Abstract
Objective:
A small-scale exploration of the use of video analysis to study robberies. We analyze the use of weapons as part of the body posturing of robbers as they attempt to attain dominance.
Methods:
Qualitative analyses of video footage of 23 shop robberies. We used Observer XT software (version 12) for fine-grained multimodal coding, capturing diverse bodily behavior by various actors simultaneously. We also constructed story lines to understand the robberies as hermeneutic whole cases.
Results:
Robbers attain dominance by using weapons that afford aggrandizing posturing and forward movements. Guns rather than knives seemed to fit more easily with such posturing. Also, victims were more likely to show minimizing postures when confronted with guns. Thus, guns, as part of aggrandizing posturing, offer more support to robbers’ claims to dominance in addition to their more lethal power. In the cases where resistance occurred, robbers either expressed insecure body movements or minimizing postures and related weapon usage or they failed to impose a robbery frame as the victims did not seem to comprehend the situation initially.
Conclusions:
Video analysis opens up a new perspective of how violent crime unfolds as sequences of bodily movements. We provide methodological recommendations and suggest a larger scale comparative project.
There is now a tradition of research into robberies of various forms (in private homes, shops, or on the street), which seeks to understand how these interactions unfold. A classic is Luckenbill’s (1981) study of how shop robbers try to frame the situation as a robbery. This work shows that whether robberies end up in violence or not depends, among others, on how successful robbers are at imposing a “working agreement” on the victims: They should not just accept that they are victims, but they should also understand and fulfill the role that robbers have in mind for them in various stages of the robbery situation. Subsequent work has elaborated on these stages (Bernasco, Lindegaard, and Jacques 2013; Copes et al. 2012; Jacobs 2012), and crucially, the use of weapons in relation to the degree of violence involved (Cook 1981; Feeney 1986; Kleck 2005; Kleck and Delone 1993; Kleck and McElrath 1991; Wells and Horney 2002). In this article, we delve further into the robbery situation. Based on video footage of shop robberies, we provide close-up analyses of the bodily and emotional dynamics of robberies and the role of weapons therein. While our study is explorative, also given the small sample size, the contribution we aim to make is threefold.
First, while prior analyses have relied on interviews and judicial case files to reconstruct robbery situations, we use real-time visual data. This material not only allows for the validation of the results of prior studies that were based on these other data sources but it also enables us to probe deeper into the robbery situation by repeated and variable viewing (slow and fast motion, forward and backward) of the event.
Second, previous interactionist work did not consider the role that bodies and emotions play in robbery situations. To our knowledge, only Lindegaard and colleagues (2013) paid attention to emotions in robberies. However, their study did not focus on how emotions impact the dynamics of the situation. Our assumption is that emotions play an important role in the establishment of a working agreement between robbers and victims and hence influence the outcome of the robbery. We aim to infer (changes in) the emotional states of robbers and victims from the sequences of body movements and postures they display, specifically with regard to the use of weapons.
Third, with respect to weapons, prior U.S. work has demonstrated that the display of (lethal) weapons by robbers tends to reduce the severity and likelihood of violence (Cook 1981;Kleck and Delone 1993; Luckenbill 1981; Wright and Decker 1997; but see Bernasco et al. 2013, who report that in Dutch robberies, weapon use by robbers increases the likelihood of violence). We intend to specify this work by investigating how exactly weapons are being used. We are going to argue that in robbery situations, weapons should be seen as body extensions, which are crucial in the robbers’ attempts to attain “emotional dominance” (Collins 2008) which enables them to impose a robbery frame on the situation. Emotional dominance is the sense that one can impose one’s will unto the other, that one is able to make the other follow one’s rhythm, and that the other will not resist (any longer).
The main question we aim to answer is how does the use of weapons as part of robbers’ body posturing influence their attempts to attain emotional dominance? Our study is empirically grounded on Closed Circuit Television (CCTV) footage of 23 armed robberies available at the Netherlands Institute for the Study of Crime and Law Enforcement. In what follows, we will first outline the conceptual framework and specify the research questions. After that, the data, sampling, and procedures of data analysis are explained. Two sections report on the empirical results, each focusing on two distinctive stages in robberies. In the concluding section, we discuss the use of video data for the analysis of violent crime in general and robberies in particular.
