By Adelheid Heftberger

Freed from the rule of sixteen-seventeen frames per second, free from the limits of time and space, I put together any given points in the universe, no matter where I’ve recorded them. My path leads to the creation of a fresh perspective of the world. I decipher in a new way a world unknown to you. [Dziga Vertov (1984, 1923)]

My focus in this contribution lies primarily on how web-based film studies can be used for analysis and visualisation. Of course, I am aware of recent initiatives to make rare footage and documents available online, (1) or blogs, where film clips are accompanied by insightful articles. (2) At the same time, I am also familiar with crowd sourced (or expert sourced) online projects from which valuable information can be gathered. (3) In my experience, the Internet’s potential lies not only in sharing information, but also in providing technological tools, often developed through scholarly initiatives, for others to use. If Film Studies as a branch of the humanities doesn’t shy away from opening itself up to scientific methods, inspiring processes can be borrowed and adapted from other disciplines.

The Digital Humanities may well be hailed as a new set of academic disciplines that might bring the ‘Two Cultures’ closer. But, in my experience, there is still a lot of work to do, and much more communication will be necessary amongst scholars. The following thoughts and speculations on seven questions of concern to Film Studies in an age of digital humanities are based on my experiences of working in a film archive (The Austrian Film Museum, Vienna) and as a researcher in the interdisciplinary project Digital Formalism. (4)

1. What is ‘traditional’ offline research in Film Studies?

In the beginning was the film. Or, to be more precise, the film event. If we wanted to define ‘traditional’ offline film studies, we would have to go back to a time when scholars would sit in the cinema in the dark, desperately trying to scribble something down in their notebooks. The text written afterwards would be based on one’s memory and the ability to decipher one’s notes. Of course, researchers would visit archives to study film too, but this was often complicated, and limited by financial resources and time. In short, access to many films was difficult, and even if one could gain access, there was rarely a possibility to see each film more than once. Furthermore, film is a visual medium, so how do you quote a film in your text? One must rely either on an abundance of words to describe an image or scene, or on frame enlargements to illustrate the point being made.

Over time, as access to audiovisual material has become more open, and the means and skills necessary to manipulate said material more commonplace, the writings of many film scholars have moved beyond solely verbal to more visual, or multimedia, forms. The limitations of having to use only words to describe a film have always been present in ‘traditional’ Film Studies. I would argue that this problem is particularly prevalent in film history, where the contemporary audience may not always be familiar with the actors or the setting in question. Often, an even better way to explain one’s point, and one unsurprisingly quite frequently used by film scholars, is to produce a video essay. In any case, the discipline of Film Studies, offline or online, is always based on first watching the film, and then reading the related written documents – reviews, biographical information, authors’ notes, charts, and ephemeral literature, like advertisements, leaflets or posters. What has changed for the scholar at first glance is the availability and accessibility of his object of research and, of course, how the results are then published.

2. What’s the material side of it? – Image quality and the artefact

If archives are willing to put their collections online and offer useful tools for navigation, scholars save time and travel expenses. What is lost, though, is obviously the possibility of examining the original artefact, for example, a film print. This is not a film-specific issue, of course, but a general archival one. In reality, archives will rarely grant permission to touch originals. Instead, researchers are usually confronted with reproductions, like copies on microfilm or electronic media. Here, I will only tackle the issues of digitisation and of implementing standards within archives, which are nonetheless very important questions. If a museum or archive wants to open up to scholars and other interested parties, it needs to make sure that the quality of the images is state of the art.

