Earlier today, the National Gallery launched 200 Paintings for 200 Years – 200 catalogue entries placed online, most for the first time. All in all, this comes to about 2.2 million words; 2,850 images; and references to 8,130 publications, 1,325 archival documents, and 860 exhibitions. As this is something that my team and I have been working on for quite some time, I thought I’d say a bit more about the project.

What’s gone online?

First of all, there are completely new entries for seven paintings (in order of artist’s name):

• Dillian Gordon on The Wilton Diptych (NG4451)
• Letizia Treves on Artemisia Gentileschi’s Self Portrait as Saint Catherine of Alexandria (NG6671)
• Mary McMahon on Lawrence’s Charles William Lambton (‘The Red Boy’) (NG6692)
• Francesca Whitlum-Cooper on Liotard’s The Lavergne Family Breakfast (NG6685)
• Sarah Herring on Manet’s Portrait of Eva Gonzalès (NG3259)
• Emma Capron on Massys’ An Old Woman (‘The Ugly Duchess’) (NG5769)
• Gregory Martin, Nina Cahill and Bart Cornelis on van Dyck’s Equestrian Portrait of Charles I (NG1172) – the first entry in a new catalogue of the Gallery’s Flemish paintings

We have also published a new catalogue of all ten of our Raphaels: Carol Plazzotta and Tom Henry, The Sixteenth Century Italian Paintings, Vol. IV, Raphael (2022).

We have taken entries from all the catalogues published since 1991 which still have entries that have not been superseded (most recent first):

• Susan Foister, The German Paintings before 1800 (2024):
  • Altdorfer, Lucas Cranach the Elder, Dürer, Elsheimer, Hans Holbein the Younger, Huber, Liss, Lochner, the Master of the Life of the Virgin, the Master of the Saint Bartholomew Altarpiece, Maulbertsch, Pacher, the South German school and Wertinger
• Sarah Herring, The Nineteenth Century French Paintings, Vol. I, The Barbizon School (2019):
  • Calame, Corot, Courbet, Daubigny, Diaz de la Peña, Dupré, Huet, François Millet, Jean-Francois Millet and Théodore Rousseau
• Humphrey Wine, The Eighteenth Century French Paintings (2018)
  • Boilly, Boucher, Chardin, Danloux, Jacques-Louis David, Drouais, Fragonard, Lancret, Perroneau, Subleyras, Claude-Joseph Vernet, Vigée Le Brun and Watteau
• Giorgia Mancini and Nicholas Penny, The Sixteenth Century Italian Paintings, Vol. III, Ferrara and Bologna (2016)
  • Costa, Dosso Dossi, Francia, Garofalo, Girolamo da Treviso, Mazzolino, Ortolano and Francesco Zaganelli
• Lorne Campbell, The Sixteenth Century Netherlandish Paintings: With French Paintings before 1600 (2014)
  • Beuckelaer, Bosch, Pieter Bruegel the Elder, Gossaert, Marten van Heemskerck, Catharina van Hemessen, Jean Hey (the Master of Moulins), Massys, the Master of Saint Giles and Provoost 
• Dillian Gordon, The Italian Paintings before 1400 (2011)
  • Barnaba da Modena, Cimabue, Jacopo di Cione, Daddi, Duccio, Giotto, Giovanni da Milano, Lippo di Dalmasio, Ambrogio Lorenzetti, Lorenzo Veneziano, Margarito d’Arezzo and the Master of the Borgo Crucifix
• Nicholas Penny, The Sixteenth Century Italian Paintings, Vol. II, Venice 1540-1600 (2008)
  • Jacopo Bassano, Bordone, Schiavone, Jacopo Tintoretto, Titian and Veronese
• Nicholas Penny, The Sixteenth Century Italian Paintings, Vol. I, Paintings from Bergamo, Brescia and Cremona (2004)
  • Bartolomeo Veneto, Lotto, Marziale, Melone, Moroni, Previtali, Romanino and Savoldo
• Dillian Gordon, The Fifteenth Century Italian Paintings, Vol. I, [1400-1460] (2003)
  • Fra Angelico, Domenico Veneziano, Giovanni di Paolo, Filippo Lippi, Masaccio, Masolino, Pesellino, Pisanello, Sassetta and Uccello
• Humphrey Wine, The Seventeenth Century French Paintings (2001)
  • Philippe de Champaigne, Claude, the Le Nain brothers, Mignard, Patel, Poussin and Valentin de Boulogne
• Judy Egerton, The British Paintings (2000)
  • Constable, Gainsborough, Hogarth, Jones, Lawrence, Reynolds, Sargent, Stubbs, Turner, Wilson and Wright of Derby
• Lorne Campbell, The Fifteenth Century Netherlandish Schools (1998)
  • Bouts, Campin, Gerard David, van Eyck, Geertgen tot Sint Jans, Justus of Ghent, Juande de Flandes, Marmion, the Master of Delft, Memling and van der Weyden
• Neil MacLaren, revised and expanded by Christopher Brown, The Dutch School: 1600-1900 (1991)
  • ter Brugghen, Cuyp, Hals, Hobbema, Honthorst, de Hooch, Kalf, de Keyser, Leyster, Rembrandt, Ruisdael, Saenredam, Steen, Willem van de Velde the Younger and Vermeer

