Mapping Victorian Homes and Haunts: A Methodological Introduction

The past 15 years have seen an increased interest in the possible insights that digital mapping offers those studying Victorian literature and culture. While it is true that as far back as 1948, H. C. Darby mapped Thomas Hardy’s fictional Wessex onto historic Dorset, in recent years there has been much more work at the intersection of geography and literary studies.1 Important projects include, but are not limited to, Lancaster University’s ‘Mapping the Lakes’ project, the Stanford Literary Lab’s ‘Mapping Emotions in Literary London’, and the ‘Political Meetings Mapper’ project, which maps the Chartist movement.2 As David Cooper, Christopher Donaldson and Patricia Murrieta-Flores point out, the increasing ubiquity of digital maps in our everyday lives has not only changed our ‘daily spatial practice’, it has also ‘had a demonstrable influence on the way scholars are thinking spatially, and understanding spatial patterns and networks’.3 While more Victorianists are thinking spatially than ever before, there is room for greater understanding of how to use current mapping tools to identify historic buildings with increased precision for our research. Our project, ‘Mapping Victorian Literary Sociability’, brings together collaborators in English (Bourrier), Geography (Jacobson), and Libraries and Cultural Resources (Brosz, Peller, and Reiche) to examine the effect that propinquity – physical proximity leading to frequent interaction – had on the careers of women writers. Funded by a grant for interdisciplinary research from Libraries and Cultural Resources at our university, we began work in November 2018 alongside four research assistants from English and Geography (Anderson, Jarmula, Lapins, and Macaulay). Our aim was to geolocate the addresses of 50 key Victorian writers, illustrators, editors and publishers from birth to death in order to better understand the role of neighbourhood social networks in the careers of women writers (see Figure 1). Home addresses are especially important for women writers, who, as Joanne Shattock points out, would largely have networked through at-homes rather than the male spaces of the university or the club.4 Traditional literary sources such as letters and memoirs give us glimpses of these neighbourhood networks, but these sources do not tell the full story since people were less likely to have an extensive correspondence with their neighbours, and letters and diaries from early in a writer’s career, when she is cementing her connections but before she becomes well known, are less likely to survive. In this short article, we share our methodology and workflow for geocoding Victorian residences using a wide variety of online geospatial resources. A sample of Victorian literary addresses in London. Background map: ‘Guide to the International Exhibition’, Supplement to the Illustrated London News, 24 May 1862. Published by John Dower. Map analysis and rendering in ESRI ArcGIS. Figure exported to 24-bit 300 dpi tiff. As address data from the census has become available alongside a multitude of online mapping platforms, Victorianists have begun to make serendipitous connections between neighbours. Ana Parejo Vadillo traced the interactions amongst bohemian circles, including Karl Marx and his daughter Eleanor, and the poets Mathilde Blind and Rosamund Marriott Watson in St John’s Wood in the 1880s.5 More recently, Mary L. Shannon has uncovered a thriving print industry on Wellington Street, which hosted more than 20 newspapers or periodicals between 1843 and 1853, including the offices of Dickens’s Household Words, the Radical publisher G. W. Reynolds, and the journalist Henry Mayhew.6 Sarah Bilston has argued that the suburbs were a fruitful place of collaboration for women writers later in the nineteenth century; the suburb of Shortlands was home to writers including Dinah Craik, Jane Panton, and later, Enid Blyton.7 Many Victorianists are now using Google Maps as a private way of keeping track of their research, with no plans to publish these maps beyond using them to gain a sense of the movement of a particular writer or the population of a neighbourhood. Here, we share our methodology for locating addresses in both space and time with greater precision, and for recording the uncertainty and ambiguity inherent in this process. We use resources including but not limited to historic paper documents that have been digitized and aligned in geographic space. Because our research involves geolocating residences down to the level of the exact building whenever possible, it is not possible to automate this task in bulk using a script. Indeed, some steps in our methodology are not radically different from traditional archival research; though digitization has made it possible for us to complete our research on Victorian Britain from a distance, and at much greater speed. As Cooper, Donaldson, and