A TYPICAL MEMBER OF THE PUBLIC!
This is, in the immortal words of Sellars and Yeatman, “A GOOD THING”. But it presents a problem. Whenever I write a book or article, I seek to draw attention to the (ahem) sheer brilliance I have exhibited in solving tricky puzzles in the evidence, or revealing the complexity of an issue. When I build a database, which is partly to be judged by its “impact” on the general public, or simply because what is the point if no one is using it, I am equally at pains to ensure that no one is put off having a go “because it is too difficult for me”. Somewhere prominent on the first page will be a shiny red button akin to the label Alice in Wonderland discovered on the bottle instructing “Drink ME”, labelled something like “Search the Database/Catalogue”. By all means use it to get an initial taste of what lies in the collection. But remember that there are probably all kinds of things you need to know about the database/collection BEFORE you search seriously in order to i) design a good search and ii) interpret the results you get.
I’ll give you one example. In my clerical database, each cleric has a field for date of death. From our records we have no direct record of death – only of records that indicate that someone had died by a particular date. In some cases this will be pretty close to the actual date, and where we have seen a tombstone, may actually record it – but some others may be years after the event. To help the user, a pop up set of codes tells them how closely we think the date in the date of death field relates to the actual date, and there is a long essay in the help section explaining why this is the best we can do; notes flag the issue wherever relevant. But week in week out we get complaints from users telling us we have got the date of death for some cleric wrong. More generally, so complex are some of the evidential issues in our data that there are probably 100,000 words exploring them in our help and guidance section. I suspect fewer than 1 in a 100 of our users has read a word of them. But unless they have, there is a risk that the results of their queries may be fatally flawed. You won’t need to read every word of the guidance pages of an online resource, but checking it out may save you from key errors. It is almost always aimed at academic users who need to understand more about the resource than someone simply looking for a record with their great-great grandfather’s name in it.
So the two key take-aways: i) always look at the guidance, especially if it is labelled ‘Before using the resource …’; ii) use the advanced search engine where one is on offer – it will almost always help you design a better query.
- YOUR RESEARCH IN AN ONLINE DATABASE IS ONLY AS GOOD AS YOUR SEARCH STRATEGY
Most research articles and books will spend some time setting out their methodology and approach in the introduction, as it is a key factor in a reader’s assessment of the validity of the results. For any research involving searching a digital archive, such as a Newspaper or text archive, defining the research question is only the first issue – you also need to think hard about whether the search you make is capable of answering it properly. You made need to explain not just that you ‘searched for x’ in a database, but ‘how you searched for x’. This underlines my advice in section 1: you will only be able to design such a search adequately if you avoid trying to replicate using Google (or even actually finding things via Google!): always use the Advanced Search engine.
Let me give an example: say I want to look for mentions of a ‘Francis Smith’ in a newspaper archive. Simply searching for ‘Francis Smith” may yield far too many results, many of which will relate to thousands of other Francis Smiths. So At the very least I will want to limit my search by restricting it to the time he was alive (unless I am interested in posthumous matters). But in the eighteenth century, that is just the start. Spelling was often inconsistent in the period: so he may well appear as ‘Francis Smithe’, or ‘Francis Smythe’, or ‘Francis Smyth’, or ‘Ffrancis Smith / Smithe / Smythe / Smyth’. His name may be abbreviated: ‘Fran. Smith/Smithe etc’. Might there be a long ‘s’? Could he just appear as ‘Mr Smith of [wherever he lived]’? Or could that be ‘Mr. Smith’? (If you hadn’t noticed, with a full stop after ‘Mr’ – some database search engines will take this in their stride – others won’t).
You will need to give any such search a lot of thought – even more so if you are looking for something more abstract, such as references to an emotion or policy issue. What words will be most commonly associated with it, and how might they be spelt? Time spent on this will pay huge dividends – and if it is not done, you may come a cropper.
So: if your project involves searching a set of data, spend time designing the search, and discuss it with your supervisor.
A classic digital MA project is to search a text archive such as a newspaper archive or poetry database to see how often reference is made to a topic/person and reveal patterns, sometimes completely counterintuitive ones. Some of the best dissertations I have read have done this. But also some of the worst! Where they fall into the latter category, it is sometimes because the research questions have not been designed carefully enough (see 2 above!). But it may also be for two other reasons – which provide my third and fourth points.
