Introduction
On October 13, 2013, the regional court of Bonn, Germany, displayed military video material in a public hearing. The court had to decide on compensation claims presented by the Afghan citizens Mr. Abdul Hannan and Mrs. Qureisha Rauf against the Federal Republic of Germany. The plaintiffs claimed damages for the killing of their relatives in a bomb attack on a sandbank in the Kunduz River, Afghanistan, on September 4, 2009.
Germany has no courts-martial. German soldiers are tried before civilian courts. However, it was not until investigations into this NATO airstrike, the so-called Kunduz airstrike, that the German judiciary employed the law of armed conflict to assess the criminal liability of German Bundeswehr soldiers. Up to this point, all legal investigations into the lawfulness of military activities employed domestic criminal law.Footnote 1 In civil matters, German courts had previously only seen a single case in which individuals claimed compensation for damages caused by the German armed forces.Footnote 2 Given the novelty of assessing military conduct by the means and principles of international law, the legal actors in the domestic courts had to enter unknown territory. Next to the inevitable requirement to work with a body of law that most participants had only occasionally or never worked with before, a range of additional demands appeared. The case required legal actors to establish professionally competent ways of telling what constitutes good military practice. To do so, they needed to work out the lines between legal and illegal conduct.
During the investigations into the airstrike, it had become clear that the German military had ordered the killing of a large, but undetermined number of civilians who it had wrongly identified as Taliban fighters. The legal evaluation thus included explicating what it means to take “precautionary measures” (Art. 51, Protocol I, Additional to the Geneva Conventions) and to “do everything feasible to verify that the objects to be attacked are neither civilians nor civilian objects” (Art. 57, 2, a, i, Protocol I). Consequently, the court's decision can be read as a legal commentary on military identification work.
How did the legal actors approach this unfamiliar matter? How did they assess the lawfulness of the airstrike? In the civil procedure, the burden of proof rested on the plaintiffs, who based their claims for compensation payments on the commander's “grossly negligent violation of international humanitarian law.” They argued that on the basis of the visual material available, a “reasonable soldier” would have identified the airstrike targets as civilians. The court accepted the lawsuit.Footnote 3 Following the argumentation of the plaintiffs, the court treated the military video material recorded during the airstrike as key evidence in determining whether the military commander violated humanitarian law. In this article, I show how the judges and lawyers jointly organized and assessed the military video material.
During the hearing, defendants and plaintiffs agreed on one central point: the people who were targeted had been visible only as “infrared dots.” All participants in the hearing thus abstained from assigning the prescribed categories of civilians/combatants. In the written follow-ups to the hearing, the plaintiffs turned the infrared dots into possible civilians, while the court concluded that they were possible Taliban. This article is concerned with this categorical finding and how it was accomplished through various ways of knowing what a “military viewer” would have made of the pictures.
After providing a short overview of the conditions under which the airstrike became a national scandal, I examine the role of the hearing in the categorization accomplished during the procedure. To reconstruct the procedural categorization work, I combine two ethnomethodological analytics: I use a trans-sequential approach to membership categorization analysis (MCA). I draw on my ethnographic field notes of the court hearing as well as legal documents.
Following the legal categories through different working episodes of the proceeding allows the examination of judicial politics as a matter of organizing the inclusion and exclusion of contested issues: highly controversial questions surrounding the airstrike were excluded from the hearing. As the written follow-ups show, these issues were fundamental for assessing the visibility of civilians.
The Airstrike and the German Afghanistan Debate
Debate over the Kunduz airstrike took place in various political, administrative, and legal spheres, expanding from the original concern about the particulars of the airstrike itself to encompass broader issues regarding the government's information policy and public-relations techniques. In the course of what became a highly public scandal, several members of government resigned, including the Minister of Defense and his deputy.
For the German public, the airstrike was not primarily problematic as a potential war crime, but as evidence of an act of war as such. The incident in Kunduz did not fit the typical events reported from Afghanistan before, in which German soldiers were ambushed during a patrol and had returned fire. Until the Kunduz airstrike, German officials had claimed that German soldiers were not at war in Afghanistan and that they had limited the use of lethal force to situations of self-defense (Deutscher Bundestag Drucks 2011; Friesendorf Reference Friesendorf2012). Treating the airstrike as an act of war required a substantial reconfiguration of the legal framework that had governed the German engagement up to this point. In the criminal investigations into the Kunduz airstrike, the Attorney General's Office categorized the situation in Afghanistan as legally being a noninternational armed conflict: a conditio sine qua non for applying the laws of armed conflict (Generalbundesanwalt 2010). On the basis of official reports and press accounts, the Attorney General's Office determined that the German soldiers had found themselves in a situation that met the legal definition of an armed conflict. Facts on the ground had thus overruled the political mandate.
Type of Activity: Defensive-Offensive Strike
Despite (or maybe due to) the investigative efforts and the considerable degree of transparency the investigations generated, it is still difficult to tell what actually happened that night in Kunduz. Even the most basic questions, such as whether the strike was a defensive or offensive action, remain in dispute. What appears uncontroversial is this: the German commander had declared a “troops in contact” situation and called US fighters to a sandbank in the Kunduz river to provide close air support. Later investigations showed that there were no ISAF troops at the sandbank or nearby, and that the German camp was more than 7 km away (linear distance). The human intelligence source did not report on a planned attack, but named four Taliban leaders as on site (Deutscher Bundestag Drucks 2011).
