Landscape and Local History Group

Bulletin 8: Sources and Research Methods: Some New Approaches

June 2021

Contents

Geographical Information Systems for Historians William Franklin

The Archive at Your Desk: : Researching local and urban landscapes using digital sources Denise McHugh

You can prove anything with statistics’ Ken Sneath

The next issue of the Bulletin will be published towards the end of August and articles on any topic from members will be greatly welcomed. Please let Evelyn Lord have them by 16 August: evelyn.lord@ntlworld.com

All articles are © the Authors, 2021.

Published by the Cambridgeshire Association for Local History

Geographical Information Systems for Historians

William Franklin

Summary

Traditionally, historians have relied on the human eye to interpret old maps and the human hand to produce maps that display complex data in a meaningful way for readers. The latter can take many hours, and for every permutation of data to be displayed, a new map hand drawn. The advent of free-to-use Graphical Information Systems (GIS) now give historians new opportunities both in the interpretation of data and its display. This article, based upon my own use, considers the use of GIS in the study of landscapes.

Maps

While data from many types of documents can be used spatially, it is in the study of the historic landscape through maps that Geographical Information Systems are a particularly useful tool. I make no apology for highlighting this particular use above others in this paper.

There are a variety of types of map that can be used for such study: county, estate, enclosure and tithe, as well as maps made for specific uses, such as the route of railways and turnpike roads or for travelling from place to place. All have advantages and limitations of use. What they depict, how it is shown and, equally importantly, what they omit, depends on the type of map and its purpose. Some maps are linked to a specific process, such as enclosure or tithe commutation, while those grouped under the umbrella of estate maps might have been commissioned for a variety of reasons. These issues and more are discussed in detail in Catherine Delano-Smith and Roger Kain’s English Maps, while Sarah Bendall provides a thorough examination of the estate map.1 It is not the intention to repeat general principles already established by these scholars or to give a history of map making here.

Care must be taken when extracting information from maps as, regardless of how detailed or finely drawn they might be, all maps require interpretation. When originals are unobtainable, copies must be used. Copies need to be treated with caution: negative images will have lost any information contained within colouration and shading; estate copies of enclosure or tithe maps might omit information not relevant to that particular estate; estate copies of earlier maps, even from the same estate, might include changes made in the period since the making of the original but this may not be indicated on the map.2 All maps are somewhat cavalier in their treatment of features beyond the remit of that particular map. For example, a c1812 map of Eltisley possibly produced for an early attempt at enclosure does not show a large, moated area next to the church, now known to exist at that time, and shows another large moat at right angles to its actual position.3 Here the map maker was interested in the fields and furlongs of the parish and not other aspects of the historic landscape such as moats.

The condition of a map can also affect its usefulness. Those held in modern archives are stored in climate-controlled atmospheres and are carefully protected and repaired. But some have evidence of neglect and damage prior to their deposition; they might be dirty, torn, water and mould damaged, or written upon, and for some the damage is irreparable, leading to a serious loss of information. Many suffer from simply being old and faded, while those behind glass are extremely difficult to photograph and it is good quality, high resolution photographs of maps that we want when working with Geographical Information Systems. As we shall see later, we do not necessarily require a whole map to be photographed as one image, but we do require the whole map in as many images as it takes.

Traditionally, local and landscape historians will have viewed every detail of a map taking notes and sometimes laboriously made a tracing. With the advent of photography, photographs could be made of all or part of a map, which alongside notes, (particularly important if the photographs were black and white), were then studied in detail. However, the issue remained about how to place features in the modern landscape especially if that landscape had significantly changed. The historian using a geographical information system will still encounter this difficulty, but to a lesser degree.

Aerial Photographs

Like maps, aerial photographs provide a snapshot of an area of landscape at a particular moment in time. Those taken by Cambridge University, the Royal Airforce and the Luftwaffe may be particularly useful as they pre-date the shift to prairie style farming and more recent large scale housing developments. From such photographs we can trace certain ephemeral features such as field patterns, pastures, parks, and woods.

Geographical Information Systems

Having considered our primary sources, let us now consider the tools: the Geographical Information System (GIS). GIS have been around since the late 1960s and started to become popular tools in the 1990s when they were used by commercial mapping companies, institutions, and government departments such as local public health departments. GIS is computer-based software and as computing has developed, so has the software. As recently as fifteen years ago GIS software was very expensive and well beyond the pocket of most historians. A collaboration of programmers from around the world interested in GIS changed all that. In about 2005 they produced some basic but free software called Quantum GIS (now known as QGIS), which has developed over the last fifteen years into a program that is a match for any commercial system and remains free.4 As free software, it is available for the use of historians and archaeologists to use.

But what is a Geographical Information System? Quite simply, it is a specialised form of database in which the data in the form of shapes, known as points, lines and polygons are usually, although as we shall see not always, drawn on screen and on to a map. Like any traditional database, the attribute data is stored in rows and columns, and provides a structure in which the features in the database can be queried and analysed regarding what and where they are.

GIS and Historical Research

The integration of the disciplines of history and geography has developed since the 1950s in order to incorporate both spatial and temporal elements of history, thus allowing for further research and analysis than might otherwise be possible in a short period of time and providing both historians and geographers with a visual means of analysing complicated information.

Local and landscape historians recognise that all human interaction with others and the environment has inherent spatial characteristics, although not always explicitly addressed in the research. Siebert, studying the spatial history of Tokyo, noted the importance of the geographical component of historical studies:

There is a simple, two-part answer. First, humans are spatial beings. Individuals, groups, and institutions exist and interact in natural human environments that occupy space over time. Human history necessarily includes a spatial dimension, which historians often overlook. Second, geographic information systems are designed to record spatial features and related information, display them, and analyse their conditions and spatial relationships. These capacities enable spatial historical research and extend its analytical reach. 5

GIS therefore offers numerous possibilities for expanding our knowledge and examining more widely some of the issues that intertest us, whether that be at a local level in mapping social status, tenure, and income etc. in a town or village, or mapping taxation data at a district or county level, studying the effects of enclosure or a whole range of other topics.

Things to know about using a Geographical Information System

GIS software is nothing but a blank screen without base maps. The Ordnance Survey makes available a range of free digital maps that can be used in conjunction with most GIS (including QGIS) as a base map.6 In addition, Google Earth (basic maps, terrain maps and Satellite imagery) as well as a range of other regularly used computer mapping services can all be used in QGIS. A word of caution, free maps from the Ordnance Survey use the British National Grid (BNG) as their coordinate system, whereas most free computer mapping systems such as Google Maps use a system known as Pseudo-Mercator. The two projections are different, and the same feature is often many metres apart.7 I do not intend to discuss the differences in coordinate referencing systems here as that would require a very long article. What can be said here is that GIS can add a large range of free maps and satellite applications within them, and if you have bespoke maps or images created in, for example, Google Earth and then exported as a Keyhole Markup Language (.kml) file, these can be easily imported.

A geographical information system, like any project or database, requires a bit of planning, both in terms of what data you want to capture and where you store your data. Every layer and photo should be kept in a specific location, so that when you open the GIS it loads the associated files. Not doing so will cause the program to stop loading until you have relinked each item at start-up, which could prove tedious if you have a lot of layers on your map.

Broadly, GIS software packages handle two types of file: Raster, and Vector. Any paper map you photograph and load into a GIS will be a Raster file. Its properties are fixed, whereas points, lines and polygons as digitally drawn features are Vector files and can be manipulated in many ways. A modern digital map is built up of a series of vectors displayed at the same time, for example, a river will have its banks outlined by lines which are held in one file, whereas the coloured infill of the river will be a polygon and stored in another file. If we are building up a map using historic data it is vector data we are primarily concerned with, because each vector file will contain a table of data behind it, whereas we cannot assign data to a raster file.