Murrieta-Flores point out, following David Berry’s point that the term ‘digital’ may soon be redundant, ‘we are moving to an age in which maps – of all kinds – will invariably be digital’.8 Access to thousands of digital maps can help us surmount the challenges of identifying historic buildings, many of which have been demolished over the last 150 years. The process we describe is recursive and research-intensive. However, through trial and error over the course of our grant, we were able to streamline this process. Peter Peller, Director of Numerical & Spatial Data in Libraries and Cultural Resources, charted a workflow that the co-investigators and research assistants followed together in weekly three-hour lab-sessions.9 The first step in the process is to collect the addresses themselves, which we kept track of in a Google Sheet that was accessible to the whole team simultaneously. In most cases, the most extensive information about the residences of Victorian writers will be the addresses in their correspondence, which can tell us not only about more permanent addresses but also about extended visits to family and friends which would have brought writers into new circles.10 At the moment of writing, digitization of historical directories, maps and spatial gazetteers has been piecemeal, and finding the right information is often a matter of luck; but as digitization continues these may become an increasingly valuable source of information.11 At this stage, ideally we would have a street name and number, or at least the name of the house and the area it is in (e.g. The Elms, Upper Tooting); at times we could only find the neighbourhood or city that a Victorian writer lived in, and in these cases we made spatial approximations and recorded this ambiguity.12 Once we collected the addresses, we consulted a few sources that give the latitude and longitude of a residence with great precision almost immediately. For addresses in London, the first port of call is the Blue Plaque scheme. For addresses outside of London, Open Plaques, a community-based crowd-sourced database of plaques and historic markers around the world, can provide further information.13 Finally, some Victorian homes are historically significant enough to be found in the Historic England database.14 A house name and area (e.g. Otterbourne House, Charlotte Yonge’s home in Otterbourne, Hampshire) is often enough to turn up the coordinates of a listed house. One thing to be aware of is that different countries and mapping agencies use different coordinate reference systems. Because Historic England uses the British National Grid System, and we chose to give all our final geo-coordinates in terms of latitude and longitude, these must be converted using an online tool such as the British Geological Survey Coordinate Converter.15 For the majority of the addresses in our project, which did not have a plaque or a listing, we had to turn to other digital tools to determine the historic location. The most important of these was the side-by-side geo-referenced map viewer developed by the National Library of Scotland (NLS), which shows historic maps alongside contemporary maps from sources including Bing and Open Street Maps.16 Because of the development of their innovative technology, GeoReferencer, the NLS has been able to deliver a vast number of maps online. Currently, there are approximately 220,000 maps in 600 different geo-referenced layers.17 The interface shows all of Great Britain, dynamically revealing more detailed maps as users zoom in. The most detailed information is shown in London at a scale of 1:1250 where individual buildings, pavements, and street furniture are visible. Using the NLS side-by-side viewer has allowed us to trace changes in neighbourhoods over the years. For example, two of Frances Hodgson Burnett’s childhood homes in Manchester were not only demolished but were part of streets that changed shape (see Figure 2). The result is that two of her homes, 141 York Street and 9 St Luke’s Terrace, which were once on different streets, are now both on Cheetham Hill Road (numbered 385 and 364 respectively). Even when streets have not changed, often they have been renamed; as shown in Figure 3, George Eliot’s London lodgings on 21 Cambridge street are now on Kendal Street. National Library of Scotland side-by-side viewer illustrating techniques used to track childhood address changes of Frances Hodgson Burnett. Left panel: Ordnance Survey 25 inch series 1892–1914. Right panel: ESRI world topographic map series. Maps available at https://maps.nls.uk/index.html. Reproduced with the permission of the National Library of Scotland. Licensed under CC by 4.0. Map derived from screen capture of NLS’s website, imported into Corel Draw 9.0 and exported as 24-bit 300 dpi tiff file. George Eliot’s original address at 21 Cambridge Street is now renamed Kendal Street. Left panel: Ordnance Survey 1:1,056 series 1893–1894. Right panel: ESRI