In his testimony before the parliamentary inquiry and the Attorney General's Office, the commander offered a dual justification for his decision: he had wanted to use this chance to kill the Taliban leaders he knew to be on the sandbank (offensive), but he had also feared that the fuel tankers might be used to launch an attack against his camp (defensive). The Attorney General's Office declined to rule on whether the event was an offensive or defensive strike, arguing that under international law both justifications posed legitimate reasons. It did not consider possible violations of rules of engagement, but held that these were “organization internal regulations” (Deutscher Bundestag Drucks 2011, 68) and therefore irrelevant for the assessment of a possible war crime.
Categorization in Warfare
The legal actors of the civil proceeding sought to determine whether civilians were or were not recognizable on the videos. In the following, I briefly comment on the principle of distinction in international humanitarian law (IHL), and outline how I use MCA to study the categorization of the people on the videos as a practical accomplishment of the legal interaction.
Legal Categories of Armed Conflict
The Geneva Conventions stipulate that “parties to a conflict must at all times distinguish between civilians and combatants” (Art. 48, Protocol 1, Additional to the Geneva Conventions). Civilians are entitled to protection unless they take a direct part in hostilities. State practice has established this rule as a norm of customary international law applicable in both international and noninternational armed conflicts (Rule 1, Customary IHL). And yet, there are different ways that governments successfully turn alleged civilian casualties into legitimate deaths under international law. The following are three such examples.
-
1. Civilian status denied: Who is considered a civilian for the purposes of the principle of distinction? The conditions of warfare are often characterized and problematized as new (Münkler Reference Münkler2004; Kaldor Reference Kaldor2013) or asymmetric, and it is said that these conditions eroded many of the distinctions made in IHL (Odermatt Reference Odermatt2013). Some argue that the distinction between civilians and combatants cannot be drawn within these settings and that they need to be replaced by an understanding of status as a continuum (Dershowitz Reference Dershowitz2006). In this regard, “civilianity” becomes a matter of degree, rendering babies more civilian than thirty-year-old sympathizers, who are still considered more civilian than those who materially support terrorists. Lena Jayyusi (Reference Jayyusi, Dupret, Lynch and Berard2015, 281) has described how Israeli officials constructed a “landscape” in which “civilians not only can turn out to be disguised combatants … but [can] be part of the combat as willing or unwilling shields.” The “mutual transmutability” of the legal categories, she argues, produces “shadow figures” with a defeasible status.
-
2. Tolerable cost of action: Civilians can be said to have been recognized before an attack but may have been considered an “acceptable” cost of action against bona fide military targets and thus meeting the IHL principle of proportionality (Art. 51, Protocol 1).
-
3. Invisible civilians: When ex-post investigations find that civilians were killed, governments may accept the after-the-fact status, but argue that these civilians were killed by accident because they were mistakenly seen as military targets.
With the Kunduz airstrike, all acknowledged that civilians were among those killed and wounded in the attack.Footnote 4 They were not treated as an acceptable cost of action, but as invisible civilians.
Distinction and Visibility
How is invisibility approached as a legally accountable and assessable matter? As Jayyusi (Reference Jayyusi, Dupret, Lynch and Berard2015, 282) points out: “In IHL, a logic of visibility organizes the categories and the discursive terrain in which they are employed.” By adopting this logic of visibility, proponents of modern warfare advocate the benefit of technological systems that can provide better vision, as this advertisement for a new drone puts it:
U.S. military drones have been criticized for their lack of precision, leading in some cases to unnecessary civilian deaths . … [W]e can build a prototype drone that has more accurate image-capture and image-processing abilities than the current generation of drones … capable of performing strikes with surgical accuracy, thereby greatly reducing the potential for unnecessary civilian casualties. (Panopticopter project information)
Presentations like this render the problem of distinction a matter of technology: if the “image-processing abilities” are advanced enough, the problem of civilian deaths can be “greatly reduced.” At the same time, the high rates of civilian casualties resulting from these “surgical strikes” cast doubt on the promise of clean warfare.
Investigations into possible war crimes also utilize a logic of visibility. In the aftermath of combat-related killings, military video material is often considered as the best evidence of what really happened. In the Kunduz airstrike compensation proceeding, no one involved was questioned. Instead, the judge effectively gave the video footage the status of a material witness (Schuppli Reference Schuppli, Franke and Weizman2014) that would talk about the events.
Analyzing Practices of Categorization
To study how the video material was “made to speak” about the status of the civilians, I draw on an ethnomethodological approach to the practices of categorization. MCA, first explored by Harvey Sacks (Reference Sacks, Gumperz and Hymes1972c), examines how members use categories to organize their knowledge of the world. People deploy categories in attributing blame (Watson Reference Watson1978; Housley and Fitzgerald Reference Housley and Fitzgerald2003), in detecting lies (Lynch and Bogen Reference Lynch and Bogen1997), in making gender relevant (Stokoe and Smithson Reference Stokoe and Smithson2001), in identifying troublemakers (Lepper Reference Lepper1995), or in telling stories (Sacks Reference Sacks, Gumperz and Hymes1972c). Categories allow people, situations, and problems to be described in mutually intelligible ways. They are interactionally grounded and thus are shared; and because they are shared, they offer the possibility of scrutiny—the making of inferences, assessments, and judgments. In my analysis, I draw on the following three concepts of MCA.