Entering data into a GIS is time consuming. With a database, data is entered and analysed relatively quickly (assuming you have reasonable typing skills), but then producing your data manually in ways that visually tell the story or highlight the points you wish to make can be very time consuming. Using a GIS means drawing the features and populating the database at the same time. That is time consuming, whereas analysing the data and producing a wide range of visual data is relatively easy.

Before considering how we might use a GIS there are some things that need to be defined. I have already said that vectors such as points, lines and polygons are the visual aspects of a data table, points allow the user to assign a precise location to something. A point is a dot or other symbol displayed at the vertex at which the lines of longitude (x) and latitude (y) intersect. The user can assign a value to a point, such as a feature (a well, spring, tree etc.), or a place (hamlet, village, town, or city), etc. Lines, sometimes called polylines, can be drawn to outline features or show the presence of something linear, such as a road or hedges. Polygons are a series of connected lines that represent an area such as the outline of a building, the extent of a town or the area of a county.

GIS and historic maps

You are probably by now thinking: "how on earth do I get the historic map I photographed into my GIS?". The simple answer is by digitising it. Within its tools, QGIS (along with most other GIS packages) contains a tool called a georeferencer. In QGIS, when this is selected a second window pops onto the computer screen into which you load a digital photograph of a map. You then must look at your photograph and the modern map side-by-side and identify features common to both. Let us say your picture contains a churchyard and surrounding streets. You select the corners of the churchyard, one at a time, by putting a marker on a point in the photograph and on the corresponding place on the modern map. When done, the program overlays the photograph of the old map on to the modern map.

Once overlaid, it is possible to make the photograph transparent so that you can see the differences between the two maps. If you photographed a very large map in sections, you might have to repeat the process for every picture. It is therefore best, whenever possible, to have one photograph of the whole map you are digitising. But please, make sure you can read every bit of text, because if you cannot, your assigning data to each feature as you digitally trace it will be much slower as you constantly need to refer to more detailed photographs of part of the map or notes you have taken.

Figure 1. The Green at Wicken overlaid onto the modern (colour) digital Openstreet Map

The same procedure is used to digitally overlay historic aerial photographs.

Once your photographs have been digitised it is then possible to think about digitally tracing them. GIS software allows the user to draw points, lines, and polygons (user defined shapes) on screen. As you digitally trace or mark on the map you can add a lot of data all of which is stored in a database, so some thought is required in determining what you wish to capture.

Some examples of how this can be used include:

Point data - A local archaeology group goes fieldwalking and collects pottery brought up by the plough. They use a 10m grid and end up with a bag of pottery for each grid. The number of sherds and types of pottery varies per grid. A grid reference is taken for each grid using a handheld GPS. The pottery is then identified, and a single dot is placed at each location for each type of pottery. The record form they created in the GIS software allows them to enter in the pottery type, the number of sherds, a period (Iron Age, Roman, Saxon etc.) and a date range as well as some notes. They can then produce one or a series of maps to show the distribution of the pottery over the fields walked. Similarly, a historian interested in mills could plot all the known mill sites for a user defined area and record against each point on a map detailed information regarding that site.

A live example of this can be seen in figure 2, which shows the results of field-walking in Eltisley by members of the Cambridge Archaeology Field Group. In this example, only the distribution of Roman pottery is shown. Each red dot represents one 10m grid where Roman pottery was found. It does not show the number of sherds of pottery or the types and dates of the sherds found, although that data is in the underlying database and could easily be displayed.

Figure 2. Example of the display field-walking finds.

Line data - A landscape historian with an interest in ecology or hedgerow dating might digitally trace field hedged boundaries on an early map that still exist. He or she could then add in data regarding the number of species from their field work, the data being displayed as solid, dashed, or dotted lines and in a variety of colours.

Polygon data - as noted above, a polygon is a user defined shape that is drawn over an area - in this case over an old map that you have photographed. For example, you might wish to look at land ownership and distribution at a point in time, or buildings in a settlement at different times. For the former, a polygon is drawn over each strip or enclosure and in the database is stored the name of the occupier of the land, the acreage, and any other data you might want to add. If your map is a pre-enclosure map you can then show the distribution of land for each occupier across the parish (see figure 3, overleaf).

GIS Database

Because the data entered is stored in an underlying database, it can be interrogated outside of the GIS program. I usually take a copy of the data to work on in either a spreadsheet or a database. That way I am assured that I do not accidentally undo any of my hard work. In this way a data set need only be recorded once, but used in two distinctly different ways, one to display spatially and one for interpretation and display statistically and in tabular form. If in deciding the data to be captured for your study linking fields, such as a Parish ID are included,

Figure 3. Soham - 1656 Robert Hamond's lands

Note:- In this example the 1656 map has been digitised and each strip in the open fields overlaid with a polygon. the database of polygons was interrogated to isolate only Mr Robert Hamond's lands. The resulting polygons, each showing a piece of land occupied by Mr Hamond have then been overlaid on the 1886 OS Map. As the database contained every piece of land and every

then it becomes possible to link the data in the database behind the GIS layers and other data or wider analysis.

Importing existing data

There might be cases where you have a spreadsheet full of data that you want to display on a map. With GIS software this is very easy to do providing firstly, you have columns of data for northing and easting, or a column containing the location as a British National Grid reference, and secondly that you save the spreadsheet as a Comma Separated Value (.csv) file. All GIS software packages can read in easting and northing data and most can read in British National Grid references. As the data is read into the GIS program a point (dot) is placed in the correct location on screen.

Lidar and landscape studies

Alongside digital versions of more traditional mapping techniques, there are several new technologies that have emerged and continue to emerge. These are of particular use to the landscape historian and archaeologists, and can be used in conjunction with GIS mapping systems. The most readily available is that of LiDAR. LiDAR or light detection and ranging uses a pulsed laser to calculate an objects variable distance from the earth’s surface and can generate accurate 3D information about the earth’s surface. For many years the Environment Agency has been using airborne LiDAR scanning to look at areas at possible risk of flooding. In recent years the popularity of this data, which can be patchy, has led to the Agency expanding the original remit to include the whole of the UK landmass, although the original completion time for this may well have slipped due to Covid-19.

LiDAR data for those areas completed to date is available freely on the Environment Agency website (https://environment.data.gov.uk/DefraDataDownload/?Mode=Survey) and can be downloaded either as Digital Terrain Model (DTM) images or Digital Surface Model (DSM) images.8 Put simply, the difference between the two is that the Digital Terrain Model has had the tree cover removed, the Digital Surface Model has not. The Environment Agency LiDAR images come in either 25cm, 50cm, 1m or 2m accuracy. The 25cm is the most detailed but has the smallest area covered, while the 2m LiDAR is the least detailed but has the greatest level of cover across the country. The data is provided in tiles and these can be joined togther in the GIS program to give a coverage over a larger area.

When LiDAR data is received the file appears as plain, flat, black and white photograph. To realise the potential of LiDAR data, it has to be manipulated. The most common method of doing this is to use a methodology called Hillshade. This method alters the angle of the sun known as the azimuth.9 In GIS programs this usually involves changing the settings of two fields on screen. The two examples below (figure 4) show the southern half of the parish of Isleham. The left picture is the file as loaded into the GIS program and the right hand picture is the Hillshade model. In this instance the altitude has been set at 45 and the Azimuth at 325 degrees.