world topographic map series. Maps available at https://maps.nls.uk/index.html. Reproduced with the permission of the National Library of Scotland. Licensed under CC by 4.0. Map derived from screen capture of NLS’s website, imported into Corel Draw 9.0 and exported as 24-bit 300 dpi tiff file. Not every Victorian residence can be easily found using the NLS side-by-side viewer. The most significant limitation of the tool for Victorianists is that the maps only date back to 1885. In many cases, such as when we know a group of terrace houses was built in the 1850s and not demolished until the twentieth century, this is enough to pin down a residence with a reasonable degree of accuracy and certainty.18 However, outside of major urban centres we often found we had only a house name and settlement to start with. It is important in these cases to take note of landmarks mentioned by the biographical and historic sources. Fan sites and local history sites, which scholars tend to overlook, can often provide valuable information. For example, early on in our project we were looking for Dinah Craik’s birthplace, Longfield Cottage, which we knew was in Hartshill on Stoke-on-Trent. We started with Google Maps, which gave us a Longfield Road in contemporary Stoke-on-Trent. (We found that streets are often named after demolished historic houses, so this was a promising start.) We took this information back to the NLS side-by-side viewer to find that Longfield Road did not exist in the nineteenth century. In cases like this, other landmarks can be used to locate a historic building. A local Staffordshire history site showed a picture of ‘Longfield Cottage Ice House, Hartshill, Stoke-on-Trent’ and mentioned that the cottage once stood opposite the Noah’s Ark Public House, and that it became the Hartshill Orthopaedic Hospital before it was demolished in the 1960s.19 Google Maps revealed that the Noah’s Ark Public House still exists, and we were able to use those coordinates to navigate to a map on the NLS side-by-side viewer that showed both the pub and the hospital. As a final step, we often compared historic photographs to the contemporary Google Street View to see if the building remained the same. As this example illustrates, the work of geolocating Victorian addresses, especially those of lesser-known historical figures living outside of London, is often recursive and labour intensive. In fact, we would caution Victorianists, who may feel very comfortable working at this level of detail due to their training in close reading and archival research, to be sure that their research question requires this level of precision. If a researcher wanted to map how many publishers moved from Edinburgh to London, it would be wildly inefficient to geocode exact addresses when coordinates in the city centre would suffice. Depending on the research question, geographers can work with anything from millimetres to continental scale; we would suggest that interested Victorianists consult with a colleague in human or historic Geography or a librarian in spatial services in determining the parameters of their project. For our project, which revolves around the question of how near neighbours may have fostered the careers of women writers, this type of accuracy was important, and we made every effort to locate the exact building. When we were particularly stumped by an address, other sources that we turned to included Open Street Maps, Old Maps Online, and the nineteenth-century fire insurance plans recently digitized by the British Library.20 Many of these maps are not yet geo-referenced. To determine the coordinates, we went back to the NLS side-by-side viewer, which we would then double check in Google Maps. We chose Google Maps as our final source of latitude and longitude, despite critiques of the interface as one-dimensional and sanitized, because of its accessibility and its coverage of almost the entire globe.21 Collecting the information for the Victorian addresses in a shared spreadsheet was useful for working with data in a large team since we could all access it simultaneously. However, we cannot trust that resources like Google Sheets will be available for years to come, and we needed a more stable form of representing the data for future users. We also quickly found that there was a good deal of ambiguity in our data; we could not always be certain of the exact dates that an author lived at an address, nor could we always be certain of the exact location of a historic building despite our best efforts. Encoding the data using the standards of the Text Encoding Initiative (TEI) accomplished the dual goal of preserving our data and clearly marking its ambiguities for future researchers.22 The TEI is a set of guidelines and defined elements and attributes used to encode textual information in XML, as well as a community of practice. Ingrid Reiche, Metadata Librarian, developed the template for our TEI gazetteer in