-
• Membership categorization device: Categories work as ordering devices because they are kept in collections that are perceived as going together (Schegloff Reference Schegloff2007): for example, mother, father, and child belong in the collection family. All categories can be kept within different collections. Accordingly, the category child can also be used within the collection stage of life (baby, child, teenager, etc.). Rules of application provide for the recognizability of the chosen collection. The collections, along with the rules of application, make up a membership categorization device.
-
• Category-bound activities: Membership categories, as used by the parties to an interaction, are “inference rich” because they make certain kinds of predicates and activities relevant as part of the adequate description of activities (Sacks Reference Sacks1989, 272). Mentioning a certain activity can thus indicate a specific category membership, and vice versa: mentioning a category can make certain activities relevant, showing them to be “category bound,” for example, a crying baby (Sacks Reference Sacks, Gumperz and Hymes1972c, 337).
-
• Standardized relational pairs: There is a class of categories that are commonly used together whereby mentioning one can make the other relevant, for example, teacher/student or mother/child. These standardized relational pairs carry specific rights, obligations, and expected behavior among the category pairs.
Analytic Approaches to the “Military Viewer”
To determine whether civilians could be seen on the videos, the participants of the legal proceeding sought to establish what “someone like the commander” would have made of the visuals. To this end, the legal professionals produced the category of the military viewer. In my analysis, I am interested in how the participants accomplished the figure of the military viewer, and I draw on the following three approaches to discuss the relation between the military viewer and the recognizability of civilians.
-
• Viewer's maxim: Sacks argued that the use of categories relies on “relevance rules,” one of them being the viewer's maxim: “If a member sees a category-bound activity being done, then, if one can see it being done by a member of a category to which the activity is bound, then: See it that way” (Sacks Reference Sacks, Gumperz and Hymes1972c, 338). Thus Sacks's focus is on the recognizability of an account whether within a story/narrative or through observable activities. For example, a woman who is picking up a crying baby can commonsensically be seen as the mother of the baby. Typically, no one will feel the need to go and ask the woman if she really is the mother. Sacks's interest is in the unproblematic understanding of events by listeners, readers, and viewers. Here, recognizability does not rely on the viewer's special competences or activities, but on culturally shared assumptions. The viewer's maxims are based on common and recognizable attributes attached to categories.
-
• Creating observables: Although recognizability may be taken for granted in our everyday affairs, there are situations in which creating recognizability requires extra work. The military, which is instructed to employ the standardized relational pair civilians/combatants in the sense making of an operational field, may have to create recognizability actively in its identification work. Sacks (Reference Sacks and Sudnow1972b) analyzed how police officers are able to foresee a line of action and make inferences about people's moral character.Footnote 5 They need to identify criminals under the condition that “persons seen are (differentially, to be sure) oriented to the character of their own appearances as grounds of inference as to probable criminality” (1972b, 283). This is to say that people have an understanding of what would make them look like a criminal, and based on that they adopt an appearance that does not account for their criminal intentions.Footnote 6 Inferring a person's criminal intention under these circumstances seems to involve more than inferring from the observable activity to the action-bound category as suggested in the viewer's maxims. What the police need to accomplish first and in situ is to render someone's appearance an observable that permits warrantable inferences about that person's moral character.
-
• Professional vision: With the term “professional vision,” Charles Goodwin (Reference Goodwin1994) refers to the ways members of a profession shape their phenomenological environment into profession-specific objects of knowledge. Along these lines, agreeing on something that everyone can see is the outcome of professional practices such as coding, highlighting, and representations that are articulated, and can thus be studied as an observable activity. Goodwin describes how police experts in the Rodney King trial employed graphic practices to instruct the jury to see activities in the video-recorded incident that no one could see before. One key device to achieve this was structuring the ongoing action in tiny sequences of rising and falling violence that turned little motions of Rodney King's body, as he was beaten, into visible acts of aggression. In relation to Sacks's work on police assessment, Goodwin's example deals with a secondary categorization by the jury that is taught to see like the police. Similarly, the participants of the trial analyzed in this article set itself a comparable task: they want to see like the military in order to assess the commander's categorization.
My analysis shows how the judges and lawyers drew on different approaches to the military viewer during different working episodes of the proceeding.
Expanding MCA to the Procedural Course
Conversation analysis developed as a research program that investigates members' methods of meaning making on the basis of tape-recorded talk. Employing a combination of MCA with conversation analysis, MCA studies in law and society scholarship have focused on specific moments in the procedure, and analyzed transcripts of hearings (Winiecki Reference Winiecki2008; Licoppe Reference Licoppe, Fitzgerald and Housley2015) or specific documents like expert reports (Wolff Reference Wolff1995; Schank Reference Schank2012/1213). I expand the analytical context of MCA and analyze the categorization work of the civil hearing in its relation to a chain of other events that constituted the civil proceeding. Face-to-face interactions during court hearings usually attract a lot of attention, but the practical relevance of these hearing activities, I argue, can be understood only when analyzed as events in ongoing processes. Most of procedural work is done on the basis of highly formalized and written turns: letters, statements, protocols, claims, decisions, and the like. Trans-sequential analysis treats legal proceedings as an organized form of making, sustaining, and challenging relevancies across “processual events” within an “eventful process” (Scheffer Reference Scheffer2007, 183). I show that the categorization work was not resolved during the hearing, and I argue that the meaning of the hearing's activities can be understood only by looking at the participants' interpretations of this event. In this case, the participants' interpretation of the categorization work in court could be found in the written statements following the hearing: the plaintiff's statement and the decision of the court.