Figure 4. Isleham, showing the raw LiDAR image (left) and the Hillshade image (right)

.

Most often a variety of different setting for the azimuth have to be tried to work out the best results for your needs. Hillshade is the most common method of analysing LiDAR data, but there are others, all designed to pick out the detail from the image and the data behind it. Other models for manipulating LiDAR include, the Simple Local Relief Model, Sky View Factor and Multi-Dimensional Hillshade.

Because LiDAR contains height data it can be used to provide topography maps, it particularly useful when looking at topics such as settlement location and drainage. LiDAR is also particularly good at picking up historic field information such as headlands and furlong boundaries, although it is best used in this respect alongside field work as occasionally what might appear as a furlong boundary, is actually the boundary of a field created at enclosure (and sometimes both) and this can only be resolved by going out and looking, and the location of other features such as the roddons of former watercourses. At the beginning of this article, I highlighted the problem of placing features on historical maps in the modern landscape, where that landscape had changed significantly. LiDAR can help in this aspect. Because it shows up features such as furlong boundaries and former stream courses, it can be used to increase the accuracy of the digitisation of a historic field map during the georeferencing process.

By using a combination of the LiDAR digital terrain model with maps reconstructing a historic landscape it is possible to create a 3D map of an area over which the reconstructed landscape can be draped (figure 5.). This is particularly useful when considering soils or geology and historic land use.

Figure 5. 3D map of Kelmarsh Northamptonshire

This map was produced for a talk by David Hall, given at Oxford in 2018 (with permission) and shows the reconstructed medieval field system draped over the LiDAR DTM. The underlying map is the 1880s Ordnance Survey map, and the dots are settlement areas identified from field-walking. The dark grey band is where LiDAR data is absent.

Conclusion

Geographical Information Systems offer the historian the opportunity to collect huge amounts of data in one place with relatively little effort when compared with traditional methods of research. They also allow for rapid production of visual data in many forms or views, that can be easily output and incorporated into books, papers, and teaching materials etc. The recent availability of free data from aerial mapping, LiDAR and satellite imagery combined with GIS mapping offers the historian unique opportunities to present data as never before.

For the landscape historian, the range of data that can be entered into a GIS and the analytical tables that result from it, can be mapped to produce highly accurate plans that can be integrated with modern digital OS mapping, all of which can prove particularly useful in fieldwork and for studying landscape survival.

The archive at your desk: Researching local and urban landscapes using digital sources10

Denise McHugh

Over a year into an undeniably historic pandemic is a timely point at which to consider digital primary sources for research, as the internet has been the only real archive resource available to us during the last fifteen months. Over the last two decades the growth of the digital environment has provided unprecedented ease of access to historical material and is a very real boon to those wishing to engage in researching and understanding past environments and the experiences of them.11 The expansion of digital tools, and increased public participation in these, has produced a vast amount and variety of sources for the historian of place. This wealth of resources offers unparalleled research freedom to the local or landscape historian without the need to even leave their desk. From the digitising of long-appreciated maps and manuscripts to the innovative crowd-resourced digital memory bank or the sharing of private family films and photographs, the range, quantity, quality and nature of historic material has never been greater or more easily accessed. While many of these sources offer more egalitarian access to primary sources, they do tend to privilege material from the modern period, particularly the twentieth century. This paper considers some of the digital resources and platforms which enable the historian of place to access new types of evidence and fresh perspectives. I will first examine how the digital world has enhanced our access to extant historical material before turning to consider the historical evidence collated, or even created, by users of digital platforms, using the market space as a focus.

The apparently infinite resources offered by the internet and web-based activities mean that it is vital for the researcher to first understand and evaluate digital archives and collections and the historical material presented therein. The homogeneity of presentation whereby all types of material are experienced in the same manner through the screen make this understanding even more important as the ease and immediacy of production, the visual convenience and equivalence are deceptive.12 In some ways all sources appear equal on a desktop screen; we cannot tell, for example, the size, weight or smell or feel of a town guide or similar publication.13 This sort of thing does matter as we cannot know if the pages on the screen were designed for a train passenger’s pocket or a rich man’s library table. We should also appreciate the agendas behind the collection, digitisation and presentation of material; the Victorian Census data and Piri Reis Map are good examples of primary sources where their accessibility and formats have been shaped and driven by political, economic and cultural concerns and powers. The processes of digitisation and access privilege certain historical sources, or even parts of sources, and the nature of search engines and digital traffic further influences what and how information is presented to us in our research.14

The digitisation of ‘traditional’ sources, that is to say official and institutional records and accounts, or visual materials that are constructed as socially valuable enough to be housed in ‘brick’ archives, has proven to be both responsive and vulnerable to wider social and economic change. The institutional and establishment embracing of digitisation in the early years of the twenty-first century slowed and shrank significantly in the years of austerity following the global crisis of 2008. Conversely, in this last year, the Covid 19 pandemic seems to have given a new impetus to open online provision and access to records.15

Thinking about types of digital sources

When researching we can consider digital historical sources to fall into three very broad categories; firstly, fairly familiar material that the historian might find in any official, institutional, corporate or private archive that has been digitised rather than ‘born digital’, usually in a process driven by funding, access and public profile considerations.16 This material is often held behind institutional, professional or paywall online barriers. The second category is historical material that has been sourced, collected, or even generated, by the very existence and use of digital technology. This material has been directed and curated within particular parameters and so I think we can also consider a third category – the unintentional digital archive. The material here is generated or shared by people who do not necessarily see themselves as contributing to making an archive but who nevertheless provide the historian with valuable data.

In relation to the first category, these are extant catalogued archive sources that are selected and digitally imaged for web publication and are intended for professional and amateur researchers. These records may or may not be re-catalogued, but the advent of in-text recognition searches and ‘fuzzy logic’ have reduced the reliance on traditional indexes and records while also producing a further narrowing of availability and hierarchy.17 An example useful for landscape historians is the RIBApix website which makes images from the Royal Institute for British Architects (RIBA) collections openly available.18 Other examples such as Ancestry.com or British Library Newspapers offer access to archival material for subscription.19

Within this category of digital resources, are technologies used in partnership with the internet; an excellent example here is Geographic Information System mapping (GIS) which uses space satellite technology for digital mapping. Along with Google Maps and Street View (handy for checking historic buildings and streetscapes from home) GIS and related software, has revolutionised the use of historical maps.20 The ability to overlay accurately maps from different periods enables urban local and landscape historians to build a visual understanding of spatial change.21 The University of Edinburgh’s Mapping Edinburgh’s Social History (MESH) projects provides an exemplar for the usefulness of this from the level of the individual street or close, up to a whole city geography.22 The British Library’s Born-digital Ordnance survey mapping initiative is increasing access to digital local maps and OS base layers.23

One major difference between the digital internet archives and a physical ‘holds’ room is the flexibility which enables themed or related sources and records held in different physical spaces to be brought together for the researcher to access. Good examples of these are the Lincs to the past digital resource which combines material from Lincolnshire Archives, libraries, museums and the Tennyson Research Centre.24 On a larger geographical scale, but more tightly defined by genre, is the Historical Directories of England and Wales site that enables the researcher to read trade directories preserved in multiple archives and libraries, now curated together equally presented, equally accessible and legible as we ‘turn’ the digital pages.25 An initiative with a global reach might be the British Library partnership with Google to make a quarter of a million books freely accessible digitally.26 In terms of achievability, travel time saved and expenses avoided, this level of access and choice is unprecedented.7

