Losing films

Production still for 4 Devils, from http://www.lost-films.eu

The keen-eyed among you may have noticed that there has been no announcement for the Bioscope Festival of Lost Films. I’ve decided not to pursue the festival any further. It ran for two years, and attracted quite a bit of interest, but I was never quite happy with the way it worked, and then there were the various people who contacted me subsequently who wanted to know where these long-lost films were, and who couldn’t grasp what was going on. And then there was the embarrassment of one of the films that I wrote about not being lost at all – or so I understand (I’ll identify it and report on its happy rediscovery when more concrete information emerges). All in all, it’s time to move on to other things.

But while we’re on the subject of lost films, it would be handy to summarise where things are these days, in particular to reassess the Lost Films site created by the Deutsche Kinemathek. Established in 2007, the ‘Lost Films’ project set out to gather information on films believed to be lost, with archivists and historians invited to add information, documents etc. What started out as a wiki with a strong emphasis on lost German films has grown into a fully-fledged portal, “aimed at collecting and documenting film titles, which are believed or have been declared ‘lost'”. 3,500 titles are now listed, many of them described in detail and illustrated by photographs and documents, and while the emphasis remains with German cinema the range has extended greatly to include lost films from around the world. Not all are silent films, but given the biases of time, the great majority comes from the pre-1930 era.

As an example of what one can now find on the site, take a look at one of the most keenly sought of all lost films, F.W. Murnau’s 4 Devils (US 1927). Starring Janet Gaynor, Charles Morton, Barry Norton and Nancy Drexel, this circus drama was Murnau’s second American film, and it was issued both as a silent as with a Movietone score (not with Murnau’s co-operation) with synchronised sound effects, music and dialogue sequences.

Lost Films gives us a synopsis (including different versions of the ending), cast and credits for both silent and sound versions, release information including length and censorship details, premiere dates, and this commentary by the Deutsche Kinemathek’s Martin Koerber on the film’s ‘lost status:

A print of this was last seen in the 1940s in the Fox warehouse in Los Angeles. According to the files on this title in the Fox papers at UCLA, the print was given to Mary Duncan, lead actress. Legend has it that she either burned it or drowned it in her swimming pool. We can still hope this is an urban legend. No one has traced Mary Duncan’s things, she died only in the 1990s, and there may be heirs. Janet Bergstrom’s video essay about this lost film is a fascinating story, presenting all the surviving stills and sketches and other evidence. This is a bonus on the SUNRISE DVD (or on some of them, anyway.)

Attached to this record are a mouth-watering 172 digitised documents:stills, programmes, advertisements, audience questionnaires (see above), censorship documents, drawings, screenplay (just a sample page, alas), distribution documents and more. The documents come from the Deutsche Kinemathek’s F.W. Murnau collection, and obviosuly not every lost film will be so richly documented, but nevertheless Lost Films has become a rich research resource, not just for specific lost films but for silent film history in general.

There are various ways in which you can search for films – by title, director, country, or year. One needs to be aware of the bias towards German production, and that the numbers of lost films per country is not a measure of overall loss – so, for example, there are currently 1,931 German films listed, but only 42 Japanese, yet the survival rate for Japanese silent film is catastrophic – at least 95% of all production is estimated to be lost (more on such figures in a moment). There are 752 American films listed, and 174 from the UK. Not all are silents, do note. Many of the American and British titles come from the Presumed Lost section of Carl Bennett’s Silent Era site.

Unidentified film no. 145 from Lost Films, a drama of Europeans in Japan, thought to date c.1925, possibly American

Lost Films invites you to register and then contribute information, including adding your own digital materials. It has a helpful links section (though I wish they’d spell my name right), which covers not only lost films but report on films which have been re-discovered. There are different kinds of lost film – those which we know about but no longer exist; those which exist in incomplete form or some non-original version only; those which exist but are hard to find (or see); and those which exist but are effectively lost because they are unidentified. So Lost Films has a section on films that require identification, all illustrated with stills, and a number (chiefly American) happily now identified.

A number of the stills has been contributed by the Nitrate Film Interest Group, and off-shoot of the Association of Moving Image Archivists (AMIA), which is using Flickr to host images of unidentified films in the hope that someone somewhere can identify a name, place or film company logo. There are some highly knowledgeable folk out there who can, and there is pleasure to be had simply in witnessing how sharp-eyed some people are. Do have a go, becuase if you don’t recognise the performers you may be able to spot a location, or a form of dress, and a piece of décor which may provide the vital clue.

There are other lost film resources out there, either aimed at the film history enthusiasts or the archivist looking to identify the mysterious or to help complete a national collection.

  • Moving Image Collections – Lost Films
    The AMIA and Library of Congress-support film archives portal includes a lost films section (all American), and a complementary found films section too.
  • The Silent Era – Presumed Lost
    The aforementioned Silent Era service list hundreds of lost silents from around the world, with credits, references and technical information, plus updated information on films previously listed as lost which survive complete or incomplete somewhere.
  • Recherche des films perdus
    Listing (in French) of films which were listed as lost in 1996 but which have been discovered in archives around Europe by the LUMIERE project (the list is arranged by country of archive, then by country of production).
  • Der Stummfilm in Lateinamerika
    This handy site on Latin America silent cinema (it’s in German) includes a listing of key lost films from Argentina, Brazil, Mexico etc.
  • Wikipedia
    Wikipedia has a selective list of lost films, arranged by decade, which many of the most sought-after titles. It also provides lists of incomplete or partially lost films and rediscovered films. (Much of the information derives from The Silent Era).