Approaching the Military Viewer
In the following, I will show how the participants of the hearing coproduced the military viewer in the course of the public hearing.
Hearing: The Staging of Competence
Judges and lawyers came together to watch the videos and discuss their interpretation in front of an audience. The session included expert testimony provided by two military experts and one Afghanistan expert. In this way, the hearing was set up to test the plaintiff's substantial empirical claim: could the civilians be identified?
I was seated in the first row of the stands and took field notes on the spot. For more than a year I had been following the course of the casework, studying the procedural documents, and conducting ethnographic interviews with the lawyers of the victims.Footnote 7 Here is what I recorded in my field notes:
Two large screens have been put up on two sides of the room. The screens show sequences of the video recording that the plaintiffs' counsels have submitted for the hearing. I see the contours of a barren landscape from a high camera angle. At times the pictures have a slightly higher resolution than the fixed images that circulated in public media; then the image flickers, turns greenish, and the perspective changes. Alternately, I concentrate on the pictures and then on the people in the room. Many in the room stare intently at the pictures, while others can't focus; they look out of the window or glance at their papers or hands. The video shows little black bars or spots moving over a lighter colored ground.
I later realized that the sequences had been shown in a chronologically mixed order; episodes after the bombing were followed by episodes from before, and so on. Furthermore, the videos were taken by a variety of (sometimes unidentified) sources from various locations.
What Everyone can See
The joint “seeing” is instructed by occasioned comments (seeing instructions) by the lawyers from the complainant bench. For example: “What we can see here very clearly is how people run on to the sandbank from all directions.” … Later, the complainants reformulate their remark about the running people: “You can see here how curious people came running onto the sandbank from all directions.” Members of the counsel for the defense talk audibly to each other: “No, this is nothing you can see, eh?” A short discussion takes place among the lawyers for the defense, and they agree on the point to be made. Then, directed at the judge, one of them says: “I would like to remark that it is not possible to see the intention of the people. So that these are curious people, one cannot see that in the pictures.”
The judge takes up this intervention and translates it for the complainants (mediating): “So yes, you probably got this information from the investigations you undertook, that is how you come to speak about curious people. But here and now we want to focus on whatever can be seen in the pictures. So we only want to talk about what we can see.” The point is accepted by the parties.
What is the relevance of potentially seeing the curiosity of the people? The attribute or feature of curiosity may be clearly category relevant—or even category bound—in the context of an anticipated Taliban operation. Assuming that it was only insurgents who came running to the sandbank, attributing curiosity to them seems incongruous. The insurgents were said to have formed a group integrated by a common plan and a common task. The members of this group were thus expected to share knowledge about the joint undertaking in which they were engaged. The categorization of insurgents as produced and used in this account is thus one depicting a highly integrated, closely knit, and well-organized group. By implication, it would seem to exclude random passersby who just occasionally happened to be insurgents. Being curious about what was going on—and thus, by corollary, not knowing what was going on—would therefore highlight someone as not belonging to the group.
In response to this possible counter to the defense account, the judge provided an excuse for the complainants' arguments and rendered it a mistake; that is, for any reasonable viewer of the video, this interpretation must have been a mistake. The judge then issued an instruction: participants of the hearing were to look at the video and see only whatever was there to be seen without interpreting it through the lens of additional, external information. The demand was to strictly decouple “what we already know” from “what it is we can see.” The judge's instruction was a response to a particular kind of evidential problem (Mair et al. Reference Mair, Elsey, Watson and Smith2013). Principles of fairness dictate avoiding reconstructive interpretations of understandings with the help of information that was at the time unavailable to the parties whose understandings were being examined. The video was thus to speak for itself.
The Military Experts
The pictures included various technical numbers and graphical signs. Quite obviously, this was material from a technical professional field, requiring some expertise to make sense of it. Military experts working for each side commented on the footage, drawing on their specialist knowledge to make what was being shown more intelligible for a lay audience. This is how their remarks are recorded in the official protocol:
Mr. Rose points out that during an operation, a military group would not behave in a way that can be seen in the pictures. In particular, they would group around the objects that ought to be protected, and in particular, they would stay at a larger distance from the protected objects. They would also not move upright across the terrain, but would try to find shelter, and would try to create shelter with the help of a spade, if necessary.
Rose was the military expert employed by the plaintiffs. He is a former air force lieutenant-colonel of the Federal German Armed Forces without actual combat experience. Rose provided some sense of what a military group involved in an organized operation would typically look like. The record notes Rose's testimony in the form of a list of criteria. Leaving the presence of spades aside, Rose's proposed criteria referred to rough movement formations that did not require detailed images in order to be identified. On this basis, Rose rendered a military operation perfectly observable in principle and, more significantly, suggested that the distinction between insurgents and civilians was equally observable.