The Interactional Dynamics of Robberies and the Role of Weapons Therein
Seen from the interactionist perspective on crime (see Felson 1982; Felson and Steadman 1983; Felson and Tedeschi 1993, Wilkinson and Fagan 2001), robberies, like all interactions, are performances in which actors try to present a situational identity; a claim on how others should value and treat them (Goffman 1959). Thus, a successful framing of the situation (from the viewpoint of the robbers) requires cooperation between robbers and victims, while avoiding open antagonism or resistance. The task for robbers is to impose a “working agreement” (Polk 1999) on the victims which defines how they should interact in the robbery situation. As noted above, this requires emotional dominance on the side of the robbers (Collins 2008).
Analyzing 261 robberies, Luckenbill (1981) identified four stages that robbers and victims must accomplish together in order to create a successful robbery, that is, a robbery in which the robbers get what they want at the least possible costs in terms of the use of threat, violence, and antagonistic tension—assuming that shop robbers follow instrumental reasoning, which may not always be the case. First, the robbers must move into copresence with the victims; second, the victims and robbers must transform their encounter into a robbery frame; third, one or both parties must transfer the valuables; and fourth, the offenders leave the setting with the valuables. This article focuses on the first two stages only for two reasons. First, if robberies fail, this occurs most frequently in the second stage (cf. Luckenbill 1981), making this stage the most important for the outcome of the robbery. Second, the second stage appears to be the longest, the most violent, and complex in most robberies and therefore seems most relevant to our purpose.
Let us take a closer look at Luckenbill’s first two stages of the robbery situation. The first stage, moving into copresence with the victims, means that the robbers move into striking range without raising suspicion to avoid perhaps unmanageable opposition (Luckenbill 1981:29). Robbers tend to use two tactics in this stage (see also Wright and Decker 1997:100). One of them is “speed and stealth”: The offenders prepare for the strike without being noticed by the victims (the stealth aspect) and then suddenly they move into copresence with the victims, having their weapons ready (the speed aspect), “disguise” is the second tactic. It involves the hiding of the robbers’ intentions, so that the victims perceive them to be a “legitimate part of the setting” as customers or as passersby who request information (Luckenbill 1981:29; Wright and Decker 1997:100).
In the second stage, the robbers must transform the situation into a robbery, which involves “a succession of moves between the offender and victim” (Luckenbill 1981:31). First, the victims should suppress opposition, and the offenders must be in control of the behavior of the victims by means of force or the threat thereof, they must be emotionally dominant. One way to impose the robbery frame is by using incapacitating force or what Luckenbill described as “bodily pain which debilitates or immobilizes the victim for a time” (1981:31). Another, more common way is to express a command, backed up by a threat. Luckenbill also found that the more lethal the weapons used by the offenders, the more likely the victims submitted. This is also his explanation for why so many offenders with nonlethal weapons like clubs or bare hands open with incapacitating force. They do this to gain dominance despite their lack of lethal punitive resources.
Later quantitative work in the United States confirms these findings. Kleck and Delone’s (1993) study on weapon use in robberies showed that the more lethal the weapons displayed, the less opposition and violence occurred. They also found that the likelihood of victim injury was associated with victim resistance by means of physical force or weapons other than guns, but not when victims used guns (Kleck and Delone 1993). So when either offenders or victims wield a gun, violence and injuries are reduced (see also Wells and Horney 2002). However, these findings were not confirmed by Bernasco et al.’s (2013) study of 256 robbery situations in the Netherlands based on interviews with 100 robbery offenders. They found that weapon use increased the risk of violence during robberies. Bernasco et al. (2013) showed that in 30 percent of the robberies, the offender started the robbery by using violence with the use of weapons, while in 70 percent of the cases, the offender used violence at a later stage only. While violence at the start of the robbery was not related to victim resistance, the violence in later stages occurred mostly (50 percent of the cases) after the victims resisted as a means of (re)enforcing compliance (Bernasco et al. 2013).