Finally, it’s worth noting that four entries contain updates written by the original author or current Gallery staff (named in the list below) which summarise recent research and supplement the original entry:

• Lorne Campbell: van Eyck, The Arnolfini Portrait (NG186)
• Dillian Gordon: Cimabue, The Virgin and Child with Two Angels (NG6583)
• Dillian Gordon: Margarito d’Arezzo, The Virgin and Child Enthroned, with Narrative Scenes (NG564) 
• Helen Howard, Kristina Mandy, David Peggie and Imogen Tedbury: Francesco Pesellino, Story of David Panels from a Pair of Cassoni (?) (NG6579-NG6580)

You can find links to each catalogue, and the entries that have gone online, at https://www.nationalgallery.org.uk/paintings/catalogues.

Acknowledgements

This has been a long-term project, and many people have contributed to it. The individual authors are listed above; but staff from the Gallery’s Conservation and Scientific Departments have also contributed to the catalogues. In the case of the new, born-digital entries, they are: Paul Ackroyd, Rachel Beard, Rachel Billinge, Lynne Harrison, Catherine Higgitt, Helen Howard, Larry Keith, Marta Melchiorre, Britta New, David Peggie, Marika Spring and Hayley Tomlinson.

Sarah Derry, Laura Lappin and Linda Schofield worked as the project and copy editors on recent entries, whilst proofreading was carried out by Sarah Chatwin, Robert Davies, Karen Francis, Sara Hall, Katie Holyoak and Melanie Marshall. Picture research for external images was by Suzanne Bosman, Maria Ranauro and Becca Thornton; and Rachel Billinge, Rachael Fenton and Claudia Fruianu helped with the supply of internal images. New images were created for the born-digital entries by the Gallery’s Photography & Imaging team, assisted by the Art Handling team.

However, the bulk of the work to get the final desktop publishing and Word files ready to go online was carried out by the Collection Information team’s Online Catalogue Assistants, Annetta Berry, Cynthia So and Hoyee Tse, who added all the tagging required to make the online pages work, and coordinated the images. Carlo Corsato helped with mark-up in a pilot phase, and Katie Holyoak assisted with image coordination.

On the technical side, Jane Hyne provided original files, and SunTec India carried out conversion from QuarkXPress or InDesign to basic TEI-compliant XML. Initial work on XSLT templates to process the initial XML files into something richer and repurposeable was carried out by Kate Byrne, and the bulk of the XSLT templates which help automate the tagging and convert the final XML files into HTML for publication were written by Jeremy Ottevanger of Sesamoid Consulting. The pipeline for ingesting the final XML files into our CIIM middleware, running the final XSL transformation, and making texts and images available to our website, was carried out by Jacob Naylor at Knowledge Integration.

The entries have been incorporated into our new website collections pages: the initial interface designs were made by Numiko, in collaboration with the Gallery’s internal UX/UI team. Over the last year, Caroline Kha and Antonio Sauro worked with Lucinda Blaser, Nejra Hadzimejlic and Jim Gettrup to iterate the designs for integration; Nejra and Jim also built and implemented the final designs and the data integration. (Alongside the new catalogue entries, the redesigned collections pages include new designs for the image viewer, video components, and the display of groups and series of paintings.)

Finally, Charlotte Eaton provided legal support; and Chantel Dendie, Alan Wright, Surbjeet Golam and Mie-Kuen Hong helped with all the financial and personnel-related administration for the project.

How we did it

Jeremy Ottevanger and I described the process at greater length in a paper presented to the 2021 CIDOC conference (actually held in May 2022), ‘Traditional catalogues and digital possibilities: the National Gallery’s online catalogues project’. To summarise:

The first stage is to take the final texts and covert them into a more useful format – in our case the highly-flexible eXtensible Markup Language (XML), using the well-established and incredibly rich structure defined by the Text Encoding Initiative (TEI). Using XML as our base file format lets us extract and reformat parts of the texts as required for various purposes – for example, we have extracted the provenance sections of catalogue entries to put provenances online for all our paintings for the first time (I will publish more on this shortly).