For this expert, the appropriate question to ask would have been: “Is there a military operation visibly going on?” Why did Rose try to see a military operation here? Looking for a military operation can be interpreted as being rooted in at least four logics: (1) a legal logic (IHL refers to ongoing hostility), (2) a moral logic (a condition of ongoing hostility has implications with regard to the “who started it” question and the inevitability of the strike), (3) a logic of recognizability (by “engaging in hostility/engaging in a military operation” the Taliban become an observable category), and (4) a cross-procedural-history logic (here, testing what the commander had testified in other procedures; namely, that he was able to see military activities).
The record proceeds with the remarks of Mr. Marosz, the military expert for the defense. Marosz identified himself as a lieutenant-colonel with ten years' working experience in the analysis of aerial photographs.
Mr. Marosz confirms that a group of people can be seen here, gathered around an object. He points out that the image gets darker the warmer the pictured body or object is. He explains: “I also cannot recognize military conduct in the classical sense, but also no typical civilian conduct. I recognize two groups of people, sort of two knots of people, and individuals that move back and forth.”
Marosz described how he interpreted the pictures. According to the record, he began with some very basic technical information on how to read infrared pictures. He denied the recognizability of “typical civilian conduct” and “military conduct in the classical sense.” Unlike Rose's testimony, for Marosz, this understanding was not explicitly criteriological: he did not spell out what civilian or military conduct would look like. Instead, he confined his account to what he could see in the pictures: an empirical description of the dots as groups of people and their movement as “back and forth.” He refrained from drawing conclusions about the status of the people and made no links between observable activities and categories of people. His reference to typical civilian conduct came without explications on what civilian conduct would look like on these pictures. It suggests that the recognizability of civilians and the recognizability of people engaged in a military operation were of a similar kind.
The expert for the defense highlighted the technical character of the material and contributed largely to an understanding of the technical specifics by delivering instructions on how to read the colors, numbers, and signs. He thereby rendered the problem of vision a technical problem: a problem of understanding the technical equipment and the technical language. The plaintiff's expert, in contrast, concentrated on the scenic sense making of movements. He maximized the interpretative value of the pictures by turning the question of distinction into a question of recognizable activity (here, an ongoing military operation).
What did the two experts try to prove? In neither of the two alternative ways of characterizing what was happening on the sandbank were the dots cast as aggressive, as preparing for an attack, as being part of a military operation, as carrying weapons, or otherwise acting in a hostile fashion. The plaintiff's expert tried to convince his audience that the observable activities would not resemble a military operation. I would have expected the expert of the defense to do the opposite: to argue that what can be seen indeed looks like an ongoing military operation. However, the expert for the defense readily agreed that he could not see a military operation and he did not show any attempt to make Taliban a recognizable category in the material. His interpretation thus incorporated the points made by the plaintiff's expert and he merely added that the videos equally did not show civilians. I was puzzled by this line of argument and how it related to the plaintiff's argument. The written statements that followed the hearing also show how these two argumentations were used in favor of the respective side's case.
After the video screening, an expert for Afghanistan was questioned on typical civilian and Taliban characteristics.
The Afghanistan Expert
In the everyday practice of attributing features and activities to categories, attributes need to be recognizably appropriate, but they do not necessarily need to be correct (Jayyusi Reference Jayyusi1984). However, in this case, the court tried to specify properties that could correctly—that is, forensically or scientifically—distinguish civilians from the Taliban. It wanted to distribute specific characteristics to the categories they must belong to. As the chief judge put it:
Some say that no one could have expected the population to be out at night because it was the time of Ramadan. Others say that the people are especially active at night during this period of feasting. So this is all a bit unclear now, and we would like to hear a certificated Afghanistan expert on this matter.
The questioning of the Afghanistan expert started with questions about his credibility and reliability,Footnote 8 eliciting details about his qualifications, career trajectory, how much time he had lived and worked in Afghanistan, his language skills, and his work for the United Nations and the German embassy in Afghanistan. His answers to this list of questions provided a display of his competence.
| 1 | Chief judge | So what can you tell us about the Taliban? They are not a standing army with a membership card and all, are they? I mean, I shouldn't think of them as being like our military with a uniform and all the trimmings? |
| 2 | Afgh. expert | I've been in direct contact with the Taliban, up to the time of the Taliban rule . … They are not organized in the form of a standing army, but more like a guerilla movement which operates from within the population and are very hard to distinguish. There are no exact numbers . … |
| 3 | Chief judge | What is a usual size in which a Taliban group operates? |
| 4 | Afgh. expert | Well, from what I know they usually operate in very small groups. Twenty men is already extraordinarily big. |
| 5 | Chief judge | So it has never occurred that they have operated in bigger groups, for instance with several small groups doing something together? |
| 6 | Afgh. expert | Well, over the last 12 years there were as far as I know only three operations that included up to 60 Taliban. But these were extraordinary operations, like the storming of Kandahar. There were about 60 Taliban involved in this, but this was a really extraordinary undertaking. The hijacking of fuel tankers is not quite an operation like that, I suppose. |
| 7 | Chief judge | And so when something gets stuck, like the fuel tankers here, is it possible that the Taliban can ask civilians for help? Or do helpers have to be Taliban? |
| 8 | Afgh. expert | … Well, you see, they did ask the population of the surrounding villages for help, that's what they did . … This is nothing unusual. In a situation like that it may be quite natural to just go and ask the people from the villages. It's also, I have to highlight this point again, the Taliban are no alien elements within Afghan society; they are part of the population. This is why it is not an unusual thing for the Taliban to call the people from the villages. |
The line of questioning started with a search for stable signs of distinction: in (1) the judge refers to a “uniform” as a visible sign (in principle) or a “membership card” as a formal record. The existence of a membership card is interesting in regard to various possible ways of organizing membership. The card could work as a possible medium for certifying membership locally or otherwise, administrating the internal organization of an in- and out-group. The example seems to be so wide of the mark, however, that it does not appear to be a serious remark. Consequently, the expert does not even bother rejecting the idea of membership cards, but instead treats it as a general invitation to talk about the Taliban's organizational structures. In (2) the Afghanistan expert makes “operating” relevant as a category-bound activity of Taliban fighters. The phrase “from within the population” prepares the ground for his assessment that they are “very hard to distinguish.” For whom is this distinction difficult? Is it difficult for anyone or for observers in general? However, this general observer does not seem to include the Taliban itself or the population. Or is the Taliban not a recognizable category within Afghan society?