While Bernasco et al.’s (2013) study of robberies in the Netherlands focused on whether a weapon was exposed during the robbery and in what stage it was exposed, studies of weapons and robberies in the United States focused on the effect of the presence of weapons in robberies (Cook 1981; Kleck 2005; Kleck and McElrath 1991; Wells and Horney 2002) without specifying how weapons were used in the course of the event. For the understanding of robberies, it seems crucial to know how weapons are brought into play in the interactional dynamics of robberies, especially in the United States, where carrying a weapon is more common.
From a phenomenological approach, robberies are a form of intentional navigating in the world; robbers have projected a trajectory which they carry out by moving their bodies. We argue that weapons play an important role in the robbers’ embodied intentionality. Not only because weapons are charged with meanings that support the robbery performance but also because they literally “point” the bodies of victims and those of the robbers themselves toward a specific set of actions. In Heideggerian (1927/2002) terms, the weapon must change from present-at-hand to ready-to-hand. This means that weapons can only support or direct robbers’ intended maneuvering to the extent they are experienced as body extensions, as integral to the robbers’ embodied being. Thus, we understand the carrying, displaying, and actual use of weapons as part of sequences of body movements and postures. This also means that weapons can both support and impede robbers’ intended course of action. Let us conclude with a note on the standpoint from which we analyze the data. Both Luckenbill’s conceptualization of the stages and tactics in robberies and Collins’s theory about the emotional dynamics of violence are external to the actors studied. While this study is thus conceptually grounded on a “data inspectors’” standpoint, we combine it with a phenomenological or “participants’” standpoint—in this case that of the robbers’. 1 More specifically, we are going to analyze how robbers carry, display, and use weapons as body extensions in their attempts to establish emotional dominance and in various stages of the robbery.
Method
Data and Sample
The data are CCTV video footage of shop robberies in the Dutch cities of Amsterdam and Rotterdam. The choice for these two cities was based on practical reasons, as more than one third of all reported robberies in the Netherlands are committed in these two cities (Overvallen NL 2016). Access to the material was provided by the Dutch Ministry of Justice and the Rotterdam and Amsterdam police forces. All robberies in the sample were therefore reported to and investigated by the police. The analyses were conducted under strict judicial conditions considering the privacy of all visible persons in the video. For this reason, we had to transform the video stills included in this article into a cartoonlike quality. Unfortunately, we do not have permission to provide a link to the raw data. The data files were stored on encrypted hard disks that were kept in a safe.
The material was collected by the last author (see also Lindegaard et al. 2016). We received footage of 127 Rotterdam cases and 48 Amsterdam cases committed in 2013 and 2014. The robbery team of the Rotterdam police force uses a central computer, on which all their CCTV footage is stored. We received a copy of their data as stored in August 2014. In Amsterdam, robbery footage is dispersed among various criminal investigations teams. As the collection of these data would require too much of our time and energies, we turned to the External Communications Department of the Amsterdam police force. This department provides and edits video material for communication to the public, asking for their assistance in investigations, among others via a national television program (Opsporing Verzocht). In this study, we used the original and unedited footage.
Of the 175 cases we received, 47 had to be dropped because the video was not working or missing, and in 78 cases, the quality of the video was poor, or there was no or only part of the interaction visible. In the latter cases, we typically found recordings of cars passing by outside the shop, or the footage showed the offender entering the shop, but it did not show what happened inside the shop. Eventually, only 50 cases remained that were suitable for analysis.
Apart from Collins’s (2008) analysis of violent interactions based on photo material and Klusemann’s (2010) video analysis of the interaction between Dutch UN military leader Karremans and his Serbian counterpart Mladic preceding the Srebrenica war atrocity, no qualitative studies of crime based on video data had been published at the moment we conducted this research, so that we could not rely on established methodological practices or exemplars. Consequently, we had to explore various analytical strategies, and some of them turned out to be less useful than we had hoped. As a result, due to time constraints, this article is based on 23 randomly selected cases of the remaining 50. However, as our analysis will show, these 23 cases were a sufficient number to analyze how weapons use and sequences of bodily movements influenced the robbery situation.
It should be noted that we cannot assess to what extent our sample is representative for the approximately 2,000 to 3,000 robberies which are committed in the Netherlands yearly (Rovers et al. 2010). Our data might be biased in several respects. First, not all shops that are targeted by robbers have CCTV cameras installed. Second, the Amsterdam cases, for which assistance of the public was sought, might contain more severe robberies in terms of violence and social impact. However, we do not consider these potential selection biases highly problematic, as our aim is not to arrive at results that are representative for Dutch robberies in general.