As Walter Benjamin argued (Benjamin 1977, 1936), the technical reproducibility of film is, in contrast to literature or painting, not only an inextricable part of its nature, but an aspect forced upon it by the need for mass distribution. What he couldn’t have known, of course, is that film reproduction processes have always led to a loss of image quality (not just of ‘aura’). It may be a truism to say that digital representation of analogue film is a tricky thing in itself: a video image can be heavily cropped and one can’t always tell from it if its original format was 35mm or 9.5mm. Some film scholars, and film historians in particular, are perfectly aware of this problem. One example of it would be to see where the print is divided into separate reels, which in many cases form individual narrative units. Such things can only be studied on the artefact itself. In the Digital Formalism project, the research team was fully integrated into the archive, so some of that knowledge could be brought to the table. As our experience on that project has also shown, however, film scholars and computer scientists are rarely interested enough in a film’s material qualities and production context. There seems to be a pervasive tendency to regard film as pure content, as if it watching it on VHS tape, DVD, as a video file, or a 35mm print made no difference. Whether this comes from the scholar’s detachment from the processes of film production, or from being used to watching material more often than not in a poor quality state, is hard to say.

What will happen in the future as screens become even smaller than those of our modern laptop-computers or mobile phones is not only a relevant question for scholars or users, but also for the institutions planning online projects. Already now one can assume that online videos are watched mostly on laptops, iPads and iPhones. So how, for example, can necessary contextual information be provided? How can one make sure that metadata is linked to the image? Should the original sound even be added, or is pop-up text an option, if many of the videos will seemingly be watched in public areas without headphones?

The Russian filmmaker Dziga Vertov, attacked by film critics for de-contextualising his documentary material, wasn’t too preoccupied with metadata. He defended himself by explaining how he regarded museums and archives as the correct place to store information about film material in the form of documents for editors, which could serve as a kind of guide to the correct ‘editing route’. As he wrote,

The allegation is false that a fact taken from life, when recorded by the camera loses the right to be called a fact if its name, date, place, and number are not inscribed on the film. Every instant of life shot unstaged, every individual frame shot just as it is in life with a hidden camera, ‘caught unawares’, or by some other analogous technique – represents a fact recorded on film, a film-fact as we call it. […] It would be completely absurd to try to have each individual shot (as a general rule) answer an entire questionnaire: where, when, why, date of birth, family situation, etc. [Vertov 1984, 1926: 57]

We can hardly disagree with the filmmaker’s point of view, and it is not my intention to quote an artist to prove that film clips don’t require metadata and contextualization. However, the quotation does serve to show that, already in the 1920s, film images alone couldn’t be trusted as hard evidence, and that there was an awareness that images without context could be prone to misinterpretation.

3. Where do we study film? – The cinema auditorium versus the laptop

So we don’t need movie theatres any more as the sole place where we can see the objects of our studies. On our laptops we are free to decide when, at what speed, whether to watch from beginning to end, or just jump to the bits we need to see. This process of course had already begun with home video, but now we can skip instantaneously over several minutes (or even hours) of content, or watch different films in multiple windows on a single screen. Already the process of watching several films at the same time might be considered a visualisation, and therefore already inspire our scholarly hypotheses by a simply comparison of the images and montage patterns.

In addition to ordinary DVDs, containing a sole work or film, there are also scholarly editions on the market, which provide us with metadata in the form of annotated commentaries. (5) The Russian Hyperkino edition is one of the most ambitious attempts to publish film classics with ‘footnotes’. (6) Useful Online-Editions of films can also be found: in May 2012, for example, the Austrian Film Museum published the earliest works of Dziga Vertov, providing both the films in digital form and translations of the Russian intertitles in German and English. (7)

A related, and very interesting initiative, this time in online film scholarship, is the Film-Educational Film archive (from the German filmvermittelnder Film) set up by German film and media scholars in recent years. This web-based database collects examples of, links to, and comprehensive information about films made by filmmakers about other films and filmmakers. As Michael Baute and Volker Pantenburg, the scholars involved in this project, have written:

These films […] follow a certain purpose. They want to illustrate, explain, ‘mediate’ what film and cinema are, how they function and what they do. These are films that have learned from other films and want to pass on what they have learned. [Baute and Pantenburg 2007]. (8)

Baute and Pantenburg’s project is part of a much larger, international movement by film scholars and archivists. Often going by different names, (video essays, audiovisual film studies, videographic film studies or DVD essays), these digital works, which are increasingly published online, are both informative and have their own aesthetic value as well.