In the next two question-answer pairs, the size of the group is discussed as usual or unusual but possible. In (3) the judge asks about the usual size of a Taliban group. In relation to what the expert states as a typical size, the presence of a big group on the scene could then be category relevant or even category bound for a non-Taliban group. In the following question (5), the judge deconstructs the general rule by asking for possible exceptions: Has it never occurred? In his answer the expert tries to immunize his statement against wrong conclusions. He highlights the point that bigger operations have been seen only in extraordinary undertakings, and indicates, in his opinion, that the case at hand is not extraordinary in this sense. In (7) the relationship between the Taliban and the population is topicalized by the judge. Is asking civilians for help a possible activity in the Taliban/population pair? In his answer, the Afghanistan expert begins by pointing to the counterfactual implication of the question: according to the facts (as known from ex-post investigation), this is indeed what had happened. He then addresses the judge's desire for a general statement on the issue, and declares that this type of activity is not generally unusual. He thus unites the just-this-ness of the event with general knowledge of social phenomena.
The questions the judge put toward the Afghanistan expert render the problem of distinction as a matter of culture: the Afghan people become an exotic group and cultural knowledge is needed to make sense of them. What are they like and how can we tell Taliban from civilians? The judge seeks to ascertain decontextualized, stable criteria of distinction, but the society the Afghanistan expert describes resists such bifurcation: it is not either Taliban or civilians who would use mobile phones, pickups, or motorcycles; it is not just one or the other who is going outside at night or celebrates Ramadan. When he points out that the Taliban are part of the Afghan population, the binary distinction of Taliban/population appears to collapse entirely.
The Military Viewer and His Special Competences: Professional Vision
The court refused to take any more evidence after this single session. It denied all other complainant's requests to enter evidence, including requests for questions to be put to the commander, the pilots, and other personnel implicated in the airstrike. The court declared the case to be decidable on the basis of the different kinds of category work performed in the course of the proceeding. Going back over that work is therefore instructive. During the hearing, the question of what a military viewer would have made of the activity on the sandbank was approached variously by:
-
1. What everyone could see: that is, people coming and going;
-
2. What military experts could see: that is, people not engaged in military activities;
-
3. What the Afghanistan expert could see: that is, the shared characteristics of civilians and Taliban.
The military viewer was thus invoked as a category belonging to what Sacks (Reference Sacks and Sudnow1972a, 37) calls “collection K”: a category set organized around an asymmetric distribution of knowledge. The standard example is that of the expert/layperson pair with its various realizations, such as a doctor/patient pair. The status of testimony during the hearing was tested with respect to this assumed unequal distribution of knowledge. The military experts and the Afghanistan expert were held to possess soldierly knowledge beyond that of a layperson—indeed, this duality is what their status as experts hinged upon. By implication, the other parties to the hearing—lawyers, judges, and complainants—belonged to a category of people who could be seen as lacking the kinds of knowledge needed to interpret the pictures correctly. The military viewer was thus defined by a domain of specialized knowledge and skills upon which the soldier's competence rested. Did this performance of special competences enable the legal actors to distinguish civilians from Taliban? During the hearing, this was apparently not the case. The military viewer as jointly produced in the hearing was not able to make sense of the video in terms of the legally prescribed categories. The only category the participants of the hearing managed to apply to the infrared dots was people—proposing a rough distinction between various possible things in the operational field: people, vehicles, animals, and so forth. The categorization work was thus left undone: the military object was not constructed as a legal object.
Undoing the Legal Capacity of the Military Object
How was it possible that the participants of the hearing did not assign the legal categories? I argue that the noncategorization needs to be understood as a product of the hearing that inextricably reflects the practical conditions of its production. What were the specifics of the circumstances in which the legal actors analyzed and used the video?