To get a sense of the sample, we now provide descriptive statistics on key variables such as the types of weapons carried by the robbers, the number of robbers and victims, the time duration of the robberies, and the locations were the robberies took place. Table 1 displays the types of weapons offenders and co-offenders carried.
Types of Weapons, Carried by Robbers.
The sample provides sufficient variation with regard to weapons, given our aim to inquire how they may affect robbers’ body posturing differently in their attempts to attain emotional dominance. While 13 offenders and co-offenders carried guns, 10 displayed knives and 4 used hammers (in one case, a hammer was used to threaten a cashier only; in another case, three hammers were used both to smash display shelves and to threaten the shop owners). In two cases, the weapon was not clearly visible, and one robber used a chair as a weapon.
The average number of offenders was 1.7, and the maximum was 3. Similarly, the average number of victims was 1.6, the maximum being 3. The robberies lasted between 10 seconds and 20 minutes with approximately one third lasting less than a minute, one third lasting between 2 and 3 minutes, and the remaining ones more than 3 minutes. Table 2 provides an overview of the locations where the robberies took place.
Types of Locations, Targeted by Robbers.
It can be inferred from Table 2 that the robbers in our sample targeted a diverse range of locations. The seven other types of businesses were a telecom shop, flower store, taxi, café, cinema, a clothing store, and a gas station.
Analytical procedures and analytical problems
Video analysis allows researchers to map different modes of bodily behavior by various actors simultaneously, thus providing an appropriate tool for the close-up analysis of social interaction (Cowan 2014:7; Knoblauch 2009). Moreover, the opportunity to repeatedly view the data at various speeds (ranging from very fast to very slow or stills) and in various directions (backward, forward) is of great value; this enables to verify the coding, thus increasing data validity (Luckmann 2006). In our research, it occurred that in some cases, even after carefully observing the footage in slow motion, new discoveries were made as behaviors were overlooked first. This is because video data are very complex, especially when it concerns multiple actors who engage in different modes of behavior. The focus of analysis in our case was on sequences of body movements that transform the situation; every movement indicating any emotional dynamic (based on literature) was included in the coding (see Ekman 1991; Hall, Coats, and Lebeau 2005; Kluseman 2010; Scheff and Retzinger 1991). Body movements that were interrupted because actors moved out of reach of the camera were not included in the analysis.
We created a codebook that contained codes for body postures and movements that indicated either emotional dominance or submissiveness (see Appendix). Our codebook was based on preliminary analyses of the same footage with other (student) observers and prior research on the emotional meanings of bodily behavior (Collins 2008; Hall et al. 2005; Klusemann 2010; Scheff and Rentzinger 1991).
We used Observer XT software to code body postures and movements. This program allows to code different modes of behavior of different individuals at different points in time (see Cowan 2014:15). Figure 1 provides a visualization of the coding of one, relatively simple, robbery situation in Observer.
Visualization of coding with Observer of robbery situation no. 1.
Figure 1 shows two robbers who first display minimizing postures. At point 12, robber 1 moves forward to victim 1 and produces a weapon. The victim responds by moving away. At that moment, robber 1 threats the victim. All the time, robber 2 displays a minimizing posture. The strength of using Observer is that it ensures high levels of coding reliability and offers tools for quantification of predefined behavioral categories. The weakness is that, due to the fine-grained coding, the situation as a whole—which provides the meaning context to understand body posturing and movements—disappears from view. To give an example, the meaning of victims raising their hands is not unequivocal; it could mean either to submit to a threat, to actively counter a threat, or to make a calming gesture toward the robber, depending on how the situation unfolds.
Rather than increasing the complexity of the coding scheme by incorporating many different codes for each meaning of raising hands and various other bodily movements, we decided to write “story lines” to capture each robbery situation as a hermeneutic unit. This decision was motivated by our research question, focusing not on the effect of weapon presence but on how weapons and other forms of body posturing and movements influence the process of establishing dominance. The story lines depict the detailed course of action of every actor involved, including our interpretations of these actions as they follow up on one another. Figure 2 provides an example of the story line of the same robbery as depicted in Figure 1, which showed the time lines created with the Observer software.