The second category of open access digital platforms is material that has been deliberately collected, organised and directed: material such as that generated by the BBC’s WW2 The People’s War project which generated, curated and catalogued testimony.27 Much of this material would never have existed in recorded form without these digital initiatives; the photographs may have mouldered in boxes and the memories lost with those they belonged to. These digital resources can also be more organic in nature where users are enabled to start their own themes and topics of discussion or image sharing thus creating a place-based digital archive of diverse quality but unique perspectives.28 The Cambridgeshire Community Archive Network (CCAN) is an excellent example of this; thirty groups from various places creating a digital local archive from the ‘bottom up’, utilising pictures, recordings, testimony and reports to document everyday life.29 Another example of this type of resource and source one less dependent upon established interest, is the social media message or chat board where users share their local knowledge and memories thus creating an archive of primary material like the Cambridge Memories Facebook page.30 These social media resources are often gated by membership, but are not so directed or targeted in terms of the contributions. The focus and nature of the contributed material tend to shift and grow constantly, reflecting participation and fashion. Materials in both these ‘directed’ and ‘organic’ resources are published and public, for anyone to read, and are thus available to historians but not deliberately created for them. It is probably fair to say that most historians feel much more comfortable with using some form of structured mass history than informal and unstable social media; a nice example is James Greenhalgh’s research on domestic civil defence in the BBC WW2 The People’s War archive.31

In the final category we have the data in public social media platforms such as YouTube which can act as search engine, media player and unintended archive for the historian. These open-access platforms accept uploads of material and contributions from users so are continually expanding and changing creating a dynamic and multi-faceted resource. As well as individual material added by the wider public, many specialist archives and institutions have uploaded resources for public access on these platforms. YouTube hosts over 220,000 British Pathé films or clips including documentaries and newsreels for example, all available to view for free.32 YouTube also contains films from regional film archives such as the South West Film and Television Archive, and, invaluable for the historian of place, amateur and club films uploaded by the public.33 While there is a clear intention by the institutional content contributors to enhance digital traffic and make their archives available, many individual users are motivated by concepts of sharing and entertainment rather than archive formation. This produces a type of ‘mixed economy’ of data which can be uploaded by institutions or individuals, can be private or public in generation, of public, local, sectional or personal interest and, of course of highly variable degrees of reliability. While the platform can remove illegal or inappropriate material there are few content parameters on what is uploaded and available. Some users may be creating or adding their own archives, but the platform is not designed as such; it is the usage of a tiny portion of the data for historical research which defines the historic archive here.

The search algorithms built into these social media platforms and digital archives can assist the local historian by ‘learning’ from each search undertaken and any ‘like’ flags ticked, thus presenting apparently suitable material first but cannot be relied upon. A YouTube search using the keywords ‘old Cambridge’ does produce films of 1950s and 1960s Cambridge but also (predictably) films about the University, current day walking tours of Cambridge, films of Cambridge Ontario, Canada and even two of the Duke of Cambridge on holiday! The local historian’s discernment and skill at sifting remain essential online and the more precise the keywords employed, the better the results. Combining specific street and town names with decades such as ‘1960s’ can produce a wealth of neglected visual material of urban landscapes both extant and lost.34

Another way that a local historian might use social media is to take a proactive approach and use or develop a social media platform such as a local forum or message board to engage with people in a particular place and ask questions using traditional oral history or social science methods.35 These initiatives often partner the knowledge of academics in particular fields with that of the wider public in pursuit of research goals. This can produce a larger sample of material and allows the historian to focus research questions and, of course, this approach requires all the usual scholarly transparency and ethical considerations. The downside of taking this approach rather than an archaeological one, is that the researcher influences the testimony and, in the case of social media, the very presence of the historian ‘in the room’ on the message board may impact who contributes and what they post.36

Examining the urban landscape

Using digital research and online sources enables the historian to develop deep multiple-perspective understandings of past urban landscapes and their evolutions. Online maps, plans, designs, reports and images reveal and explain the built environment and streetscapes, film can give a sense of being in a place but social media allows us to understand the experiences, feelings and responses of those who inhabited, encountered or traversed these spaces.

A good way to explore the different historical understandings and perspectives made possible by different digital platforms and sources is to consider a particular traditional urban space- the market. A fundamental and enduring part of the urban landscape, the market enables us to see the advantages of using a variety of digital sources for place research. We can usually identify the locations of markets from maps, charters and even local regulations, there is a possibility that the local or regional archive may have digitised some of these documents with civic pride and funding motivations. Digitised trade directories can tell us about the market days and market produce for the nineteenth and early twentieth centuries.37 Digitised newspapers from the nineteenth century onwards give information about trade, products and prices and market activity. Visual sources, including old photographs shows us the presentation of the urban landscape, the scale and relationship of surrounding buildings and the variety of surfaces and materials in the built environment. Photographs and films are probably the closest we can get to actually standing in a past space. Unlike other sources though, film has two advantages; it can capture the usage of the urban space, particularly with regard to mobility and also reveals aspects of the urban landscape which were ephemeral and temporal. Very early films such as the Mitchell and Kenyon film of Bedford Street, Plymouth in 1912 provides evidence of the many commercial activities which took place in the street itself from knife-grinding to socialising that may otherwise have gone unrecorded.38 Film shows us not just the transitory shop awnings and hackney cabs we can see in old photographs, but how people moved through and used the space, film captures the passing street trader, the temporary shop stands and transitory displays on pavements; the ebb and flow of urban life. If the film features a clock or a commentary about the time of day or night so much the better.

Having spent more than a decade using film for urban and local history we have found markets and market places, appealed to filmmakers from the beginning and continue to do so.39 The bustle of people, activity, varied materials and later, when technology allowed, sound and colour were all attractive and suitable for the film medium. For any commercial or official film, the end consumer was the public and the market space was presented as their urban locus, stage and a crucial reflection of place and community. A 1902 film of a cattle market in Rossville Street, Derry shows the variety of people, traffic, public transport and interactions at an informal but busy provincial agricultural mart.40 We see little of the built environment but traders set out goods on the pavement and in the kerb while unpenned cattle mingle with farmers, children and people crossing the road.

The quality of film improved dramatically between the two world wars and the introduction of sound allowed narration to direct the viewer’s attention, changing the film experience profoundly and adding another element for the historian. A wartime documentary film of Boston, Lincolnshire, Country Town (1943) produced by the British Council, showed a tightly organised space defined by livestock pens and regulated by ‘government officials’, telling the cinema audience that: ‘Nowadays the market’s more than a barometer of the town’s prosperity, it shows how the whole country’s faring’ in reassuring tones.41 The people seem to be less diverse in dress and gender than in earlier films; they are overwhelmingly male and there are more indications of officialdom in the form of white coats, bowler hats and clipboards. The children are still present though, climbing the rails of pens to pet sheep and cattle, using the urban space for pleasure and recreation.