We have seen that the judge guarded the exclusion of information that was not available to the soldiers at the time of the airstrike. But more than just excluding information gathered after the fact, the hearing also excluded available information and its influence on the military's capability to make sense of all the people. (1) The German command had received the (incorrect) information that all individuals on the sandbank were insurgents. (2) The German command (wrongfully) told the US pilots that they would be facing a troops-in-contact situation. A number of the features of the hearing seem to provide for this exclusion: (a) the pictures were not read together with the existing transcripts from the pilot-to-pilot and pilot-to-ground controller conversations; (b) the video sequences were shown in a chronologically mixed order; (c) both experts testified as experts for the pictures, but did not comment on the military operation.
In sum, the hearing did not reconstruct the actual military identification work. The participants did not interpret the pictures as an integral part of a flow of (changing) information available in the course of the coordination and planning of the airstrike. Instead, the hearing took the material out of the context of the military operation and objectified it as separate from and additional to the military activities.
The rigorous exclusion of the situational circumstances allowed interpreting the videos while sustaining the ambiguities about the type of attack. As I pointed out earlier, the Attorney General's Office had declined to rule on whether the event was an offensive or defensive strike. While, in their written claim, the plaintiffs had argued that the attack was offensive, the question was set aside during the hearing.
The unintelligibility of the video arose from this specific legal setting. Here, a military viewer emerged vis-à-vis a military object, and the legal actors could not put these two objects in a legal relationship. The military object appeared to resist legal coding. The hearing had thus resulted in the undoing of the legal capacity of the military object.
The participants had created this nondiscriminable object in a rather consensual manner. But as the subsequent turns show, the plaintiffs and judges did not, in fact, agree on the interpretation of the videos as legal evidence. How did they utilize the infrared dots for their argumentation?
The Subsequent Turns: Preferences Within the Viewer's Maxims
The final step of the categorization work occurred after the hearing. The plaintiffs communicated their categorical finding in a note to the hearing, which they submitted to the court. The court presented its established category in its decision. As I show below, the complainants and the court both turned to a normative understanding of military viewing and employed seeing preferences embedded in distinct viewer's maxims. Moving from the professional vision design of the hearing to the viewer's maxims allowed the complainants and the court to conclude what a military viewer should see in the scenes displayed on the video material.
Complainants: Viewer's Maxim of Caution
Because weapons could not be identified on the sandbank, it was equally impossible to conclude that a military operation was going on . … [Among other things], the way those on the sandbank were moving indicated the involvement of civilians . … [But the] German commander made it harder to determine whether civilians were present by instructing the pilots to “hide” and not engage in a show of force. (Complainants' note to the hearing, November 2013, unpublished)
In their follow-up summary of the hearing, the complainants highlighted details such as the size of the group, their vibrant coming and going, the absence of movement positively signaling military intent, and the lack of discernable weapons to argue that “from the outset, the commander would have had to expect the involvement of the population in the events.” In this regard, the complainants displayed a preference for seeing civilians in the dots. Their version of the viewer's maxim thus stipulates: “If it is possible to see some of the people as civilians, then see them that way.” I call this a viewer's maxim of caution: to prevent collateral damage, the reasonable soldier ought to exclude the possibility of civilian presence, and in order to do that, good evidence (in this case video footage) is required. The high degree of caution is fitted to a nondefensive situation: in the absence of an imminent threat, the military would have had the opportunity to strike only when they were very certain that the bombs would not hit civilians.
Court: Viewer's Maxim of Hostility
On the basis of the infrared pictures examined it was equally impossible to tell if the people that were displayed as infrared dots were old or young, tall or small. … While the court accepts that people and vehicles were moving back and forth from the fuel tankers without recognizably displaying the kind of defensive formation that would imply military coordination, it does not accept the further proposition … that the people who appeared as infrared dots [therefore] had to be [seen as] civilians. (Landgericht Bonn 2013, 67–69, author's translation)
The court drew its conclusions after considering whether the video footage had presented anything to the commander that could have forced him to see the figures on the ground as anything other than a Taliban force: “A gathering of 50 to 70 people is unlikely … but does not compel the conclusion that it could not be Taliban fighters” (Landgericht Bonn 2013, 19, author's translation). The military viewer's maxim the court thus employs may be summed up as follows: “If it is possible to see the people as ‘all Taliban,’ then see them that way.” I call this the viewer's maxim of hostility: it evokes a military ethic of symmetryFootnote 9 and suggests a general situation of kill or be killed. In this general situation, no third option is available; a participant cannot simply decide not to kill and go on as before. A soldier receives information about enemies and naturally makes plans to fight them (back). Under these conditions, civilians would need to be identified clearly as such to counter the preference for seeing the dots as Taliban. The responsibility is shifted to the civilians to provide evidence of their status.
One can see how the two viewers' maxims are built on the professional vision provided by one or the other military expert. I have shown that the defense did not aim at making the participants see anything Taliban-like. They did not try to strengthen the quality of the material as operation-relevant evidence, but minimized its value as legally relevant evidence—which was brought in by the plaintiffs. Their military expert minimized the diagnostic significance of the pictures. In combination with the viewer's maxim of hostility, this could help to render the dots as possible Taliban. The expert for the plaintiffs ruled out the visibility of military conduct, which in combination with the viewer's maxim of caution allowed the turning of the people into probable civilians. How did the complainants and the court make use of the testimony of the Afghanistan expert? Apparently, the expert had denied all stable signs of distinction between Taliban and civilians tested by the judge. Based on this, the complainants concluded that as the Taliban were said to be farmers who pick up their guns at the end of the working day and by this activity become Taliban (action-based), an ongoing military operation becomes the only observable behavior usable for distinction. The court, by contrast, seemed to have been looking for stable signs that positively identified civilians: vehicles that were only driven by civilians and times at which civilians could always be found at this specific location. However, nothing contained within the pictures was judged to aid the positive identification of civilians. In the court's decision, all behavior was thought to be Taliban-bound, or rather Taliban-possible. Even very uncommon behavior could not falsify the assumption that all the people could be Taliban.