  • For the rest, check out Lost Filmslinks page.

To review all of the Bioscope Festival of Lost Films entries, visit the Series section of this site which has all the links – and see if you can guess which one now survives. While you’re there, you can also follow the links for the Lost & Found series, which tells the stories of the discovery of lost film collections.

There are some books on the subject too: Harry Waldman, Missing Reels: Lost Films of American and European Cinema; David Meeker and Allen Eyles, Missing believed Lost: The Great British Film Search; Frank T. Thompson, Lost Films: Important Movies that Disappeared. Note also Anthony Slide’s Nitrate Won’t Wait: a history of film preservation in the United States (2000); David Pierce’s essay, ‘The Legion of the Condemned: Why American Films Perished’, Film History vol. 9 no. 1 (1997), a revised version of which appears in Roger Smither (ed.), This Film is Dangerous: A Celebration of Nitrate Film (2002); and Gian Luca Farinelli and Vitorio Martinelli, ‘The Search for Lost Films’, in Catherine A. Surowiec, The LUMIERE Project: The European Film Archives at the Crossroads (1996).

There are different kinds of lost film, and different degrees of loss. Indeed, some argue that no film can be described as being definitively lost, since by the very nature of the medium multiple copies were produced, so there is always the chance that one may be squirreled away somewhere. Lost films are found in archives, in private collections, on distributors’ shelves, in projection rooms, in people’s basements, attics or garden sheds, in auction sales, and in some romantic cases among a collection of old radios (The 1895 Derby), hoarded by a Swiss school (the Joseph Joye collection), in a Chinese flea market (The Case of Lena Smith), at the bottom of the sea in the wreck of the Lusitania (The Carpet from Bagdad), buried underneath a Canadian swimming pool (the Dawson City collection), or on eBay (Zepped).

Nitrate films in an advanced state of decomposition, from http://www.archives.gov

But just how many silent films are lost? The figure generally bandied about is 80% of all films from the pre-1930 era, this was put together quite a few years ago (I believe it was at the behest of the Federation of Film Archives, or FIAF), and it hasn’t been challenged much since. It may be correct, but it was estimated by matching titles held in national film archives with the titles recorded in national filmographies. But national film archives don’t hold everything (as any proud collector will tell you) and national filmographies have tended until recently to restrict themselves to the fiction film. Nowadays there seems to be a less blinkered approach, but as the Film Foundation says, while “a mere 10 percent of the [fiction] films produced in the United States before 1929 are still in existence … for shorts, documentaries, newsreels, and other independently produced, ‘orphan’ films, there is simply no way of knowing how many have been lost”. Where did that 10% figure come from? The American Film Institute calculated in the mid-1970s that 25% of American silent era films were lost, a much-quoted figure, but as Anthony Slide points out, “such figures, as archivists admit in private, were thought up on the spur of the moment, without statistical information to back them up”.

Solid information for other countries is hard to find, and is certainly not gathered together in any one place that I know of. Here’s a start, however:

  • Australia – 50 out of 250 feature films made between 1906 and 1930 survive in whole or in part (source: National Library of Australia)
  • Brazil – around 10% of silent feature films survives, though many only in a fragmentary state (interview with Carlos Roberto de Sousa)
  • China – 5% of 1,100 productions made 1905-1937 survive (source: Griffithiana no. 54, 1995)
  • Germany – No overall figure, but of 700 films made in 1916 just 60 survive, while one fifth of films made in 1925 are held by the Bundesarchiv (source: Bundesarchiv, ‘Lost Films’)
  • India – of around 1,330 silent fiction films made, thirteen survive, all incomplete (India Profile)
  • Japan – 95-99% of all silents are estimated to be lost (Le Giornate del Cinema Muto catalogue, 2001)
  • Russia – 286 films, out of an estimated 1,716 films produced 1907-1917, or one sixth of production, is preserved by Gosfilmofond (Silent Witnesses: Russian Films 1908-1919, 1989)
  • United Kingdom – “a huge proportion of Britain’s early film heritage is believed lost” (BFI Collecting Policy document)
  • USA – a survival rate of 7-12% for each year of the teens (feature films only), moving to 15-25% for the 1920s (Library of Congress figures from 1993, cited by David Pierce)

(Anyone who has a source for figures from other countries, or better figures, please let me know)

We need an up-to-date international set of figures, one which takes into account the most authoritative filmographic work and which makes it clear the proportion of fiction and non-fiction, professional and amateur film that we should be considering. It would need to make clearer the national differences in survival rates, and what survives in public institutions, commercial concerns, and privately (however much of a guess the latter would have to be). Where certain figures cannot be computed, we need formulae that give an indication. The methodology needs to be made clear. We need the same for the talkie era, for the television era, and now all over again for the digital era, when user-generated content is rewriting the rules for what can be produced. Only then will we know with accuracy just how shamefully we have treated the medium that supposedly is the great mark of the modern age. And we will treasure what survives all the more.