Story line of robbery situation no. 1.
While the Observer software seduces researchers to fragment video data into predefined categories, the construction of story lines invites researchers to come up with categories inductively, fitting to the situation as a whole. As our story lines involved more interpretative work (e.g., rather than “raising hands” descriptions were now of the type “aggressively pushing her back”), we went back and forth comparing the story lines with the video footage to increase reliability. In cases we disagreed on the interpretation, consensus was established by repeatedly watching the footage.
We analyzed the story lines in five stages. First, all the cases were divided into the first two stages of robbery described by Luckenbill (1981). Second, all the relevant behaviors that signaled transformations in emotional states were highlighted. Third, the robbery interactions were analyzed in a vertical way based on the multimodal transcripts coded in Observer, now focusing on the sequences of body movements between robbers and victims. Fourth, the robberies were analyzed in a horizontal way, analyzing the sequences of interactions per robbery stage, based on the multimodal transcripts again. Lastly, all these material were grouped and compared to find patterns with regard to the carrying, display, and use of weapons, in different stages of the robbery.
Results
The First Stage: Moving into Copresence
In this section, we report how robbers work toward an encounter with their victims. In line with Luckenbill (1981), we identified the speed and stealth tactic and the disguise tactic as ways in which the offenders move into copresence with their victims. In addition to describing these forms of establishing copresence, we report on the emotional dynamics expressed in sequences of body movements and posturing.
Speed and Stealth Tactic
First, the speed and stealth tactic was apparent in 16 of the 23 cases studied. In 11 of these 16 cases, the robbers had their weapons drawn when entering the scene. Of these, seven are with guns, three with knives, and one with a blunt object (see Figure 3). These cases include both the offenders who quickly aimed their weapons at the victims directly after entering the scene and robbers who kept their weapons hidden until close to the victim.
Offenders generally tended to keep knives low, at waist height, while guns were mostly kept at shoulder height when posing the initial threat. It also seemed that robbers were more likely to keep knives hidden until close to the victim probably because they were concerned about alarming the victims, as proposed by Luckenbill (1981). Compared to gun threats, threats with a knife are less compelling from a few feet away. In 10 cases, the victims maintained their normal routines when robbers had entered the scene mostly because they did not spot the offender yet. However, in three of these cases, the victims clearly saw the face-covered offenders approaching, but they just froze or maintained their normal routines.
A robber uses the speed and stealth tactic as he enters a supermarket swiftly. He looks downward at the ground, neck and shoulders slightly bent down, holding a hammer downward in his left hand as he rushes toward the cashier (still taken from case 19).
In 11 of these 16 cases, the robbers wore some sort of face cover. Covering one’s face probably plays various functions in robberies. First, there is the obvious goal to hide one’s face from being recognized, also given the widespread use of CCTV cameras in shops. Second, putting on a face cover provides a symbolic–emotional transformation for the robbers themselves, preparing them to play their role in the upcoming action. However, sometimes face covers such as hats, caps, or loosely tied shawls seem to inhibit the successful imposing of a robbery frame at least for a moment: Robbers clumsily need to adjust these garments to avoid their faces being exposed to the camera. Their fumbling and tinkering around resulted in minimizing, self-oriented body movements. While arranging, designing, and wearing proper disguises that support rather than impede the course of the robbers’ actions has been part of the preparatory tasks of robbers for a long time, the widespread use of CCTV makes this task even more crucial.
Weapons were used to emphasize forward and pointing arm movements especially when robbers proposed a threat. As extensions of the arm, weapons contributed to robbers’ aggrandizing postures. In doing so, they did not just enlarge their spatial presence but also prepared to invade the body space of the victims. At the same time, this is their claim to emotional dominance. Hence, weapons do not achieve compliance just by being possessed or displayed, but only as extensions of the robber’s body as it executes movements and postures that are experienced (by the robber) and recognized (by the victim) as being part of a robbery. Domineering postures seemed to be more associated with guns than with knives. Of the 11 robbers who carried a gun, we noted expanded body postures 11 times, while the 6 robbers who used knives displayed aggrandizing posturing only once (see Figures 4 and 5). Guns seem to expand the posture of the offenders more than any other weapon. When victims become aware of the robbers’ intentions, we arrive at the next stage, in which the robbers aim to enforce and maintain the robbery frame. First, however, we will show the results concerning the other tactic used in the first stage of the robbery.