- YOU CAN ONLY SEARCH WHAT IS IN A DATABASE: SO CHECK WHAT IS!
To state the obvious, a database of historical sources can only include those things which survive, or to which the builders had access, or chose to include. The first of these is particularly an issue with newspaper archives. You may be astonished in your results that an event which happened in January 1788 seems only to have been mentioned in the newspaper in March 1788. But you might want to check whether the copies for that month actually survive. A database will tell you the titles it includes prominently, but it won’t go out of its way to tell you which individual copies it has. You need to check. The information will be there in the supporting materials, but you would be amazed how few people ever check such resources (see 1. above!) How many copies of the same edition does a database include? (before you equate the numbers of books mentioning something with the number of hits you get). And in databases with innocuous titles such as ‘British Nineteenth Century Periodicals’ you need to check what has not been included. A good example – searching the various British newpaper and periodical archives for the nineteenth century now online to explore newpaper coverage of key religious issues might give a very false impression unless you remember that many key specifically religious newpapers with large circulations are not yet included in the available databases. Obviously in lockdown you can’t address this by checking hardcopy – but you need to be able to note that this is a factor in the results you have obtained. Once you get to the pages telling you what is included, they often give key further information about those items – thus newspaper archives including the London Gazette may need results adjusted if you are looking for evidence of ‘public opinion’, since this was an official mouthpiece.
- HOW GOOD IS THE DATABASE AND ITS OCR / TRANSCRIPTON / METADATA?
We all know how search engines search text archives don’t we? It’s not a magical process. In the early days people retyped all the text, perhaps once or even twice for accuracy. Now, the chances are is that the text is recovered using Optical Character Recognition in which the original is scanned and an automated process generates a text which can be searched from the image of the document. Sometimes the text is effectively indexed to create metadata. All of these approaches can have increasingly impressive results. But they are also often extremely poor, and financial constraints mean we often now post a much rawer text than we might have done in the past, and ask our users to help polish it up. When a search engine searches an OCR-generated text, we might be in effect looking at results that pick up lots of false positives, and perhaps miss 50% of the correct ones, as accuracy rates can be little over 50% in some cases. Where I search newspapers for articles I KNOW are there using their titles they often do not come up in the search. Usually poor OCR is the answer. So don’t give up if the search engine does not yield what you want first time – it is worth trying another approach (see 2. Above) or just browsing through the relevant issues. Also remember human transcribers can make mistakes especially if transcribing manuscript, and may even deliberately sabotage results (who had the bright idea of getting prisoners to help transcribe police records?) And even indexes are only as good as indexers – the first digital editions of some parliamentary records failed to unite records for the Earl of Orford and Sir Robert Walpole, the same person. SO, if the results look wrong
Check results that look wrong by trying a different search strategy.
- WHAT IS LOST IN DIGITIZATION?
Digital archives are wonderful. Increasingly we get razor-sharp images into which we can zoom to microscopic detail, and can manipulate in ways we never could with photos in the past. But we old fogeys who still occasionally wistfully recall six months lockdown in the Colindale Newspaper archive, trawling through endless bulky newspapers searching for elusive small ads, or days in the British Library print room, have some useful experience to bring to the online archive which some digital natives occasionally lack to their cost.
It’s always important to remember that the digital image is of a physical object, even if it is in effect a two-dimensional one like a newspaper page. It has a smell, feel, dimensions, and maybe other bits you can’t see. Whenever I use a digital image, there are therefore a number of things I ask which reflect this.
- How big is the object in real life? If we are discussing its impact or message, this can be quite important, especially for a painting or other image
- For a newspaper article, especially if it is first accessed as just the article itself
- Where did it come in the newspaper?
- What articles were next to it? (You often get the chance to view the whole page/newspaper as well as the article when you get your results – use it)
- For a manuscript letter – is there anything on the back?
- What else is in the same file reference that might be of interest or related?
- What is the file it is in, and what does this tell us about the document?
These questions not least allow us to recapture something of the serendipity of research in a physical archive which can lead to some of the most interesting finds. But they are also very important to assessing significance and context.
There are plenty of other issues with using digital resources. But as stated at the outset, the positives far outweigh the negatives. Hopefully, these reflections will help you experience more of the former than the latter.