As the initiatives described above show us, digital media offer us tools to navigate or rearrange images, enabling us to break up the narrative and continuity of a filmic work. Instead of one screen and one timeline, we now have several screens of various sizes – but equal significance – that display a multitude of different images simultaneously. Metadata can be searched using a database. Information relating to shot length or framing can theoretically be computed and visualised, thus forgoing the film image entirely. It’s possible to jump in and out of films, read comments connected to certain scenes or to one aspect at the same time, watch a related video on the same website or navigate through a film like an image-database. Film has become a true ‘time-based art’ for the audience as well as the scholar.

But I would argue that, even if the Internet is a great place to study and analyse film, it should not be the only place where we watch films. Film museums and cinematheques, like the Austrian Film Museum, regard the screening room still as the right and historically correct place for the film event. But it is certainly the case that he arrival of digital technology has now fundamentally challenged both the cinema and the film archive. (9)

4. What are we talking about when we talk about data? – Communication between humanities and computer sciences

The digital archivist Trevor Owens argues that regarding research objects as data is not as alien to humanities scholars as it may seem and suggests a number of ways in which they can use already existing methods:

We can choose to treat data as different kinds of things. First, as constructed things, data are a species of artefact. Second, as authored objects created for particular audiences, data can be interpreted as texts. Third, as computer-processable information, data can be computed in a whole host of ways to generate novel artefacts and texts which are then open to subsequent interpretation and analysis. Which brings us to evidence. Each of these approaches – data as text, artefact, and processable information – allow one to produce or uncover evidence that can support particular claims and arguments. Data is not in and of itself of evidence but a multifaceted object which can be mobilized as evidence in support of an argument (Owens 2012).

I believe that there is still a lack of knowledge among humanities scholars when they talk about ‘data’. ‘If film scholars try to be understood by machines’, was the title of one of the articles published within the Digital Formalism project (Fuxjäger 2009). What we can see at work here is not only a continuation of the well-known stereotype of the old-fashioned film scholar, usually buried in his analogue paper archive, but now trying to communicate with the somehow illiterate computer. It also reveals in one sentence, perhaps more subconsciously than consciously, one of the core problems of interdisciplinary projects.

One of the specialists in this field, Elijah Meeks from Stanford University, quotes a comment made by a graduate student, who felt ‘that oftentimes collaboration with computer scientists felt more like colonization by computer scientists’ (Meeks 2011). He argues that ‘wholesale importation of digital tools, techniques and objects into humanities scholarship tends to foster a situation where rich, sophisticated problems are contracted to fit conveniently into software’ (Meeks 2011). Also the author, visual theorist and artist, Johanna Drucker, sees a lack of, ‘humanities principles developed in hard-fought critical battles of the last decades’, and defines those as:

 the subjectivity of interpretation, theoretical conceptions of texts as events (not things), cross-cultural perspectives that reveal the ideological workings of power, recognition of the fundamentally social nature of knowledge production, an intersubjective, mediated model of knowledge as something constituted, not just transmitted. For too long, the digital humanities, the advanced research arm of humanistic scholarly dialogue with computational methods, has taken its rules and cues from digital exigencies. [Drucker 2009]

Questions of the extent to which art is even allowed to be rationalised, or quantified, are still discussed, and are still problematic, as film and media scholar Barbara Flückiger sets out:

If and how aesthetic objects can or should be measured, is part of the debate on basic principles. Whoever wants to dissect the peculiar haziness of all artistic works into measurable units makes himself easily suspicious of reductionist positivism. Regardless of the long line of attempts, dating back to the turn of the 20^th century, to base philosophical aesthetics on empiric-scientific grounds […], there seems to still be a conflict between empiricism and aesthetics, which is difficult to overcome. [Flückiger 2011: 44]

