The feature film “Tout va bien” by Jean-Luc Godard and Jean-Pierre Gorin (1972) shows an unusual interview: during a wild strike in a sausage factory, a reporter, played by Jane Fonda, talks with the women workers there. She asks about their working conditions, housework and other aspects of their working lives. Yet Godard and Gorin do not allow us to share directly in this interview. Although we see the interview, we hear a different sound, namely the inner monologue of a woman standing silently on the side. She hears how testimony is given on her working conditions as well, but although all the facts are right, this interview seems wrong to her. There is something wrong with the tone. The interview will once again affirm all the cliches about women workers. It will have exactly the same effect as all the others – and not convey anything. Although the facts are right, the tonality is wrong. In short: the interview is nothing but kitsch. Despite the best intentions, the reporter Jane Fonda does not succeed in conveying the voice of the workers – specifically because it can be heard in the media.
In an interview in the film “La politique et le bonheur” Godard himself articulates the problem again. Letting the workers speak for themselves, letting them participate in the production of the film, does not necessarily mean letting them speak up. A worker, whether man or woman, speaking for themselves is perceived in the media as not being particularly bright, in any case as a relatively pitiable specimen. They become objects of a voyeuristic gaze that is interested in “authenticity”, but not in change. No matter what they say, their roles are already predetermined: they are the ones affected, and as such, they need not be taken seriously. They can talk as much as they want, but somehow the sound is missing. Godard and Gorin are aware of this dilemma: in “Tout va bien” we hear the voice of a worker – but only in the form of her mute thoughts.
However, if interviews, as Godard and Gorin maintain, are futile under certain circumstances, this has disastrous consequences for the documentary form. No one can report on an event more credibly than witnesses who have seen it with their own eyes or heard it with their own ears. This is why eyewitness reports and interviews with experts or contemporary witnesses are among the most widespread methods intended to justify the documentary claim to rendering facts.
Nevertheless, a distrust of witnesses is chronic. For a witness can indeed speak the truth, but he or she can also lie. To insure against these imponderabilities, legal systems have again and again invented new systems of testing. An old Roman rule of law stated: testis unus, testis nullus – one witness is no witness. At least two witnesses must give concurrent statements to grant credibility to a report. Documentary forms copy these kinds of rules through the juxtaposition of multiple witness statements and perspectives to create “objectivity” with this procedure – a concept that is chronically contentious in documentary film theory. Yet even this method can only create probability, rather than certainty. For what if both witnesses are lying? Precisely the figure of the witness, through whom the viewer wants to secure the veracity of what is shown, is inherently insecure. She or he balances along their memory as on a railing that only exists in the imagination.
For this reason, the interview has always been under suspicion. According to Michel Foucault, techniques like these are historically derived from dubious examinations such as trial by ordeal or confession. According to Foucault, the production of truth was much more ruthless in the Middle Ages than today. Someone was thrown into the water – if he floated, he was right, if not, then he had lied. In technologies of truth like the trial by ordeal or the duel, chance or higher powers determine the result. More recent technologies of truth, such as the scientific experiment, religious ones like confession or legal ones like torture or a guilty plea, have more complex rules, but according to Foucault they still function according to the same principle. Documentary techniques such as witness testimony or the interview also belong to this tradition. They are based on historical, legal or journalistic technologies of truth.
Making witnesses the guarantee of documentary truth thus means taking the risk of an actually bottomless trust. For do witnesses really establish an unfiltered access to reality? Or is the witness not fundamentally opaque – subjectively colored, marked by interests, seduced by images of language, enamored of being right? Is what is witnessed not perhaps more reality as it should have been, rather than reality as it really was? Can the testimony not be torn out of the context by documentary articulations, distorted or falsely relayed? For example, without sound, like the interview with the workers from “Tout va bien”?