Amateur filmmakers were also attracted to this urban theme; in 1951 Nottingham and District Film Society made a topical film Old Market Square showing a day in the life of Nottingham’s central space from early morning to 11pm.42 Although gentrified and no longer a market but a transport hub and space for leisure and consumption, the film gives a real insight into the mundane and routine use of this area as well as a good impression of the physical environment. A different type of amateur film, a new-wave influenced art film A Shilling Life (1963) about student life in Cambridge in the 1960s shows students in a café on Cambridge Market Hill with the busy market scene in the background, buying fruit from a stall and a child examining a tortoise for sale on the market.43 Films like these can contain a great deal of unwitting background evidence particularly relating to the mundane and everyday aspects of life but as Ludmilla Jordanova reminds us, such sources were ‘expressly made to be looked at’ and so our gaze is directed and controlled; our perspective pre-ordained.44

In the post-war period films of marketplaces reflected town planning and societal aspirations and many newsreels and later television reports focused on improvements to town centres including market places. Choices made by the filmmakers reflected the nature of the film and societal norms at the time; a film about the redevelopment of Birmingham’s Bull Ring in 1964 noted the accommodation of the ‘traditional’ market stalls ‘which every real housewife would hate to lose entirely’.45 This trend arguably ‘culminated’ in the films of Milton Keynes new town which contrasted the market canvases flapping in the wind outside the vast modernist glass box of its central shopping centre.46

While these different types of digital sources can help us to map the market space over centuries, track the size and scale of plots and buildings, look at visual appearance, traffic, regulation and usage, only testimony can tell us about the user’s sensory and emotional experience of the market. The unique value of social media is to give the historian access to many different user experiences of the same place and time. Open access public social media, particularly chat and message boards, can provide a wealth of memories, local knowledge and even visual primary sources for the local and urban historian. Discussions of remembered shops, streets and even street corners and signage locate the social media posters in time and space and bring past environments to vivid life. Everyday and mundane happenings such as bus routines or adjustments for weather are recalled and clarified, long-run active discussion threads can mediate, moderate and adjust information giving the historian a level of corroborative detail and accuracy that traditional archives might not yield. These discussions often contain details of emotional responses to environments, buildings or even neon signs that we could not ascertain from any other source.47

Social media can record and share experiences, testimonies and collective responses in an unprecedented manner but can also allow us access to particular groups of people who may not be well represented in the traditional archive. Immigrant communities, professionals like nurses and sub-cultures including teddy boys or students, all share memories of places on local and community chat boards offering the historian of place a myriad of experiences. This means that we can access experiences and perspectives that may be unfashionable or under-valued as well as the mundane and unremarkable aspects that may not appear in official records.

Using user-driven social media, particularly local chat forums, for history opens us up to criticisms dealing in nostalgia; historians are inclined to distrust nostalgia, to be disdainful even, using it as a pejorative term.48 Nostalgia suggests that people’s memories of the past are not clear-eyed but rose-tinted, that things were better in the past and testimony asserts life is 'not what it was'. Not all memories are nostalgic however and representations of the past are not all positive. We tend not to collect memories from the young, indeed their experiences are often neglected, and it is only natural that the memories of older members of society are tinged with nostalgia. Rather than a defensive approach however, historians can consider it more positively. Nostalgia can create communities of memory; people motivated by nostalgia can become active members of local history and civic societies. Nadia Atia and Jeremy Davis argue that 'Nostalgia can be a potent form of such subaltern memory'.49 Historians could view nostalgia as a starting point; through nostalgic contributions to websites featuring old photographs or forum discussions about 'the worst school uniform' posters share information about everyday life and experiences that might previously have been considered too mundane to be actively researched or archived. Access to social media can create unintentional archives recording this ephemeral information. Nostalgia 'gives sensory depth to our awareness of the other places, times and possibilities that are at once integral to who we are and definitively alien to us' and offers us a conduit into people's past responses and emotions.50

One way to manage the potential problems inherent in nostalgia or collective memories is to engage the age-old historian’s skill of critical appraisal; we should subject evidence from social media to all of our usual tests. It is important to be able to locate memories in time and space, even if we can only establish a particular decade. We should look for testimony and experiences corroborated by others, we must consider issues of typicality and context. As ever we must triangulate our evidence and be rigorous with our analysis.

To return to the market place; we have seen how we can use a variety of digital sources, particularly visual sources, to ascertain market place location, size, surrounding built environment, trading days, types of goods traded, use of market space, mobility and the behaviour of users. If we want to engage with the sensory or emotional history of this urban space however, then we need the unique insights offered by testimony on social media platforms. Here is ‘Tony1941’ from Derby on a coffee chat forum: ‘In the late 1950's (sic), I use to walk to school through Leicester market. There was a coffee roaster/grinder, just round the corner from the Clock Tower, that you could smell from several hundred meters away (yards in those days!)’.51 On the Sheffield History forum, a lively debate considered which of the city’s two markets; Sheaf or Castle, held more appeal in the 1970s. Following a disagreement about levels of ‘friendliness’, this discussion was reduced to which market was smelliest: ‘The only smell I remember is the awful one of hops floating up from Whitbreads’ brewery, yuck, that was a smell. …I never thought the fish market was particularly bad though but I didn't much like the huge crab boiler, now I couldn't have done that job for any money’.52 These types of testimonies also give lie to the idea that people’s memories are always nostalgic: ‘Henrydog’ remembered ‘I worked up at the Stables Market [Camden] in the mid-Nineties. Some memories: always cold, dragging a tonne of tweed jackets up the cobbles in the rain on a clothes rail with one wheel missing…. sitting on the radiator in a pub in Primrose Hill after work until I could feel my legs again’.53 This is the history of the local and urban landscape as experienced from the inside.

This paper has suggested some ways in which we might think about or classify the vast range of digital resources available to the local urban or landscape historian and how we might navigate them. We can choose to confine ourselves to familiar sources, now in digital form, or use technology to interrogate entirely new forms of evidence. This is not a binary choice either; we can ‘layer up’ different sources engaging a variety of perspectives and historical genres. The historian of place is extremely well equipped to do this, having the advantage of clear research parameters, but this paper has shown how focus on a conceptual urban space or landscape is also helpful. Using innovative material such as unwitting evidence in film or testimony from social media platforms can also open up new pitfalls but a combination of open-minded awareness and established historical practice and rigour will anticipate these.

You can prove anything with statistics’

Ken Sneath

You can prove anything with statistics’ is a phrase often used by those with limited understanding of the subject. But how often do we historians use words like ‘usually’, ‘normally’, ‘often’, ‘many’ or ‘significant’ without sufficient thought as to what precisely we are trying to say?

The subject of this article is quantitative history, which needs to be better understood and put into practice if we are not to just take data on trust.

It is now more than half a century since historians focussed more on the lives of ordinary people as opposed to the ‘great and the good’. The advent of personal computing in the 1980s provided the means to process and analyse large amounts of data. The digitisation of primary sources and the growing availability of historical data on the internet gave a further fillip to the need for a greater concentration on quantitative approaches to historical research. Subjects where quantification is particularly relevant include population, social structure, the relative size of towns, the importance of religion, wealth and standards of living. They are all topics which social, economic and local historians regularly explore.

Some of us naturally lean more towards quantitative rather than qualitative approaches, but we must immediately acknowledge that quantitative evidence alone will almost certainly not provide a complete answer to the questions we are researching. It may well provide part of the answer, but qualitative and quantitative history are bedfellows not enemies. They are complementary in uncovering the past. More recently, other approaches such as cultural history, with its emphasis on actions and symbols, appears to have little need for quantification.54

In our new book The Origins of the Consumer Revolution, Jo Sear and I used both quantitative and qualitative approaches extensively.55 Our study of probate records was based on the largest sample of probate records undertaken in England. We analysed around 9,000 wills, probate inventories and probate accounts as well as manorial court rolls, purchasing records and diaries. We also sought to embrace the approaches of material culture by travelling to around fifty local museums to explore the goods themselves, so bringing an important additional dimension to the topic. This paper explores some of the challenges of quantitative approaches that we experienced in this study.