consultation with examples from ‘The Map of Early Modern London’ and Melanie Chernyk’s placeography for ‘Last Wyll and Testament of Isabella Whitney’ from the Northeastern University Women Writers Project.23 The main contribution of our project to already-established TEI templates for gazetteers is our encoding of spatial and temporal ambiguity. The TEI’s date element allows us to describe ambiguity in the years an author or publisher inhabited a specific residence. We used the attributes ‘from’ and ‘to’ for dates that were known, and the ‘notBefore’ and ‘notAfter’ attributes when we were uncertain about a date. For example, Figure 4 shows that publisher Robert Chambers lived on West Nicolson Street, Edinburgh, Scotland from 1812 to 1813, and at Hamilton’s Entry, Bristo Street, Edinburgh, from 1814 to 1815. This meant that he could not have lived at Boak Land in Westport, Edinburgh before 1815 or after 1818; similarly, we knew that he was living in the district of Joppa, Edinburgh at a point not before 1815 or after 1816. By using the ‘notBefore’ and ‘notAfter’ attributes we could capture these latter two residences even though the exact dates that Chambers lived at each of them remain unclear. Entry in our TEI Gazetteer for Robert homes on West Street and Bristo Street. we to about the neighbourhood networks that fostered women careers in this project, it was important to find a of encoding how our coordinates We would not to make the in the of that an author and a publisher were neighbours in the when we had only the neighbourhood and not the exact building. In order to spatial precision, we a precision element to each address that the coordinates down the exact historic the or the The precision element a of the in on a to four scale from to least We also our element the latitude and with a from for the exact building address or to for the and for the In the example the first for residence shows we knew he lived in the neighbourhood of Hamilton’s Entry, Bristo Street in Edinburgh, the for his Land residence shows we knew that it was on Street, in the of Westport, in The is and allows us to for example, all the addresses that we have only to the city level if new maps become available that would us to locate the exact the process of a spreadsheet into TEI using – an source tool for and data to spreadsheet that much like – to and the spreadsheet data into elements with attributes and using the to and all the steps this process is At the of our grant, we shared our data in a map on a site and as a but we found that the TEI a more of the ambiguities of our which is an important part of our research. The question of where writers, editors and publishers lived may like a as in an from his work on the of and in the would from a understanding of how these of a a a a a world an our first be to be to which which In which is it that scholars will be at the number of sites and the number of that up as soon as these are further for scholars to with and When we started to the of which neighbourhood and true to up immediately. We began our research with the residences of the Victorian writer Dinah Craik, had One how Dinah Craik, as a a for her first The with and who the work of the of the including and The first thing that we found when we began looking at historic maps was that lived near and as well as Dinah Craik’s family (see Figure As the were and it as though of literary London at their the in the maps suggest that was part of the though he is not named in memoirs or This finding made it to us that the spatial would a valuable to our research. research revealed that where lived as an was also the place where she would have the and the of her future George of this was from the and memoirs that have these suggest a greater role for literary and neighbourhood networks in the careers of women writers than has been Historic maps that as an Dinah lived near not only and at the but also at Cottage, Maps from Ordnance Survey 1:1,056 series 1893–1894. Maps available at https://maps.nls.uk/index.html. Reproduced with the permission of the National Library of Scotland. Licensed under CC by and Google Maps Map derived from screen capture of NLS’s website, and Google Maps imported into Corel Draw 9.0 and exported as 24-bit 300 dpi tiff file. The final of our grant was to share our methodology in a at the of the Victorian in May at the University of with the addresses of Victorian writers, and we them through the process of geolocating these historic could then their addresses to our of Victorian Literary Sociability’, which we on the final of the By our we aim to give more scholars the tools to locate historic buildings that are of interest to their research. We also to Victorianists, many of may be with digital to their spatial thinking and map these in time and space in a more of interest were by the was through the and University a for research project by The W.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.