That what is said, is not always heard, is also maintained by the feminist literature scholar Gayatri Chakravorty Spivak, because certain groups of the population are principally excluded from social articulation. Although they can speak, we simply do not hear them. The question that Spivak therefore raises is: “Can subalterns speak?” With subaltern she means especially women in situations equally characterized by colonialism and patriarchy. And the apodictic answer to her question is: no. As an example for the muteness of subalterns she draws on files from the colonial archive, which are supposed to provide evidence that Indian women during the British colonial era “voluntarily” submitted to the cruel custom of widow burning. For Spivak, however, these files do not prove the will of the women in question, but only the impossibility of expressing their will at all. For if a woman wanted to oppose the value of the colonial masters, all that was left to her was to refer to the values of the local patriarchy, which ordered her to “voluntarily” burn herself. Her own interests were thus inarticulable in both value systems, and there was no other. Spivak’s conclusion: the subaltern does not speak, she is not able to bear witness to herself. Interviews with her are therefore futile.
The situation is similar to that in “Tout va bien”. Even if the workers bear witness, they are not understood, for similar to the situation of the subaltern women, they are only considered capable of conveying “concrete experience”. Yet asking someone about their “concrete experience” implicitly presupposes that this person has nothing else at their disposal, that these experiences are raw and unreflected, and that they must be explained both to the person and to the audience. The concept of “concrete experience” dictates a certain form of a hierarchical division of labor: between those who experience something and others who are able to understand and interpret this experience. Their alleged authenticity has immediate political effects: the very voices that sound entirely “authentic” are structurally incapacitated. These problems have also been described again and again in feminist film and science critique since the 1970s. In feminist film, the oppression of women in a patriarchal society cannot simply be captured or recorded, because the so-called simple recordings are already part of the problem, argued Claire Johnson as early as 1975.
Yet structural doubt of testimony also turns us into autists. If witnessing is futile in certain cases, this not only affects documentary forms that rely on witnesses to convey certain events as well as possible. The problem lies much deeper. Witnesses not only report on the world, but also first produce it in a social and political sense. If we want to overcome the solipsism of our individual experience, we cannot do without witnessing. If we want to know what is happening in a far-away war, we usually have to rely on witnesses. On the whole, discerning a witness represents an attempt to open oneself up to the experiences of others. It is a step in the direction of coping with the paradoxical task that Wittgenstein once so vividly described: feeling the pain in the body of the other. But testimony also assumes an important role beyond individual experience. Hannah Arendt regarded “factual truth”, which testimony is supposed to give evidence of, for the condition of the societal per se. According to Arendt, this is nothing less than the “ground on which we stand and the sky that stretches out above us”. In short, testimony is unreliable and uncertain – but it is indispensable.
Testimony is often prevented or – as in “Tout va bien” – not heard at all. On the other hand, it can also bear witness to what is not allowed to be said within these power relations. “Even fatally scratched, a small rectangle of 35 millimeters is capable of saving the honor of the whole of reality,” wrote Jean-Luc Godard. It can express the unimaginable, that which has been silenced, the unknown, the saving, and even what is monstrous – and thus create the possibility of change.
And contrary to all probability, even in official historiography, in the hegemonic media, archives, discourses and histories, testimonies may be found that really should not even exist. They are not the rule, of course. But to claim, as Spivak does, that they are not even possible, means obliterating them utterly from history – and gagging even the mute thoughts of the worker from “Tout va bien”.
“Testimony contains a lacuna.” – This is the conclusion Giorgio Agamben reaches in investigating the testimonies of survivors of the Shoah. These statements represent an absolute borderline case of testimony. They are rooted in an insurmountable paradox: those who survived the concentration camp and bear witness do not see themselves as authorized to do so. Primo Levi thus writes that it is not the survivors who are the real witnesses, but the dead. Consequently, only those can bear witness to the annihilation, who fell victim to it. But they no longer speak.
This constitutive contradiction is the leitmotif for Agamben’s reflections on the role of the witnesses. Accordingly, “the value of testimony lies essentially in what it lacks; at its center it contains something that cannot be borne witness to and that discharges the survivors of authority”. Testimony is therefore simultaneously necessary and impossible, it witnesses its own impossibility. According to Agamben, this impossibility is embodied in the figure of the Muselmann. The Muselmann is a prisoner who has lost his will to live and merely vegetates at the verge of death. He is no longer capable of speaking – and what the survivor can testify on his behalf is only mediated and incomplete.