Our first task in assembling the data was to consider the bias of the sample. Ideally, when assembling a data set, we should select a random sample. A random sample involves taking a sample of cases from the total population so that each case has an equal chance of being selected. Cases selected can be chosen using random sample tables and the results from our sample can be compared to those which would be obtained from sampling the whole population. Few historians enjoy this luxury. Probate records are a good example, for wills and probate inventories did not embrace the whole population but were socially biased documents. They largely relate to the ‘middling sort’, those situated between the landed gentry and the dependent poor.56 Labourers and females were underrepresented as they either had little to bequeath or in the case of women, their property became the possession of their husbands upon marriage. As they were likely to be produced towards the end of life (wills), or after death (inventories), they were also generally representative of older members of society. In addition, Sebastian Keibek and Leigh Shaw-Taylor found that the by-employed were also more likely to be probated than those with only one source of income.57

Probate inventories were sometimes incomplete or lacking in their description of goods. A few were damaged and so vital information could not be recovered. Some inventories were truncated, such as farmers’ inventories which occasionally only recorded details of animals and crops together with their values. There was evidence that inventories were less fully recorded in the eighteenth century, so increases in ownership of goods in this period were probably even higher than the data suggested. Certain goods, particularly low value items, were not valued or included within catch-all phrases such as ‘other lumber’ or ‘hustlement’. Patterns of recording items could change over time, particularly when there was a significant change in the value of those goods. In 15 locations, recorded book ownership declined from 28 per cent in 1630–59 to only 11 per cent after 1750. In view of the substantial increase in books published, this is extremely unlikely to reflect ownership levels and is much more the result of recording practice.

These difficulties ought not to mean that the issue of representativeness should be ignored. In the 1980s, Lorna Weatherill selected 2902 probate inventories from eight areas of England. She chose inventories from the middle year of each decade from 1675 to 1725 and took documents unseen from boxes. Whilst recognising that inventories are socially biased documents, she concluded that her selection methodology resulted in no further bias being detected.58 Mark Overton et al chose inventories from 20 random parishes in both Kent and Cornwall. It appeared that their intention was to choose 200 inventories from each selected parish, but many parishes did not have that number of surviving inventories for the period 1601-1750. In Cornwall, Temple parish had only 12 and Cuby 43. In Kent, three parishes had less than 50.59 The distribution of inventories over time showed that in Kent 471 inventories were selected for the decade 1611-1620 but only 27 for the decade 1741-1750. Their conclusion was that their inventory data set was not representative of the two counties.60 More however can be done to address the problem. Comparison with other records can prove a useful tool. For example, many parishes have baptism records where the occupations of fathers were recorded, prior to the requirement of Rose’s Act of 1812. This data enables the extent to which probate inventories are socially biased to be measured. Comparison can also be made between social structure derived from inventories and estimations of social structure from hearth tax records. These comparisons enable historians to measure the extent to which the social mix of probate records deviate from the whole population.61 Keibeck devised a methodology for neutralising the wealth bias of probate inventories, which he claimed would enable researchers to extend results from just the inventoried population to all households.62

Once the sample has been selected, deciding which software to use to record and manipulate the data is also important. Mark Overton used a custom-written software package called ITEM for analysing inventories.63 We used SPSS statistical analysis software which has many advantages. It is relatively simple to use, and once data has been inputted, the package can calculate all the statistical analysis that you are ever likely to want. Furthermore, limited training will enable you to use the package in elementary approaches upon which you can build your knowledge and expertise.

Unless every item of data in a document is recorded, decisions need to be made as to what data is collected at the outset. This underlines the case for a pilot project which will identify many problems including selection of data. Students are also subject to supervisors who might suddenly come up with a bright idea of data you could collect when you are already significantly along the data collection road. Information about individuals (cases) such as sex, location or items owned such as a clock (variables) is entered into SPSS data grids in the form of a spreadsheet. SPSS requires you to record information for every identified variable for each case. Failure to do so generates an output report which identifies information missing from the system. The report in this example for clock ownership data shows that information about 3973 cases has been selected and the information is complete.

Statistics

Clock

N

Valid

3973

Missing

0

Missing entries may be a simple typing error but even if you have thousands of entries, the missing cell is easy to find, for you need only to sort the cases by the variable and the missing entries will appear at the top of the grid. Once all the data has been entered, and there are no missing entries, the data can be manipulated at the touch of a button. Suppose we want to know the incidence of clock ownership in our sample of 3973 inventories for the years 1660 and 1800. SPSS can provide the answer immediately by using the analyse, descriptive statistics, frequencies function.

Clock

Frequency

Percent

Valid Percent

Cumulative Percent

Valid

0

3108

78.2

78.2

78.2

1

865

21.8

21.8

100.0

Total

3973

100.0

100.0

By using the cross tabulation function it can provide the answer by each location or by sex as in the examples below.

Clock * Location Crosstabulation

Count

Location

Total

Bristol

Mid Essex

Gloucs

Shropshire

Hunts

Yorkshiree

Clock

0

174

181

392

349

926

1086

3108

1

37

32

72

99

189

436

865

Total

211

213

464

448

1115

1522

3973

Clock * sex Crosstabulation

Count

sex

Total

F

M

Clock

0

497

2611

3108

1

66

799

865

Total

563

3410

3973

The data will always reconcile, for there are 3973 cases and the incidence of clock ownership is 865 or 21.8 percent. The example given is incidence of clock ownership (indicated by 1) rather than numbers of clocks owned. Very few inventories recorded more than one clock. However, if the variable was for example ownership of sheep, you may wish to record numbers of sheep rather than incidence of ownership. SPSS enables you to easily calculate average (mean, median or mode) numbers of sheep owned, in other words, the size of flocks.

One of the frustrations of using probate inventories is the grouping together of valuations. One of the very few exceptions to this tendency was clocks, for which a separate valuation was frequently given. Valuations for 605 clocks were recorded in our inventory data set. SPSS instantly calculates median clock values and by period. These values showed little change, being £1.10s in both periods 1690–1719 and 1720–49 and only rising to £1.11s 4d by the second half of the eighteenth century.64 This was consistent with a much smaller sample of clocks recorded in probate inventories in the archdeaconry of Lewes in East Sussex. Glennie and Thrift found that the median value of 52 timepieces in the 1730s was also £1.10s.65

SPSS will also calculate measures of dispersion. Distributions of individual observations can be widely dispersed or closely clustered, but it is possible for different distributions to have the same mean. The mean can be unduly influenced by extreme values. A few unduly large flocks of sheep, or yeomen with excessively high inventory values, can produce misleading mean values. Median values can provide a better perspective, but a solution is to calculate the standard deviation which describes the amount of dispersion of the data around the mean. Historians uncomfortable with mathematical calculations might be daunted by the challenge, but SPSS will calculate the standard variation for you by merely ticking a box. Ticking other boxes will produce rank order dispersal measures such as quartiles or deciles. The coefficient of variation enables the degree to which variables differed from their respective means to be compared. In our inventory sample data for the years 1675-1725, blacksmiths had the least, and widows the greatest degree of variation in their inventory values. (Table 1) It was perhaps not surprising that widows had the highest coefficient of variation, as their deceased husbands represented a wide range of occupational and status groups.66

TABLE 1: INVENTORY VALUES FOR SELECTED STATUS AND OCCUPAIONAL

GROUPS 1675-1725

(Overleaf)

More advanced statistical techniques, such as logistic regression, can be used to predict group membership of a binary outcome. SPSS does all the relevant calculations for you. Overton et al used logistic regression to attempt to predict the ownership of selected consumer goods using inventoried wealth and location as independent variables. They divided material wealth by 50 and used 1 for Kent and 0 for Cornwall.67 For comparison purposes the same approach was used or Huntingdonshire and Yorkshire.