Both accounts can be understood as categorically adequate accounts when read in light of the underlying type of activity. Each account attaches normative significance to the empirical facts established in the hearing by placing the categories into a moral narrative of what happened. The posthearing versions of the military viewer are both embedded in and constitutive of the conflict setting, the Afghan conflict. They are enmeshed in a set of relationships morally organized in terms of a differential distribution of rights, obligations, risks, and responsibilities. The claimants had rejected the assumption of self-defense and considered the videos as part of an offensive strike. The extremely high degree of caution they advocated suits the assumed absence of threat. The court did not reject self-defense, but preserved the dual justification, referring to a military viewer in general, and to a general understanding of moral and legal rights and obligations in armed conflict. However, the court's viewer's maxim can be best read against the background of a defensive strike. The low degree of caution matches the perceived seriousness of the threat.
Conclusion
In closing, I summarize my findings and highlight the value of a trans-sequential perspective on MCA for law and society studies as well as for legal practitioners. My analysis investigated the details of a particular moment in a global struggle for rights, in which participants engage in the juridification of warfare. The court accepted that the victims' relatives are entitled to claim compensation before a German court. In this regard, the trial can already be viewed as a step toward the legal accountability of the military and the individualization of IHL. However, the way in which the court analyzed the military viewer displayed an indifference to the actual practices and proceedings of the military. This created an extensive blind spot of legitimate military action and conduct as the court interpreted the videos as separate from and additional to the military activities.
What actually happened that night in Kunduz? What was the purpose of the airstrike and how did the military proceed to identify the target? Legal disputes typically do not evolve around a single disputed fact. More often, they contain long chains of controversial claims about real-world events. To operationalize a case and make it manageable for legal assessment, the court needs to organize and schedule the dispute. The schedules include closure markers, which provide the opportunity for immediate termination of the proceedings. In the case at hand, the court's conclusion that the video material does not show civilians effectively worked as such a closure-relevance point. The court used this finding to terminate all further taking of evidence. For the complainants, this meant that their case was instantly lost. For the defense, it constituted a shortcut to winning. In effect, the disputed question, “what do the videos show?” obtained the privileged position of being negotiated in a public hearing, while the other disputed facts, like the purpose of the strike, were never discussed in this arena.
To determine what the videos show, the legal actors sought to establish what “someone like the commander” would have made of the visuals. During the hearing, the judges maintained a very strict policy of dealing with one issue at a time: they singled out the issue of what the videos show by excluding all (disturbing) circumstances. Isolated and purified like this, the military viewer of the hearing could not see more than infrared dots, and lawyers and judges abstained from applying the legally prescribed categories. The way the court set up the video interpretation resulted in the undoing of the legal capacities of the military object. Notably, this is precisely what the legal actors made of the visuals. In contrast, the military material of the airstrike does not suggest that the military had worked on an undoing of legal capacities during the airstrike. Here, the videos were part of an identification process in which the categorization of the people as all insurgents had been established and stabilized.
Vis-à-vis the overhearing audience, lawyers, judges, and experts agreed to see just dots and did not stage the confrontation of possible Taliban/possible civilians as an observable happening. The hearing created and maintained a level of reduced complexity in which the actual dispute had faded away. In the written follow-ups, however, the plaintiffs turned the infrared dots into possible civilians, and the court concluded that they were possible Taliban. The complainants and the court both turned to a normative understanding of military viewing and employed seeing preferences embedded in distinct viewers' maxims. I have argued that the different seeing preferences are tied to different assumptions about the situation in which the military viewer looked at the videos. As the complainants considered the videos as part of an offensive strike, they suggested a very high degree of caution and a strong preference to see civilians.
Contrary to how the military viewer was produced in the hearing, I propose that military viewing needs to be assessed as a situated, concerted, rule-based activity through which the category of civilians finds its conditions of possibility to become visible. Civilians can become visible or probable only within the course of specific activities of making them visible and probable. In the practical settings, knowing civilians is a matter of knowing enough for all practical purposes. Based on a reconstruction of the military situation, a legal evaluation should thus explicate what was required to verify that civilians are not on the ground. IHL provides the legal prerequisites for such a situated approach to the assessment of military conduct. The Geneva Convention provision that obliges parties to an armed conflict to “do everything feasible to verify that the objects to be attacked are neither civilians nor civilian objects” (Art. 57, 2, a, i, Protocol I, emphasis added) clearly relates the verification standard to the situational demands. The legal obligations are formulated as a matter of investing more or less work into the identification of military targets. Along these lines, it is not the video as an object that makes civilians visible and observable, but the identification work that potentially creates these observables. Making civilians visible is a practical accomplishment: it needs to be done. For courts, moving closer to the actual military practices would imply developing a normative understanding of how identification work should be carried out.