Consequently, as I have gathered from my own experience, starting a project involving computer scientists, humanities scholars and archivists, requires a clear idea of the different vocabularies and methods, research goals and publication practises. The humanities, in particular, might face the problem of not being able to produce ‘clear’ and ‘computational’ tasks for the computer scientist to solve. Or, as Drucker suggests, it is an intrinsic part of their discipline to be not one-, but multi-dimensional. Although it may sound like common sense, I nonetheless intuit that the different disciplines might not be ‘open’ enough to interoperate as they should and could. Returning to the title of Anton Fuxjäger’s article, I’d like to state that it still isn’t the machines to which we have to make ourselves understood, but our fellow researchers.

5. What can we analyse in film? – Editing as an example

The above doesn’t mean, though, that discrete data can’t be retrieved from literature or filmic works. On the contrary, this mission has been a part of literature or film studies since the 1920s. In the Digital Formalism project we took our starting point from the school of Russian formalism and the texts written by influential scholars like Boris Ėjchenbaum, Jurij Tynjanov or Viktor Šklovskij, to name just some of them. (10) Transcripts and notations by filmmakers can be read, as film scholar Barbara Wurm argues, as practices between personal style and techniques determined by historic-cultural grounds, which disclose a form of ‘tacit knowledge’ (Wurm 2009). Vertov himself wrote:

A kinok who has conceived a film epic or fragment should be able to jot it down with precision so as to give it life on the screen, should favourable technical conditions be present. The most complete scenario cannot, of course, replace these notes, just as a libretto does not replace pantomime, just as literary accounts of Scriabin’s compositions do not convey any notion of his music. To represent a dynamic study on a sheet of paper, we need graphic symbols of movement. [Vertov 1984, 1922: 9] (11)

The film scholar and film historian Yuri Tsivian, founder of Cinemetrics, (12) draws film scholars’ attention to the fact that film is a quantifiable medium. He considers editing the ‘only artistic technique born and developed within the medium itself’, and invites scholars to see the rational side of film studies: ‘We know a good deal about theories of editing (mainly from Soviet montage theories of the twenties), but, ironically, what we normally hear about editing as a practice amounts to a handful of famous examples taken up from these theories.

There is a reason for this. Studying editing is not an easy matter. Editors are like tailors; before they cut, they measure. Footages and meters are staples of cutting-room talk. In this sense editing can be said to be an exact art, and not every student of film history is ready or eager to masquerade as a scientist. In addition, film scholars are more used to working at a desk or in a film viewing hall than they are at an editing table provided with a frame counter’ (Tsivian 2008: 765). Among the Internet platforms dedicated to measuring and analysing film, Cinemetrics is without doubt the most thriving. One can hypothesize that the reason for this lies in the fact that the Internet basically works according to unwritten ‘offline’ rules. The popularity of such websites is due either to their being hosted by renowned scholars, who can draw many students and peers in, or to being linked to well-known and respected universities or institutions.

6. What can we ask from data? – Explaining is analysis

It is not only the case, obviously, that Russian Avant-garde or highly formalized films, like the metric films of the Austrian filmmakers Kurt Kren and Peter Kubelka, can be subjected to film analysis. But what do we want to achieve or unveil in film analysis anyway, be it online or offline, computer aided or manually annotated? With regard to the Digital Formalism project, it could be noted that the technologies of computer science, data mining or visualisation are sometimes at a loss without film scholars and film historians first asking questions.