A robber poses a threat in an aggrandizing posture. The way the robber holds the gun, with his elbow slightly bent, extending his arm both sideways and forward, contributes to an aggrandizing posture (still taken from case 10).
Typical way of holding a knife. The still shows how robbers typically use their knives, pointing at belly height as they pose their threat, which does not contribute much to aggrandizing posturing (still taken from case 25).
Disguise Tactic
In the disguise tactic, robbers attempt to take victims by surprise by acting as regular customers. We found this tactic in 7 of the 23 robberies. In five of these cases, the robbers acted as if they were about to make a regular shop transaction or asked victims for information. In the other two cases, the robbers only looked around the store before initiating the robbery. Robbers who presented themselves as regular part of the “shop setting” share body postures that we interpreted as lacking emotional dominance. First their gaze: In five of these seven cases, the robbers avoided eye contact with their victims (while the averted gaze of these robbers may result from their efforts to remain unrecognizable given the shop camera, it may still have the effect of being perceived as not very dominant in the eyes of victims). Second, in five of these cases, robbers kept their hands in their pockets when strolling around or approaching the victim. In these cases, the offender took out a weapon from these same pockets when they were close to the victim. In two other cases, the disguise tactic seemed even more counterproductive, given the aim of imposing a robbery frame later. Here, the robbers hesitantly walked back and forth in the shop with their bodies stiffened, gaze averted, and hands kept in their pockets once they made contact with the victims.
These body postures not only seemed discordant with the outward moving, expanding mode of bodily behavior that is characteristic of attempts to attain emotional dominance, they also did not fit very well to the normal customer–shopkeeper interaction, in which both parties generally make eye contact and in which customers show their hands, ready to receive or give. Nevertheless, not once a victim recognized and countered a robber before he would present himself as such.
Although we should be prudent in making numerical claims, the results suggest that, contrary to the speed and stealth tactic, robbers who disguised their intentions first seemed more likely to use knives rather than guns. To be precise, of the robbers who deployed the disguise tactic, four used knives, two guns, and one an unidentifiable object for smashing a display case. The use of knives in the disguise tactic also means that robbers need to move closer to the bodies of their victims to pose their threats.
Our analyses so far suggest that the disguise tactic creates more emotional and body work for robbers: They have to transform their fake and sometimes even hesitant customer postures to the compelling body movements that are needed to impose the robbery frame. Weapons seem to play an important role in this transition, but hiding them and then displaying them in a convincing way require substantially more emotional and body control as compared to the speed and stealth tactic. Our data also suggest that guns and knives have different effects on the interaction and are related to the tactics which were used. Guns, rather than knives, seem to provide more opportunities for offenders to make aggrandizing body movements and postures, while robbers with knives had to be in close proximity to the victim to pose a threat convincingly.
The Second Stage: Imposing and Maintaining the Robbery Frame
After the robbery is announced, robbers have to continue imposing their dominance throughout the remainder of the robbery. Our analyses suggest that two scenarios exist in the offenders’ quest for dominance in this second stage: A scenario in which the robbers gain and maintain total dominance throughout the robbery, and one in which their dominance is contested. These scenarios are not necessarily related to the success of the robbery from the viewpoint of the offenders. To give an example, even when offenders were able to dominate the whole interaction with a gun, it happened that robberies failed due to the fact that the valuables were not accessible or the robbers just left for unknown reasons. On the other hand, it also occurred that while robbers encountered heavy resistance from the victims, they still got the valuables in the end.