Agamben comes to the conclusion that the splitting of the testimony cannot be annulled. The witness becomes a witness to the extent that he reports on the impossibility of the testimony. What these witnesses express is the dilemma that forms their task at the same time. The testimony is not only impossible, it is also indispensable at the same time. Even if the Muselmann does not speak, someone must speak for him.
Yet as though that were not enough, Agamben’s book ends with the most surprising of all testimonies, namely testimonies that should not really even exist. One of them begins with the words, “I am a Muselmann”. Those who bear witness in these testimonies are the ones of whom it is said that they are not able to. As though by a miracle, though, a few of them returned to life and tell of their existence as “Muselmänner”. At the same time, these improbable reports bear witness to the intermittent failure of fascist violence as well as to its extreme effectiveness. As absolute exceptions they confirm the rule. According to Agamben, they do not cancel out the paradox, but “fully” verify it. Although they are actually not possible, they exist.
Another case, this time in relation to images, is recounted by the art historian Georges Didi-Huberman. In his text “Images in Spite of Everything” (“Images malgré tout”) he describes pictures that are actually improbable: the only four pictures that have been passed on that were made by prisoners of the concentration camp Auschwitz. There is no lack of other pictures of Auschwitz. Although the camp was a territory in which uncontrolled photography was strictly prohibited, there was not only one, but two dark rooms there. The largest portion of the photos made there were for police purposes. Some 40,000 pictures survived the destruction of the archives before the liberation of the camp.
The only four remaining pictures made by prisoners, on the other hand, belong to a completely different context. They were made by a member of the so-called special commando to provide visual proof of the mass murders. In order for the photos to be taken, the Polish resistance first smuggled a camera into the camp. Then a complicated plan was devised to distract the SS guards. Only then did a man named Alex succeed in taking the four photos. Two of the photos show corpses being burned outdoors. Another shows women on the way to the gas chamber. The last photo is the most mysterious. It shows branches and a bit of sky.
These photos not only depict some facts of mass murder, but also express the circumstances and the perspective of their own creation. Especially in the fourth picture only showing blurred branches and sky, haste and danger virtually imprinted themselves in the photographic grain. The historical constellation, the situation of total surveillance, darkness and danger in which these photos were taken, is expressed in this picture: through perspective, blurring, loss of control over the frame.
Naturally, these photos by themselves are not proof in a criminalist or strictly historical sense. It is only through the reconstruction of the historical surroundings that it can be seen, to a certain degree, what the photos represent exactly, where they were made and when and in which context. Yet when this context itself has a pictorial expression, it shows the endless efforts required to produce these few and fragmentary images.
These pictures are just as improbable as the statements of testimony from the so-called Muselmänner. Not only a tremendous amount of epistemic violence was used to prevent them – and still they exist. To maintain that this is not even possible is simply false. The result of this kind of assertion: even documents that were produced despite all resistance are erased from history. Yet even the four pictures from the prisoners remain, in a sense, mute. For although it was possible to smuggle them out of the camp, they remained without impact.
 Elisabeth Cowie accordingly describes the documentary identification with “those affected” as ambivalent empathy, which serves the caring and sympathetic self-image of the viewers. Witnesses are not needed for this so much as victims. This role determines that the victims are supposed to be “really helpless” and “voiceless”, and that they are not supposed to be able to present arguments and analyses, so as not to compete with the viewer – and the film – as knowing subjects (cf. Elisabeth Cowie, “Identifizierung mit dem Realen – Spektakel der Realität”, in: Marie-Luise Angerer / Henry P. Krips [eds.], Der andere Schauplatz. Psychoanalyse – Kultur – Medien, Vienna: Turia + Kant 2001, p. 151–181, here: p. 169.)
