Of all the ephemeral objects we build for ourselves in the digital world, among the most precious and at the greatest risk of disappearing are databases. Databases are organised and queryable collections of data. The larger ones, along with their cousin the content management system, govern our world and manage who we are, since their very existence decides that we must be categorised in a particular way so that the relationships we have with other people, objects and concepts can be drawn and managed.
Such monster databases as drive Amazon, Facebook, Google and the like are not going to disappear. But smaller databases, created to manage an area of learning, are at risk. They cannot be archived in the way other web objects can be archived, because the web does not expose the underlying structure which enables database queries to be made. There are means to preserve the contents of databases though assorted digital preservation strategies, but they preserve the essential characteristics of the data, not the living, designed entity. Databases need to be hosted, managed, designed, and – to remain meaningful – to be constantly and unendingly updated. A reference book, once published, is what it is, a statement of how things were at a particular point in time. A database has no such fixed status – shark-like it has to keep on swimming, or else it dies.
There have been thousands of databases generated by academic projects over the past two decades. Research funds have been made available to collate information on a subject in the form of a structured list, because you can never go wrong with a list. It’s one of the basic frameworks through which we understand the world, ordering things by value, time, number, category or whatever. Everything finds its meaning and its place in a list, and a database is just a list extended into its most useful form.
These databases are wonderful things. They have been the bedrock of my own researchers, and among my favourites in the film history field are Going to the Show, Filmportal, Catalogue Lumière and the American Silent Feature Film Database. I’ve helped create or sustain a few databases myself, happily all still available online: News on Screen, The International Database of Shakespeare on Film Television and Radio, The London Project.
But once the academic investigation is done, and the research funds are spent, someone has to keep the database going. Usually that’s down to a university, which will keep the database chugging away in its servers, but the money to keep the appearance fresh fresh and the content updated will have gone. Sometimes a volunteer will devotedly be able to do just enough to keep things ticking over, but most – though not all – academic databases eventually fade away, untended and eventually forgotten. They become digital rust.
I say all of this by way of tribute to Karel Dibbets (1947-2017), the great Dutch cinema historian, who died earlier this week. Karel, based at the University of Amsterdam, was scholar of notable accomplishments, writing a history of Dutch cinema to 1940 and editing a volume, Film and the First World War, to which I refer over and over again. He was also the most charming person to know, wise, understanding and always the best company. And his greatest work was a database.
I have written in praise of the website Cinema Context, which Karel established, on several occasions. Indeed my words of praise about the site, which documents film in the Netherlands since 1896 by linking films to the cinemas where they were shown, can be found on its front page, put there not long after I reviewed it on my Bioscope blog in 2007. This is what I wrote:
What is the finest film reference source on the Web, for all film let alone silent film? With all due respect to the Internet Movie Database, I think it is Cinema Context, a Dutch site created by Karel Dibbets and the University of Amsterdam. Describing itself as “an encylopedia of film culture”, the site documents film distribution and exhibition in the Netherlands in 1896. It does so through four data collections, on films, cinemas, people and companies, derived from painstakingly researched data on nearly all films exhibited in Dutch cinemas before 1960. The research team located film programmes from 1896 onwards in each of the major Dutch cities, entering all film titles, names, dates, cinemas etc, and then ingeniously matched this data to the records of these films on the IMDb.
The result is an incomparably rich resource for tracing films, the performers and the producers across time and territories, opening up whole new areas of analysis. Cinema Context also contains comprehensive data from the files of the Netherlands Board of Film Censors 1928-1960. As the site states: “Cinema Context is both an online encyclopaedia and a research tool for the history of Dutch film culture. Not only can you find information here about who, what, where and when: you can also analyse this information and study patterns and networks. Thanks to Cinema Context, we are now able to expose the DNA of Dutch film culture.” Naturally, it is available in both Dutch and English.
This is the new film research. Every nation should have the same.
I still stand by the words, all the more since many extra features have been added since 2007. Cinema Context is dedicated to demonstrating the interconnectedness of film, and by extension is devoted to expressing the principle of meaning through linking, or the logic of context. Film, as with any other subject, only has meaning through its relationships with other subjects. Otherwise all you have a topic that exists only in splendid, self-justifying and ultimately meaningless isolation. It has to be said that too much of film studies exists in just that state of splendid isolation, and the passion that the few, such as Karel Dibbets, have felt for mapping (in every sense) cinema’s relationships has yet to make its way into the film studies mainstream. But it will do, in time, and if the databases are allowed to survive.
Happily Cinema Context continues, maintained now by Professor Julia Noordegraaf at the University of Amsterdam. May it always exist. A book, once published, remains on a library shelf, to be consulted forever more. The same must be promised for databases. More than web texts, which do little more in the communication of knowledge than printed texts had done before, the database is of one the great new academic gifts of our time. We should be doing all that we can to preserve them. They are labours of love, and the labourers are few.
RIP Karel, with many thanks.
- There are various database preservation projects and solutions out there, including LOCKSS (‘Lots of Copies Keeps Stuff Safe’), database-preservation.com and the Open Preservation Foundation
- I wrote two blog posts on Cinema Context – the original, quoted above, in 2007, and a follow-up with more information in 2011
- Dutch film scholar Ivo Blom has written a fine, long tribute to Karel Dibbets on his personal site (with a great photo of Karel where he always liked to be, at a good restaurant table): In memoriam: Karel Dibbets (1947-2017)