TABLE 2: LOGISTIC REGRESSION RESULTS FOR INCIDENCE OF CLOCK OWNERSHIP

1600-1649

1650-1699

1700-1749

1750-1800

CLOCKS (Kent and Cornwall)

Wealth

1.24

1.36

1.60

N/A

County

8.87

20.89

9.25

N/A

CLOCKS (Huntingdonshire and Yorkshire)

Wealth

-

1.31

1.16

-

County

-

0.20

0.75

1.58

Hunts market towns

-

-

0.44

-

Yorkshire Pontefract

-

-

3.25

4.15

The table above gives the estimated odds ratio (exp(B)), for incidence of clock ownership in Huntingdonshire and Yorkshire and the results of Overton et al’s analysis of Kent and Cornwall. Only estimated odds ratios that are significant (p<0.05) are included in the table. Overton et al found that location was a far greater influence on ownership of goods than wealth. Their most extreme example was window curtains in the period 1650-99. An increase in the wealth of an inventory of £50 would only change the odds of curtains being present by 1.15 times but the inventory was 54.65 times more likely to include window curtains if it related to Kent rather than Cornwall.68 This ‘county effect’ was present throughout their results because consumer goods were far more likely to be found in Kent than in Cornwall. The markedly contrasting counties of Kent and Cornwall produced quite different results to those for Huntingdonshire and Yorkshire.

The use of SPSS does not eliminate the need for historians to make interpretive decisions. Meaningful output requires accurate input. We must ensure that transcription errors are kept to a minimum. I recently reviewed a volume of St Albans wills and inventories for the Economic History Review.69 On opening the book, I immediately scanned the index and my eyes lit up when I saw a clock listed, in the inventory of Elizabeth Cooke dating to 1613, an extremely early date for a domestic clock. Checking the relevant inventory transcription revealed that the ‘clock’ was grouped with clothing and was clearly a cloak, not a clock. In our collection of published inventories used for our book The Origins of the Consumer Revolution, we also had two ‘clocks’ of an even earlier date. The ‘clock’ in Thomas Coll’s inventory (probably 1606) and again in Margaret Pyd’s inventory (1559) were also clearly cloaks.70

We must also define what and how goods are counted. When recording items associated with the new hot drinks of the seventeenth century, tea, coffee or hot chocolate, do we count only items that specify the new drinks in their description such as ‘tea-cups’? How do we deal with imprecise items such as ‘china’? In most cases, we considered incidence of ownership rather than numbers of items owned. However, economic change was reflected in the number of chairs or sheets owned or the size of herds and flocks. This presented problems where certain goods were recorded among generic headings. How do we quantify sheets when most are included among ‘linen’? Crops such as wheat and barley were usually separately identified but, in some inventories, were listed only as ‘crops in the field’. If a probate inventory only recorded ‘sheep’ rather than a specific number, but also gave a valuation, the historian must decide what to enter into the data grid. You could calculate the approximate number of sheep from the valuation, or you could merely indicate that there was more than one animal. Valuations assisted interpretation of inventory data. An inventory for Thomas Ward (1765) listed a ‘horse bridle and saddle’.71 This could be interpreted as just a bridle and a saddle, or the phrase may have meant that there was a horse present as well. The valuation was key to the correct interpretation. These examples show that historians’ interpretive skills are also important in quantitative approaches to the past.

The objective of this article was to encourage those who might be hesitant about grasping the quantitative history nettle to do so. The article provides insights from our recent study using probate records to examine the origins of the consumer revolution. To study this topic, as in many others, historians need a wide range of skills including the ability to transcribe, an understanding of medieval Latin, appreciation of material culture, religious knowledge as well as grappling with the enormous historiography of the topic. Quantitative approaches are also a vital tool. For those who would like to pursue the topic further I would recommend Pat

Hudson’s History by Numbers (London: Hodder ,2000) and Charles Feinstein and Mark Thomas Thomas Making History Count (Cambridge: CUP,2002). They were both written for historians and their examples are all history related so will make us feel at home!

1 C. Delano-Smith & R. J. P. Kain, English Maps: A History, (London, 1999); S. Bendall, Maps, Land and Society: A History, with a Carto-Bibliography, of Cambridgeshire Estate Maps, 1600-1836, (Cambridge, 1992).

2 Examples of all these issues have been found. The loss of colour and shading is especially unfortunate when the full implication of how much information that a map can contains is seen in originals.

3 C.A. 515/P this particular map was copied for the tithe survey in 1841 and used as the basis for the enclosure award map in 1865.

4 The developers are funded by donation and they include some European governments.

5 Loren Siebert, "Using GIS to Document, Visualize, and Interpret Tokyo’s Spatial History,” Social Science History" 24/3 (2000): p.538.

6 The free Ordnance Survey maps for use with GIS can be found through the Open Data Portal, https://www.ordnancesurvey.co.uk/business-government/tools-support/open-data-support

7 The historic maps on the National Library of Scotland website were developed with Google and use the pseudovector referencing system.

8 There is also an option to download the LiDAR images as Cloudpoint files. Point Cloud data contains all the reflected LiDAR information but filtered out into eight bands. Point Cloud data can be used to visualise combinations of these bands. I have not covered that here as it is a more complex option for anyone not used to GIS.

9 The Azimuth is the direction of a celestial body from the observer, expressed as an angular distance from the north or south point of the horizon.

10 With grateful thanks to Dr Lucy Faire for her helpful comments on the first draft; this article arises from our long-term research into British town and city centre usage in the mid-twentieth century.

11 R.J. Morris, ‘Document to database and spreadsheet’ in Research Methods for History, ed. by Simon Gunn and Lucy Faire, 2nd edn (Edinburgh: EUP, 2016), 141.

12 Tim Hitchcock, ‘Confronting the digital: or how academic history writing lost the plot’, Cultural and Social History, 10:1 (2013), 9-23.

13 For a discussion of this see Hieke Huistra and Bram Mellink, ‘Phrasing history: Selecting sources in digital repositories’, Historical Methods: A Journal of Quantitative and Interdisciplinary History, 49:4 (2016), 220-229.

14 Huistra and Mellink.

15 The National Archives news, ‘Covid 19 Archives Fund launched’, 4th December 2020 <https://www.nationalarchives.gov.uk/about/news/covid-19-archives-fund-launched/> [accessed 13 June 2021].

16 Hitchcock, p.10; for a discussion of how this has impacted the digitisation of various nineteenth-century newspapers see Bob Nicholson, ‘The digital turn’, Media History, 19:1 (2013) 59-73 (p.59-60).

17 Andrew Prescott, ‘I'd Rather be a Librarian: A Response to Tim Hitchcock, “Confronting the Digital"’, Cultural and Social History, 11: 3 (2014), 335-341.

18 RIBA, ‘RIBAPIX’ <https://www.architecture.com/image-library/?Action=Cookie> [accessed 12 June, 2021].

19 Ancestry <ancestry.co.uk>; British Newspaper Archive (BNA) <https://www.britishnewspaperarchive.co.uk/ >.

20 Keith D. Lilley, ‘GIS, spatial technologies and digital mapping’ in Gunn and Faire (eds), pp.128-129.

21 Lilley, p.123.

22 ‘Mapping Edinburgh’s social history’, University of Edinburgh <www.ed.ac.uk/history-classics-archaeology/research/research-projects/mapping-edinburgh-s-social-history> [accessed 12 June, 2021]; Richard Rodger and Eric Grosso, ‘Mapping and Managing Edinburgh's Social History (MESH)’, (2014).