 Cf. e.g. Klaus Arriens, Wahrheit und Wirklichkeit im Film. Philosophie des Dokumentarfilms, Würzburg: Königshausen & Neumann 1999, p. 19. Testimony as a principle already occurs in the Gospel of Luke, when the apostle maintains that his account is true, because those who have passed it on to him “saw it themselves from the beginning”.
 The so-called cognitivists seek to salvage a limited concept of objectivity or “approximate truth” (cf. Brian Winston, Claiming the real: the Griersonian documentary and its legitimations, London: British Film Institute 1995, p. 247). This understanding is implicitly based on liberal-pragmatic and partially even communitarian understandings of “common sense” (cf. Carl R. Plantinga, Rhetoric and Representation in Non-Fiction Film, Cambridge: Cambridge University Press 1997, p. 212); for instance in Noël Carroll’s article “Der nicht-fiktionale Film und ‘postmoderner’ Skeptizismus” (in: Eva Hohenberger [Ed.], Bilder des Wirklichen. Texte zur Theorie des Dokumentarfilms, Berlin: Vorwerk 1998, p. 35–69). According to Carroll, it has become a “popular sport” to cast doubt on the claim to objective information about the world. Always the same recurrent arguments are used for this, such as the selectivity of the organized film material, which means the arbitrariness of the selection of segment, shot, duration and rhythm of montage, which makes claims to objectivity futile. Carroll attempts to refute this argument by pointing out that there are standards for objectivity, which not only relativize this selectivity, but also make it appear as the perfectly normal framing process of the argument. He says that as long as certain criteria of objectivity are met, then it is entirely possible to speak of objectively rendering factual realities in the documentary medium. Bill Nichols, however, describes that these intersubjective constructs of agreement can also be interpreted as power relations: “In this case, objectivity obscures the point of view of the institutional authority itself. Here one finds not only the inevitable concern about legitimation and self-maintenance, but also historical and topic-dependent forms of self-interest, which are frequently not admitted, but are instead much more effectively disguised as tendencies and preconditions.” (Quoted by Carroll 1998, p. 55) Standards of objectivity are consequently less indicative of a “healthy intersubjectivity” (Carl R. Plantinga, Rhetoric and Representation in Non-Fiction Film, Cambridge: Cambridge University Press 1997, p. 219.) than of societal special interests and power relations, which are presented as common welfare. Raymond Williams investigated the emergence of the dichotomy between objectivity and subjectivity in the 19th century. At that time the concept of objectivity was linked with that of facticity and thus also with positivist and realist discourses and forms of representation (cf. Raymond Williams, Keywords: A Vocabulary of Culture and Society. London: Fontana 1976, p. 310 ff.).
 Cf. Michel Foucault, “Technologien der Wahrheit”, in: Jan Engelmann (Ed.), Foucault – Botschaften der Macht. Reader Diskurs und Medien, Stuttgart: DVA 1999, p. 133–144; and Toby Miller, Technologies of Truth. Cultural Citizenship and the Popular Media, Minneapolis: University of Minnesota Press 1998, p. 4.
 Cf. Michel Foucault, “Technologien der Wahrheit”, op.cit., p. 134–137.
 In its full formulation, the question is: “On the other side of the international division of labor from socialized capital, inside and outside the circuit of the epistemic violence of imperialist law and education supplementing an earlier economic text, can the subaltern speak?” – Gayatri Chakravorty Spivak, “Can the Subaltern Speak?”, in: Cary Nelson / Lawrence Grossberg (eds.), Marxism and the Interpretation of Culture, Urbana: University of Illinois Press, p. 271–313, here: p. 283.
 Cf. ibid., p. 275.