We then run a series of transformations, defined as Extensible Stylesheet Language (XSL) Transformations (XSLT), in conjunction with manual editing, to add all the apparatus that make the texts work as digital resources: we add identifiers to the various sections of text, pages, footnotes, images, bibliographic references, etc., as well as the references to them. Whilst much of this can be automated, a significant amount of work can only be done by hand. Generally, the identifiers follow standard patterns so that they are both human-readable and can easily be parsed during subsequent processing. This is an iterative process, and culminates in the extraction of all the bibliographic items, exhibitions, glossary terms and abbreviations referred to in the text into a single ‘database’ file, where they are collated with the items extracted form other catalogues so that we have a single resource for future reference ( no need for future authors to provide full references again) and to which we could easily add further information – e.g. links to the Gallery’s archive or library catalogues or to online texts.

At the same time, we assemble all the images for the entries, produce standardised derivatives, and name them according to a convention in line with our other identifiers; this makes image management more straightforward.

Finally, texts are ingested into our CIIM middleware, where they go through a final transformation into HTML web-page files and associated metadata. This process brings different sections of the catalogue into each entry: all the bibliographic references, exhibition references, definitions of glossary terms and expansions of abbreviations; but also the text explaining how the entry is arranged, and any appendices which are referred to in the entry. The CIIM also transforms any internal hyperlinks into links to other files or CIIM entities; links the entry to the CIIM records for the relevant publication, artist and painting; and identifies whether this is a new entry or a revised version of an existing one, and assigns persistent identifiers (PIDs) to the specific question and to the latest version of any particular entry.

The Gallery’s website then extracts the relevant HTML data, metadata and images from the CIIM and uses them to create the final web page.

(We have already used the same pipeline to publish the online exhibition catalogue Fruits of the Spirit.)

Editorial policy

In an effort to avoid choosing all the longest entries, our rationale for selecting entries from the published catalogues to be put online was to choose one third from among the Gallery’s most important paintings, and two thirds from paintings that were interesting, but often overlooked.

When converting the previously-published files, we tried to stay as close to the original texts and arrangements as possible, whilst also creating online entries that are self-contained, so that sections like bibliographies, appendices, etc. are brought into the individual entry webpage from elsewhere in the catalogue.

Editorially, we corrected obvious typos, but otherwise left the text unchanged. We also acquired new versions of the various images, which meant that the credit lines had to be updated to match the image suppliers’ current requirements, and have been moved into the relevant image caption. Images, which often fell in the middle of running text in order to sit well on the page, were moved to the next paragraph break after their original position – with the exception of the main image of the subject of the entry, which was moved to the head of the entry. (This explains why the page numbering may indicate empty pages.) Captions which applied to more than one image were sub-divided so that each image has its own caption.

Importantly, we have not updated the texts to reflect current opinion: they reflect the state of knowledge at the time they were written. However, in a few cases, the original author or current Gallery staff have added an ‘update’ section to their entry which summarises recent research and supplements the original entry; these are listed above.

What’s next?

As is always the case in a project of this complexity, there are a few bugs and inconsistencies which we plan to iron out over the next couple of months, mostly to do with formatting.

The most significant of these is a result of the conventions followed in the original texts regarding references, which mean that we have had to assemble bibliographies from various places in the document. The way we currently apply tags to do this means that the addition of hyperlinks to references appears less systematic than it actually is. We plan to improve the handling of references in the near future.

In the longer term: the current images are temporary derivatives. When planning the project, we anticipated being able to link to images using IIIF Image API-formatted URIs. For various reasons, the infrastructure required to exploit the affordances offered by the International Image Interoperability Framework (IIIF) is not yet in place at the Gallery; once it is, we hope to be able to incorporate higher-resolution, zoomable images, and improve the layout of images which are best compared side-by-side.

As mentioned above, it would be good to add links to our library and archive catalogues, or to online texts, to the bibliographic references. We could also look at linking to our online records for exhibitions (more on this shortly, too), and to our online glossary.

Finally, whilst the research project referred to in our earlier presentation, which would have automatically tagged references to paintings, people, places, etc., has taken a different direction, we still hope to be able to do this in due course. (Adding the tags manually would have doubled the time required for us to mark the texts up.)

Updates

• 20 March 2025: acknowledgements updated to provide greater detail about the evolution and implementation of the web design