Total Dominance
Total, uncontested dominance of the robbers occurred in 12 of the 23 robberies. In 8 of these 12 cases, the interaction started with body postures that clearly signaled dominance by the robbers, followed by body postures that indicated submissiveness by the victims. The fragment below illustrates a typical case. When he is two meters away from her, he aims the gun at her face, arm stretched. She stands behind the counter and puts her arms in front of her chest, moving a bit backward. (case 10)
The following fragment describes how a robbery frame is kept in place with repeated use of body movements by the robber: Robber 1 walks to a cash register and hits a metal part of it twice. Immediately victim 1 looks at him and holds her head a bit back. She is sitting at the cash desk next to the one the robber had hit, approximately three meters away. The robber stands upright, looks straight at victim 1 and points his hammer at her face, his arm stretched. He then turns his back to her and walks a small circle, looking at robber 3 who came in behind him. Victim 1 feigns back a bit further.…Victim 1 slowly gets out of her chair and walks away. Now robber 1 looks at her and moves in her direction, again pointing his weapon. Then he shortly points at the cash register in the back, after which he aims it at her again. He now stands against the cash register, pointing his hammer at the chair in front of it. Victim 1 moves back to her chair and robber 1 stops aiming his hammer and fumbles to open a bag. (case 19)
In 9 of the 16 cases in which robbers used the speed and stealth tactic, they were able to maintain dominance throughout the whole robbery using mostly guns (six) but also knives (three) and in one case a hammer. In three of the seven disguise tactic cases, the offenders did not encounter any resistance and enforced and maintained the robbery frame successfully in the second stage by using a single threat with a knife, a gun, or an unknown object.
Contested Dominance
In 11 of the 23 cases, victims contested the robbers’ attempts to impose a robbery frame. We identified two types of situations. First, in five cases, the robbers did not succeed in attaining emotional dominance initially as expressed by their own body postures and/or weapon usage. Second, in five cases, the robbers failed to impose a robbery frame altogether—they did not make clear to the victims what was happening: Either robbers did not give sufficient time to the puzzled victims or they went straight for the valuables without confronting the victims. In these latter situations, victims did not show signs of submissiveness because they did not seem to comprehend the situation. The following fragment shows one of the five cases in which a victim initially resists the robbers’ attempts to impose a robbery frame: Suddenly the robber draws a small knife, holds it in front of his belly and moves closer to the victim looking straight at her, his head forward. The counter is between them. The victim stays put, holds her ground, and reaches out with one hand for a split second, countering the knife. Now the robber repositions the knife at head height and leans forward again which makes the victim move back a bit and raise her hands in front of her chest, palms towards the robber. (case 4)
The other cases demonstrate that imposing a robbery frame crucially requires a working agreement between robbers and victims. Here, the robbers seemed too eager to impose dominance without noticing that the victims did not understand what happened to them. Consequently, they did not show the expected signs of submissiveness. This happened in the following example. The robber enters the scene with his gun drawn: He points the gun at victim 1 while walking in on him fast. Victim 1 does not seem to respond, he keeps exactly the same posture he had before robber 1 entered the scene. Together with victim 2, victim 1 stands behind the cash desk. Also, victim 2 does not seem to react and stands with his hands behind his back. Robber 1 now points his weapon at victim 3, who stands in the middle of the store. Victim 1 takes a step back as if his reaction to the experienced gun threat was delayed. Robber 1 reaches victim 3, puts the gun on the temple of the victim’s head and pushes him backwards violently with his gun. Victim 3 initially only moves back under force, but then he grabs the robber and resists. They start a struggle in the back of the shop. (case 8)
As the sample size is small, we cannot rely on statistical measures to test relationships. However, our qualitative exploration does show some tendencies we want to highlight here. First, most weapon threats support the forward moving by the offenders. Second, it seems that guns allow for expanding body postures more easily than knives. As we have noted above, robbers tend to hold knives at belly height when they approach their victims, while guns are more often displayed at shoulder height, requiring more extension of the robbers’ arms and thus expanding their bodies. Thus, apart from their more lethal power, the appropriate handling of guns requires a larger body space, which provides additional support to robbers’ attainment of emotional dominance (see Figures 4 and 5).