 Cf. e.g.: “It is an idealistic mystification to think the camera can capture ‘truth’ […]” (Claire Johnston, Notes on Women’s Cinema, London: Society for Education in Film and Television 1975, p. 28). As Rosi Braidotti explains in her book Nomadic Subjects, a binary concept of gender also generates a binary world view, in which knowledge is reserved to a normalized subject, which is coded masculine and associated with universality, rationality, capability of abstraction, consciousness and incorporeality. In comparison with this subject position, the feminine is conceived as deficit, as non-subject, as irrational, not capable of insight, uncontrolled and identified with the corporeal. Further pairs of opposites of this gendered relationship are, for instance, object vs. subject, active vs. passive, oppressor vs. oppressed, etc. – Cf. Rosi Braidotti, Nomadic Subjects, New York: Columbia University Press 1994, “Sexual Difference as a Nomadic Political Project”, p. 146–172; also see Teresa De Lauretis, “The Technology of Gender”, in: ibid., Technologies of Gender. Essays on Theory, Film, and Fiction, Bloomington & Indianapolis: Indiana University Press 1987, p. 1–30, and Chandra Mohanty, “Under Western Eyes: Feminist Scholarship and Colonial Discourses”, in: Feminist Review, No. 30 (1988), p. 61–88. In these texts the subject position of western ethnologist with respect to their objects is problematized within the framework of an analysis of power, which interprets documentary “knowledge” as an extension of the imperial and colonial knowledge regime. On this, see also Trinh T. Minh-Ha, “Cotton and Iron”, in: Madeleine Bernstorff / Hedwig Saxenhuber (Ed.), Trinh T. Minh-Ha. Texte, Filme, Gespräche, Munich, Vienna and Berlin: Kunstverein München / Synema Gesellschaft für Film und Medien 1995, p. 5–16, esp. p. 5.
 Ludwig Wittgenstein, Philosophische Untersuchungen, in: Werkausgabe, Vol. 1, Frankfurt/M.: Suhrkamp 101995, Section 243 ff. (p. 356 ff.).
 Hannah Arendt, “Wahrheit und Politik”, in: ibid., Wahrheit und Lüge in der Politik, Munich: Piper 1967, p. 44–92, here: p. 92.
 Jean-Luc Godard, Histoire(s) du Cinéma, Paris: Gallimard/Gaumont 1998, p. 86.
 Giorgio Agamben, Remnants of Auschwitz. The Witness and the Archive, New York: Zone Books 1999, p. 33.
 Cf. ibid., p. 33 f.
 Ibid., p. 34.
 Ibid., p. 166.
 Ibid., p. 165.
 The text has meanwhile been published in German as a book (Georges Didi-Huberman, Bilder trotz allem, Munich: Fink 2006). However, I refer here to an unpublished transcription of a lecture that was held in 2003 at the Vienna Academy of Fine Arts.
 The context often has to be laboriously reconstructed, as can be read in the controversy surrounding the way the Wehrmacht Exhibition dealt with some of its photos documenting crimes. After other historians long questioned whether the exhibited photos really showed crimes committed by the Wehrmacht (German army) or by the Soviet secret service, a meticulous reconstruction of the circumstances of the crimes, which was not at all directly evident from what was to be seen. This meant the the historians had to fulfill the task of precise reading and labeling that Benjamin called the task of the photographer: discovering “guilt” in the pictures and “identifying the guilty” (Walter Benjamin, “Kleine Geschichte der Fotografie”, in: ibid., Gesammelte Schriften, Vol. II.1, op.cit. p. 368-385, here p. 385). The reconstruction then led not only to relabeling the pictures in question, but also to a more precise reflection on the status of photographs as documents: “The photograph is regarded as the medium that purely and truthfully depicts reality. Yet the picture is always only a segment of that which took place in front of the lens, it shows a small moment from a progression of time. Like every written document, photography also requires dealing critically with the sources. Unlike the abstract text, the figurative picture suggests to the viewer that he or she is a witness to events. Photography is still too little used as a source. The problems of verifying authenticity and truthfulness seem too diverse. At the same time, missing or contradictory information in the archives reinforces the existing insecurity in dealing with pictorial sources. Methodological tools for appropriately interpreting photos have hardly been developed yet.” (Hamburger Institut für Sozialforschung, Verbrechen der Wehrmacht. Dimensionen des Vernichtungskrieges 1941–1944, exhibition catalogue, Hamburg: Hamburger Edition 2002, p. 106).