23 British Library, ‘Digital mapping’, <https://www.bl.uk/collection-guides/digital-mapping> [accessed 12 June, 2021].

24 Lincs to the past <https://www.lincstothepast.com/> [accessed 12 June, 2021].

25 Historical directories of England and Wales at the University of Leicester ‘Special collections online’ <http://specialcollections.le.ac.uk/digital/collection/p16445coll4> [accessed 12 June, 2021].

26 British Library Press Office, ‘The British Library and Google to make 250,000 books available to all’, 20th June, 2011 <https://www.bl.uk/press-releases/2011/june/the-british-library-and-google-to-make-250000-books-available-to-all#:~:text=The%20British%20Library%20and%20Google%20to%20make%20250%2C000%20books%20available%20to%20all&text=The%20British%20Library%20and%20Google%20today%20announced%20a%20partnership%20to,books%20from%20the%20Library's%20collections. > [accessed 13 June, 2021].

27 BBC WW2 People’s War <http://www.bbc.co.uk/history/ww2peopleswar/> [accessed 12 June, 2021].

28 Michelle T. King, ‘Working with / in the archives’, in Gunn and Faire (eds.), p.23.

29 Cambridgeshire Community Archive Network (CCAN) <https://www.ccan.co.uk/> [accessed 12 June,2021].

30 Facebook, ‘Cambridge memories’, <https://www.facebook.com/groups/1632041636928549/> [accessed 12 June, 2021].

31 James Greenhalgh, ‘The threshold of the state: civil defence, the blackout and the home in Second World War Britain’, Twentieth century British history 28: 2 (2017), 186-208.

32 British Pathé ‘About’, <https://www.youtube.com/c/britishpathe/about> [accessed 12 June, 2021].

33 ‘SWFTA tube’, YouTube, <https://www.youtube.com/user/SWFTATube1> [accessed 12 June, 2021].

34 For example ‘Cambridge UK 60s, 70s and 80s’, YouTube <https://www.youtube.com/watch?v=5JBW88mnE3E> [accessed 12 June, 2021].

35 This method is explained well in Lisa Blenkinsop, ‘The internet: virtual space’, in History Beyond the Text: A Student’s Guide to Approaching Alternative Sources, ed. by Sarah Barber and Corinna M. Peniston-Bird (Abingdon: Routledge, 2009).

36 Blenkinsop, p.132.

37 Historical directories.

3829 SWFTA, ‘Bedford Street and others, Plymouth, 1912’, Plymouth Library Collection @ SWFTA (ID: 73041), YouTube, <https://www.s.com/watch?v=ia62BgOsrjw> [accessed 12 June, 2021].

39 Cambridge City Council, ‘Cambridge Market: A day in the life’, YouTube <https://www.youtube.com/watch?v=ggAUQPeP3dw> [accessed 12 June, 2021].

40 ‘Cattle market in Derry (1902)’, ‘Derry City In 1902 - Ireland: Film of Rossville Street Cattle Market’, Youtube < https://www.youtube.com/watch?v=2QZhXwly4MY> [accessed 12 June, 2021].

41 British Council, ‘Country town 91943’, Youtube <https://www.youtube.com/watch?v=hqq2LxrgMsI> [accessed 12 June, 2021].

42 Media Archive for Central England (MACE), ‘Old Market Square’, <https://www.macearchive.org/films/old-market-square> [accessed 12 June, 2021].

43 East Anglian Film Archive (EAFA), ‘A Shilling life’ (1963) Cambridge, Cambridgeshire, Cat no. 1733 <http://www.eafa.org.uk/catalogue/1733> [accessed 12 June, 2021].

44 Ludmilla Jordanova, ‘Approaching visual materials’ in Faire and Gunn (eds.), p.41.

45‘Bull Ring Centre Opened (1964)’, YouTube <https://www.youtube.com/watch?v=BRRspPXZTuQ> [accessed 12June, 2021].

46 ‘Red Balloon’, Milton Keynes advertisement 1982, YouTube <https://www.youtube.com/watch?v=EOC8wtkzcrk> [accessed 12 June, 2021].

47 Lucy Faire and Denise McHugh, ‘Twelve shades of grey: encountering urban colour in the street in British provincial towns, c. 1945–1970’, Urban History 46:2 (2019), 288-308 (p.305).

48 Svetlana Boym, 'Nostalgia and its discontents', Institute for Advanced Studies in Culture, Hedgehog Review, 9:2 (2007), 7-18, p. 9 <www.iasculture.org> [Accessed 12 June 2021].

49 Nadia Atia and Jeremy Davis, 'Nostalgia and the shapes of history: Editorial', Memory Studies, 3:3 (2010), 181-186, (p.181).

50 Atia and Davis, p.184.

51 ‘Tony1941’, ‘Recollections from Leicester Market’, Coffee Forums, <https://coffeeforums.co.uk/topic/14269-recollections-from-leicester-market/> [accessed 13 June, 2021].

52 ‘Guest Carlie167’, ‘Castle Market vs Sheaf Market’, Sheffield History forum <https://www.sheffieldhistory.co.uk/forums/topic/13944-castle-market-vs-sheaf-market/?tab=comments#comment-118029> [accessed 13 June, 2021].

53 ‘Henrydog’, ‘“All London’s sub-cultures existed side by side”; Your memories of Camden Market’, The Guardian, 7th December 2016 <https://www.theguardian.com/cities/2016/dec/07/londons-subcultures-readers-memories-stories-camden-market-redevelopment> [accessed 13 June, 2021].

54 P. Hudson, History by Numbers (London: Hodder, 2000) p.6.

55 J. Sear and K. Sneath, The Origins of the Consumer Revolution in England: From Brass Pots to Clocks (London: Routledge, 2020)

56 H. French, The Middle Sort of People in Provincial England, 1600–1750 (Oxford: Oxford University Press, 2007), p.266.

57 S.A.J. Keibek and L. Shaw-Taylor, ‘Early modern rural by-employments: a re-examination of the probate inventory evidence’, Agricultural History Review, Volume 61 Issue 2, 2013, 24481.

58 L. Weatherill, Consumer Behaviour and Material Culture in Britain 1660–1760 (London: Routledge, 1996), p.3.

59 M. Overton, J. Whittle, D. Dean and A. Hann, Production and Consumption in English Households 1600–1750 (Abingdon: Routledge, 2004), pp.179-80.

60 Overton et al, Production and consumption pp.30-31.

61 Sear and Sneath Origins of the Consumer Revolution in England

62 S.A.J. Keibek, ‘Correcting the probate inventory record for wealth bias’, Cambridge Working Papers in Economic and Social History Working Paper no. 28, March 2017.

63 Overton et al, Production and consumption in English households p.19.

64 Sear and Sneath Origins of the Consumer Revolution in England p.156.

65 P. Glennie and N. Thrift, Shaping the Day: A History of Timekeeping in England and Wales 1300–1800 (Oxford: Oxford University Press, 2009), p.170.

66 Sear and Sneath, Origins of the Consumer Revolution in England pp.261ff.

67 Overton et al, Production and Consumption pp.145-7.

68 Overton et al., Production and Consumption p.146.

69 K. Sneath ‘Review of P. How and J. Harris, eds., Wills, inventories and probate accounts from Saint Albans 1600–1615’ The Economic History Review Volume 73 Issue 3, 2020, 869-70.

70 M. Reed, Ipswich Probate Inventories 1583–1631 (Woodbridge: Boydell, 1981), p.63; E. Roberts and K. Parker, Southampton Probate Inventories 1447–1575 Volume 1 (Southampton: Southampton University Press, 1992).

71 Probate inventories 1645–1800 (no accession number) York, Borthwick Institute of Historical Research, University of York.

порно зрелые мжм . làm bằng cao đẳng