Conclusion
This is one of the first small-scale and explorative studies of robberies using video data. It shows that for a robbery to be successful from the viewpoint of the robbers, emotional dominance is required to impose a working agreement on the victim. Robbers attain emotional dominance by combining expanding body postures and forward body movements with weapons that afford and stimulate these movements. Robbers who use the disguise tactic to move into copresence with the victims face the challenge to transform the situation into an asymmetrical one as they try to impose the robbery frame. This proved to be a difficult task for some robbers, especially when they were armed with knives because this weapon seems less suitable for making targeted expanded body postures. Robbers who moved rapidly into copresence with the victims as they came running in with weapons ready, mostly guns in this tactic, faced a different challenge. They had to give the victims time to adapt to the robbery frame they were suddenly proposing. Guns, rather than knives, seemed to fit more easily with aggrandizing posturing, thus offering more support to robbers’ claim to dominance, in addition to their more lethal power. In the cases where resistance occurred, robbers either did not succeed in attaining emotional dominance, as expressed by their insecure movements or minimizing postures and related weapon usage, or they failed to impose a robbery frame on puzzled victims who did not seem to comprehend the situation. We conclude that the forcefulness of armed threats to establish a robbery frame is related to the types of weapons used and to what extent these weapons afford aggrandizing body posturing. In order to explain why some robbers with a gun fail and some with less lethal weapons like hammers, knives, screwdrivers, or even bare hands succeed, the role of weapons should thus be seen as part of the whole array of bodily strategies to impose the robbery frame.
Perhaps the most important conclusion is that video analysis opens up a new, close-up microperspective on the interactions that make up violent crime: One that perceives them as sequences of body movements. Future studies that want to follow this path might benefit from the following methodological notes. First, the density and complexity of video data seem to be the prime obstacle for analyzing it properly. We propose four suggestions to make sense of such a wealth of information: (a) become familiar with the data by watching it multiple times to get a sense of the general flow of interactions; (b) work in a team, compare independently, made observations, and organize joint video sessions; (c) clearly specify what specific modes of behavior are to be analyzed based on the literature and initial observations; (d) when coding the data, focus, with each run-through, on a specific mode of behavior. Second, one disadvantage of the video footage used here is that they lack audio. Audio could have been an important additional indicator for observing the emotional signaling and impact of certain body movements and posturing. Third, while we did not pay attention to the use of physical space by robbers and victims, this seems a relevant issue to consider in more detail. Video analysis of how robbers and victims use entrances, exits, counters, display shelves, and other objects can reveal how the physical setup influences the sequences of body movements that make up violent crime. Finally, we recommend that future studies complement video data with written information of various sorts on what happened (e.g., judicial case files, news reports, or interviews). This is not only useful to verify the observations but also to gain a better understanding of all body movements that make up the situation as an integral hermeneutic unit.
We end with an ambitious suggestion for future studies. Given the contradicting results of prior studies of weapon use by Dutch and U.S. researchers, we think it is worthwhile to analyze weapon use as part of the bodily dynamics in comparative, larger scale video analyses of robberies in various societies and settings. Such comparative project should involve both qualitative, interpretative analyses and the fine-grained coding that allows to statistically assess (preferably sequence analyses given the real-time data) the relationships between various weapons and forms of body posturing of both victims and offenders.
Footnotes
Appendix
| Code | Description | Interpretation |
|---|---|---|
| Weapon visible | Weapon visible for other actors | Sign of dominance |
| Weapon threat | Weapon is directly able to injure: pointed gun/raised weapon/ready to throw or stab | Sign of dominance |
| Weapon use | Weapon is used with the intention to injure/dominate/force the other party | Sign of dominance |
| Physical force grab | Touch/hit/push/grab | Sign of dominance |
| Minimize posture | Body language: avoiding gaze (head down/bowed), postures and movements that are shrinking, head ducked down, hands cover face, chest inward, and shoulder slumped forward. Tries to hide | Sign of submission |
| Maximize posture | Body language: strong physical presence by making oneself large, for example, by standing erect and/or with hands on hips. Bodily openness. Pride; small smile, head tilted slightly (approximately 20 degrees) back, and expanded posture | Sign of dominance |
| Hands up | Hand(s) up or in front of chest; open hands. Open hand toward another actor, sign of self-protection | Sign of submission |
| Move away | Moves back or give ground to another actor. | Sign of submission |
| Move toward | Move face-front toward another actor | Sign of dominance |
| Compliance | Acts consciously in accord with the other actors’ will (give values, lay on the floor) | Sign of submission |
| Resistance | Acts consciously against the other actors’ will (give values, lay on the floor) | Sign of dominance |
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.