I do not want to assume that this is the case in every project of this kind, and it’s certainly not the case in the USA or in the UK, where special university courses in the field of Digital Humanities have been started. (13) But some answers to this highly relevant question can be found in the writings of film scholar David Bordwell, who understands analysis also to mean ‘explaining’ to some degree. As he wrote in 2000:

Analysis is a matter of breaking up whole phenomena into relevant parts and showing how they work together. Thus a film historian interested in how a particular studio worked in 1930 will distinguish among the studio’s operations (studio departments, say, or phases of the moviemaking process). An academic film critic will divide a film into parts (scenes, sequences, ‘acts’) to see how the overall architecture works. Explaining something also involves describing it. A film historian trying to explain how a studio functioned in 1930 will describe the work routines; that’s a necessary part of the explanation. An academic film critic will describe a scene in detail, for that’s necessary to understanding why it carries a particular meaning or achieves a particular effect. Analysis and description are rare in ordinary conversation and in film reviewing because of limits of time and space, but also because the film scholar is interested in something that isn’t so pressing for other parties: explanations. [Bordwell 2000]

In addition, it is certainly helpful to consider the following statements by Bordwell as a list of possible questions we can pose to filmic works:

There are distinct types of explanation in film history. A standard list would include: biographical history: focusing on an individual’s life history; industrial or economic history: focusing on business practices; aesthetic history: focusing on film art (form, style, genre); technological history: focusing on the materials and machines of film; social/cultural/political history: focusing on the role of cinema in the larger society. [Bordwell 2008]

7. Online Film Studies in the future?

It may be still true that, in Europe, hermeneutic traditions are very much alive. Therefore it’s generally still hard for humanities and computer sciences to get joint, mutually beneficial projects started. I don’t think that research funding is the main hindrance for not submitting proposals for interdisciplinary projects, however. In the case of Digital Formalism, the project received money because of its interdisciplinary approach. Yet one of the most problematic topics was how and where to publish the results of the research. Perhaps unsurprisingly, it was especially difficult for the computer sciences to find the right venues among their peers. The traditions in the different disciplines still seem firmly established and are not easily bridged, which applies both to the targeted journals as to the publication types (in print or online).

For me, the big potential of web-based film studies lies in the cross-linking of different analytical methods and approaches. To my eyes, simple online presentations of video material already invite us to compare and analyse different content simultaneously rather than just view each video one after another. The opportunity granted by having a large amount of images available enables new hypotheses and research questions, and opens up new ways of processing visual information. The analysis of mass data (macro studies) as well as investigations into a small corpus of works, or even one single work (micro studies), is aided by recent, specially developed software. (14) What we need is both a powerful infrastructure to help us view the videos and generate data, and free software to analyse and visualise it. Also, more attention has to be paid to working with ‘good’ data. This involves the way data is produced (digitisation) as well as the quality of the files available online (compression) and clean metadata. This is where archivists have to offer their expertise, from handling of analogue prints, to scanning, image retouching and database work.

One of the shortcomings of projects with a strong media and Internet focus is that not enough thought is given to the afterlife of the project. Who will host the web address and administer the content, for example? The usual procedure would be to shut down the site, but then the information will be lost to everyone. When we, in Digital Formalism, toyed with the idea of publishing the results of the project on a dedicated website, there was reasonable doubt about the ongoing impact. In the end, a book was published, which shows how strong traditions are still intact in the academic world, making true innovation difficult. Well-established and well-funded institutions, like universities, have more options to explore here than they might think.

But even the best tools provided on the Internet will not solve the problem of posing useful questions. Some of the miscommunication between film scholars and computer scientists in interdisciplinary projects like Digital Formalism may stem from the fact that there isn’t yet enough training in computation and data mining in the humanities. While digital humanities tools have already been developed in literature studies, in film studies there still seems to be little initiative. Furthermore, the roles allocated in interdisciplinary projects still have the humanities scholars posing the questions and the computer scientists answering them. This limits collaboration and, most of all, prevents a real exchange of methods and ideas. Film Studies should be regarded as a collective undertaking per se, based on technical and commercial standards. In addition, we will all have to learn from other disciplines to fully implement statistical approaches, film history, audiovisual techniques, data visualisation and basic informatics fully within Film Studies. Only then will the Web, and other digital infrastructures, become truly useful for Film Studies, providing inspiring, international networks for dissemination, participation and sharing.

Endnotes:

References:

Baute, Michael, Volker Pantenburg, ‘Look at the way he rides with his legs stretched up! Arbeit mit Stills, Arbeit mit Einstellungen, Arbeit mit Sequenzen’, kolik film, Issue 8 (October 2007).

Benjamin, Walter (1936), Das Kunstwerk im Zeitalter seiner technischen Reproduzierbarkeit (Frankfurt am Main: Suhrkamp Verlag, 1977).

Bordwell, David, ‘Studying Cinema’, Observations on Film Art, 2000. Online at: http://www.davidbordwell.net/essays/studying.php.

Bordwell, David, ‘Doing Film History’, ‘http://www.davidbordwell.net/essays/doing.php’, 2008.

Drucker, Johanna, ‘Blind Spots’, The Chronicle of Higher Education (April 3, 2009). Online at: http://chronicle.com/article/Blind-Spots/9348

Ėjchenbaum, Boris (1927), ‘Probleme der Filmstilistik’, In Wolfgang Beilenhoff (ed.), Poėtika Kino. Theorie und Praxis des Films im russischen Formalismus (Frankfurt am Main: Suhrkamp Verlag, 2005), 20–55.

Flückiger, Barbara, ‘Die Vermessung digitaler Erscheinungen’, Zeitschrift für Medienwissenschaften, 2/5 (Berlin: Akademie Verlag 2011), 44–60.

Fuxjäger, Anton, ‘Wenn Filmwissenschaftler versuchen, sich Maschinen verständlich zu machen. Zur (mangelnden) Operationalisierbarkeit des Begriffs ‘Einstellung’ für die Filmanalyse’, In Klemens Gruber, Barbara Wurm, Vera Kropf (eds.), Digital Formalism. Die kalkulierten Bilder des Dziga Vertov (Vienna, Cologne, Weimar: Maske&Kothurn, Issue 55, No. 3, 2009), 115–127.

Meeks, Elijah, ‘Digital Humanities as Thunderdome’, Journal of Digital Humanities, Vol. 1, No. 1 Winter 2011 Online at: http://journalofdigitalhumanities.org/1-1/conversations/digital-humanities-as-thunderdome-by-elijah-meeks/.

Owens, Trevor, ‘Defining Data for Humanists: Text, Artifact, Information or Evidence?’, In Stephen Ramsay, Reading Machine: Toward an Algorithmic Criticism (Champain: University of Illinois Press). Online at: http://digitalhumanitiesnow.org/2012/02/defining-data-for-humanists-text-artifact-information-or-evidence-by-trevor-owens/, February 2012.

Tsivian, Yuri, ‘What is Cinema. An Agnostic Answer’, Critical Inquiry, issue 34 (Chicago, Summer 2008), 754–776.

Vertov, Dziga (1922), ‘WE. VARIANT OF A MANIFESTO’, In Annette Michelson (ed.), Kino-Eye. The Writings of Dziga Vertov (Berkeley and Los Angeles: University of California Press, 1984), 5–9.

Vertov, Dziga (1923), ‘Kinoks: A Revolution’, In Annette Michelson (ed.), Kino-Eye. The Writings of Dziga Vertov (Berkeley and Los Angeles: University of California Press, 1984), 11–21.

Vertov, Dziga (1926), ‘The Same Thing from Different Angles’, In Annette Michelson (ed.), Kino-Eye. The Writings of Dziga Vertov (Berkeley and Los Angeles: University of California Press, 1984), 57.

Wurm, Barbara, ‘Vertov Digital. Numerisch-graphische Verfahren der formalen Analyse‘, In Klemens Gruber, Barbara Wurm, Vera Kropf (eds.), Digital Formalism. Die kalkulierten Bilder des Dziga Vertov (Vienna, Cologne, Weimar: Maske & Kothurn, Issue 55, No. 3, 2009), 15–43.

Copyright: