6.1 The BHS Research Questions in Light of Its Findings

Clearly, with the BHS and its first findings, as presented in this book, a considerable gap in European homicide research has been filled, while shedding first light on the phenomenology of (lethal) violence in this part of Europe. Although the empirical vacuum in regionally comparable homicide research has thereby been shrunk, the detected theoretical vacuum still needs to be addressed. The same goes for the empirical investigation of the power to define violence. Despite the BHS having captured the normative definition of (lethal) violence by police and its consequent redefinition by prosecutions and courts, due to lack of data from police files, missing data issues and having adopted a “narrow” concept of (lethal) violence excluding non-homicidal offenses, this line of research remains largely unanswered and calls for further criminological attention.

The BHS’s second line of inquiry, however, was successfully addressed and with the presented first findings provides for an original empirical glimpse into the realities of (lethal) violence in the Balkans. Slightly deviating from the initial cooperation agreements covering core countries of the Balkans,Footnote 1 the study managed to capture the phenomenology of (lethal) violence in six Southeastern European countries: Croatia, Hungary, Kosovo, North Macedonia, Romania, and Slovenia. This provides a strong basis for starting to unravel the Balkan-violence-paradox, while enabling critical questioning of common violent Balkan images and stereotypes.

So, in a nutshell, most of our research questions have been addressed successfully with the BHS, and many of them could already been answered. Nevertheless, many of our research questions require further country-focused and topical in-depth analysis as well as future research, while it is safe to say that in the process of conducting the study, we managed to raise more questions than answers provided. In addition to the great general scientific value of the study and its empirical findings, on a more individual note the BHS has also been a tremendous(ly exhausting) learning experience and thus an eye-opener with regard to the current criminological (un)measurability of (lethal) violence. Being interested in finding out how (lethal) violence displays itself in the Balkans, I vividly remember the disappointment I felt while skimming through the first datasets, fully realizing that overall we were not capturing, let alone measuring, the realities of violence, but rather counting some sort of its artificial (normative) construction.

If one is interested in finding out whether or why one country or region is “more violent” than another, as differences in homicide rates might imply, then it is not enough to simply look at incident, offender, or victim characteristics, let alone normative classifications, typologies, or mode of operandi. One needs to somehow capture and measure the actual violence that occurs in each incident and thereby focus on the tangibles of violence, not their (re)interpretation by various different actors, oneself included. In essence, one would need to capture the mere physics of violence and then come up with a purely criminological weighting and classification system of violence. Although with the BHS itself we did not even attempt to take this path, in the process of conducting the study, in fact, due to conducting the study and analyzing its data, this highly intriguing task emerged and is thus currently being explored by the Violence Research Lab.

6.2 The Power to (Re)Define and Deal with (Lethal) Violence

The BHS’s first line of research into criminal justice actors’ power to define violence has been discussed extensively throughout all chapters by providing numerous examples and highlighting its main conceptual, methodological, and practical research implications. Nevertheless, the empirical investigation of the “hows” and “whys” when it comes to the normative and social construction of violence and its (re)definition by different criminal justice actors calls for further research. In particular, by broadening data sourcing to include police files, while focusing not “only” on (attempted) homicidal violence, but by broadening the focus toward a much wider range of non-homicidal, yet clearly violent offenses (e.g., assault, robbery, rape with and without lethal consequence) and violent incidents which formally do not even constitute a criminal offense (e.g., justifiable killings).

Thus, if one aims at fully understanding the normative construction of lethal violence, then special attention should be paid to the dark figure of homicides as well as homicide drop-outs. European research has shown that a rather large share of (potential) homicides remains well hidden among natural (unsuspicious) deaths. It seems very likely that this is also the case in Southeastern Europe – at least for Croatia preliminary inquiries confirm a high likelihood for this.Footnote 2 The BHS findings show that once a homicide case, attempted as well as completed, gets investigated by the prosecution, the share of case drop-outs is extremely low. The same applies for the share of case drop-outs on the court level, whereas the share of case drop-outs on the police level remains largely unknown. Since for criminal offenses in general the share of drop-outs of reported cases is largest on the police level, it is plausible to assume that the same might apply for completed homicides and even more for attempted ones. This matter deserves further research attention and a much stronger focus on initial (lethal) violence detection, reporting, police practices, and criminal investigations.

Though our first line of inquiry into the power to define (lethal) violence focuses primarily on criminal justice actors, it is sensible to discuss the role of researchers at this point as well. We also have the power to define (lethal) violence and rather frequently make use of it, arguing for and against various approaches based on conceptual, methodological, and practical reasoning. With the BHS, we conceptually defined violence as the human infliction of physical harm upon another person, whereby the adjective “lethal” denotes that as a consequence another person has died. Such an approach is very narrow, as it focuses “only” on the undisputable core of violence: the infliction of physical harm. Yet it is at the same time also extremely broad, as it is not limited by any subjective element on the side of the offender (e.g., intent or motive) or the victim (e.g., consent or provocation), nor by various normative constructs (e.g., criminal responsibility or justifiability). Nevertheless, practically we sampled “only” lethal and non-lethal homicidal violence and discovered that with regard a rather large number of variables there appear to be no markedly differences between attempted and completed homicides. However, since we also found certain key variables which display noticeable differences when comparing attempted and completed cases, it remains unknown to what extent and how the lethality of a violent incident is related to different situational, criminogenic, or victimogenic factors and either good or bad fortune on the side of the victims.

After having sampled a total of 2073 cases involving 2416 offenders and 2379 victims of (lethal) violence, only 3.7% of sampled cases, 3.9% of offenders, and 3.4% of victims dropped out from our analysis due to lacking a violent element. This alone is a strong indication that attempted homicides are not simple cases of “nearly” or “maybe” homicidal violence, but that they might in fact be far more violent and severe than one would have initially assumed. This finding is further confirmed by the rather severe sentences imposed on attempted homicide offenders (98% prison sentence for males and 94% for females), which is in line with the type of sentence in case of completed homicides (98% prison sentence for males and 94% for females). Another indication of the severity of attempted homicide cases is the quite long average duration of the imposed prison sentences, with a mean of approximately 5 years, compared to completed homicides, with a mean of approximately 14 years. Another strong indication of the severity of attempted homicide cases which we detected with the BHS is the harmfulness of suffered violent victimization. The vast majority of BHS victims, actually more than three fourths, suffered heavily bodily injuries (42.7%) or death (34.8%), compared to only 16.5% who suffered light bodily injuries or even fewer 6.1% who suffered no injuries at all. Given that the ratio of completed vs. attempted homicides among the victims is 39.7 to 57.4 one would have expected a far larger share of victims with less severe or no injuries that comes much closer to 57.4, than does the detected share of 22.6%.

The BHS findings do not indicate that homicides are a unique phenomenon, something essentially different than lethal violence or the deadly outcome of violence. Based on our findings, as well as the compelling conceptual and methodological argumentation provided in the previous chapters, the exclusion of attempted homicides from homicide studies appears generally unjustified and reasonable merely due to practical research considerations and limitations. Therefore, it might be fundamentally wrong to approach the question of homicide or (lethal) violence as a dichotomy. We might rather want to approach the matter in terms of gradations or scales of a violent incident’s “homicidality,” as will be further discussed in the final section of this chapter (Sect. 6.4).

6.3 The Phenomenology of (Lethal) Violence in the Balkans

Now, what did we actually find out about (lethal) violence in the Balkans? The BHS data shows that there are no major differences between attempted (58%) and completed (42%) homicides regarding the place where the incidents occurred, their micro-locations, the time of day of the incident, the period of the week, the number of offenders and victims, (non)premeditated commission of the offense, the offenders’ main motives, the general victim-offender relationship, and the offenders’ alcohol intoxication tempore criminis. Most (lethal) violence in the BHS occurs in rural cities, closely followed by urban ones, with the least taking place in the countries’ capitals that are also the biggest cities in each of the sampled countries. The majority of (lethal) violence is committed in a private micro-location in the evening (18.00–24.00) and afternoon (12.00–18.00) on the weekends and the day prior/after the weekends. The vast majority of (lethal) violence involves only one offender and only one victim. Slightly more incidents are premeditated than committed affectively, with approximately half of the offenders’ motives remaining unclear, followed by revenge and greed. The victim-offender relationship is commonly one between non-strangers, whereas approximately the same share of offenders was alcohol intoxicated tempore criminis in case of attempted as well as completed homicides.

We did however observe noticeable differences between attempted and completed homicides when looking more closely at the specific victim-offender relationship and the type of (lethal) violence. Completed homicides are commonly found among domestics (between intimate partners and core family members), whereas in case of attempted homicides a non-domestic victim-offender relationship is most frequent (between friends/acquaintances and strangers). This somewhat more pronounced lethality of violence among domestics is well reflected in the most frequent types of violence in case of completed homicides, where other private violence clearly makes up for the majority of cases. Looking at attempted homicides we see that slightly more frequently the violence type is that of other public, closely followed by other private violence. Now, it might be that the lethality of a violent incident is under the impact of the incident’s micro-location (private), or the victim-offender relationship (domestic). Probably it is a combination of both, as well as the impact or certain offender and victim characteristics (gender, age, specific alcohol intoxication combinations of offenders and victims, etc.), that might help explain and potentially even predict the lethality of violence. Cautiousness is however in place, since the BHS findings are under the (so far non-assessable) influence of sampling decisions (e.g., excluding assaults, robbery, and rape, with and without lethal consequences), various normative constructs (e.g., intent and negligence or justification), an unknown dark figure, and missing data to name but a few.

As throughout Europe, (lethal) violence between complete strangers is the exception in the BHS (less than 15% of cases). Most stranger-violence can be classified as other public violence, followed by thievery and bar violence. The majority of (lethal) violence in all countries except for Hungary involves non-domestics (friends/acquaintances), whereby the findings for Hungary (more domestics) might be explained by the comparably atypically high share of female offenders (17%), for whom we in the BHS overall found to most frequently commit other private and separation/divorce (lethal) violence. Stranger-violence in the BHS is predominantly committed by males, whereas out of all offenders only 15% of males committed intimate partner violence, compared to as much as 39% of females committing intimate partner violence. Clearly, in the BHS males more often commit (lethal) violence, but in doing so mainly act out violently against males. Females however, although less frequently committing (lethal) violence, when acting out violently do so mainly against males (72% of female offenders) and in the form of other private and separation/divorce violence.

Despite the easier availability of firearms, due to the region’s recent war-legacy, we found that in the BHS only 13% of offenders used firearms, compared to as much as 62% who used cold weapons and even 21% who used no weapons at all. Similarly, the organized crime violence nexus could not be confirmed by BHS findings, as even after adopting a very broad concept of organized crime related (lethal) violence, only 1.6% of all offenders could be at least vaguely linked to the criminal underworld. This is in my opinion more likely to be a consequence of unknown offenders in such cases, resulting in their no-show in the BHS sample (no prosecution against unknown persons), than an accurate reflection of the realities of (lethal) violence in the Balkans. Further systematic research, for example, analysis of media reports and expert interviews with organized crime and homicide investigators would be needed in order to confirm this assumption.

Interestingly, in particularly cruel cases of lethal violence, although extremely rare in the BHS, the offenders are less frequently found to be insane or of diminished criminal responsibility, whereas the victims are predominantly female (62.5%). This share of female victims is more than twice as big as in case of non-cruel lethal violence (28.5%), where the majority of victims is male (71.5%).

Coming back to the Balkan-violence-paradox, a phenomenon where we see comparably higher homicide rates throughout the region, although the Balkans do not fit a high crime profile, the BHS findings indicate that the paradox might perhaps be explained by more frequent other private and intimate partner (lethal) violence. Whether and how this should be interpreted in the context of Balkan-typical patronage networks remains dubious at best and needs further investigation, especially with regard to reporting and investigating cases of (lethal) intimate partner violence and possible “influences” on legal qualifications, prosecutions, charges, and convictions. Here undoubtedly all the relevant criminal justice actors (police, prosecution, and courts) have considerable latitude in their qualifications and decisions which might be under the influence of patronage networks. Put a bit more frankly, it would be highly surprising if this area of criminal justice practices, just as in case of organized crime-related (lethal) violence, should prove immune to the influence of patronage networks, while the judiciary in general, alike other public institutions, is commonly found to be highly corrupt throughout the whole region.

Now, what about the violent Balkan images and stereotypes? The BHS findings do not provide grounds for such stereotypes, at least not when looking at the types of (lethal) violence and the rather exceptional occurrence of honor killings or hate/discrimination violence. Even bar violence, accounting for approximately the same share as thievery violence, is a rather seldom occurrence when compared to intimate partner violence. Contemplating about the genesis and persistence of these violent Balkan images, particularly like the one presented in the example (Sect. 3.1), I wonder whether both might not be rooted in diverse cultural aspects and lacking insights into what is regionally considered customary and appropriate behavior in contrast to what is regarded as offensive and even provocative. While the refusal to have one bought a round of drinks in a bar might appear completely trivial to someone outside the region, someone who probably not even (fully) comprehends the meaning and importance of such a gesture within the broader context of a regional culture of hospitality and honor, locals will commonly perceive such a refusal as a public slap in the face and a grave insult and provocation. Consequently, someone lacking knowledge of or empathy for such regional or local cultural aspects might easily jump to the conclusion that people from the Balkans and even Southeastern Europe, who shoot at another person for refusing a drink, are “easy on the trigger” and “bloody-minded.” But if one were to interpret the same violent incident by understanding the actual meaning of refusing a drink as a public slap in the face, this trigger might not appear so trivial anymore, even to someone from outside the region. Obviously the investigation and understanding of (lethal) violence, just as any other criminal and therefore also social phenomenon, must be imbedded in its cultural context in order to evade void interpretations that may lay the path for unfunded stereotypes. Similarly, and in line with what Liem (Chap. 2) calls for, not only (lethal) violence needs to be embedded in its cultural context, but homicide research as well. Various cultural aspects in violence/homicide research need to be discussed more frequently and much more transparently, and should probably even be far stronger accounted for, particularly if we aim at successfully joining forces in a truly European shot at homicide research.

6.4 On the Definability, Measurability, Severity, and Homicidality of Violence

Unsurprisingly, much of what we currently know about (lethal) violence, just as the findings presented in this book, depends on how and why we look at it. Some would argue that violence is any kind of (non)action causing any sort of harm to any living being. Others, myself included, would argue that violence is the human infliction of physical harm upon another person. Yet others would add to this the need for intentionality, criminal liability, or lack of justification on the side of the perpetrator and a lack of consent or provocation on the side of the victim. Thus, one might want to distinguish between violence with a lethal consequence (lethal violence) and homicides, or rather not. This is not merely a minor variation in different approaches to (lethal) violence or (attempted) homicides, but a fundamental conceptual discrepancy, a Balkanization of concepts, which undoubtedly also impacts violence and homicide research, as much as the knowledge it produces.

I wonder what the current state of art in property crime research would look like if we were to start off from comparably distant or Balkanized concepts of the terms “property” or “stealing”... we would probably end up at considering all types of crime as some manifestation of stealing, like stealing someone’s life, sexual freedom, and personal dignity. Although legit reasoning lies behind it, such diffusion of different phenomena under the umbrella of property crime would almost certainly create more confusion and misunderstandings than actual criminological knowledge. Violence and its lethal extreme is a tangible, capturable, and measurable phenomenon that occurs in reality. As such, it can and therefore should be conceptualized and then defined based primarily on objective and empirically measurable elements. Normative constructs most certainly are not among such objective and empirically observable or measurable elements. Criminology in general would do well by developing its own “criminal vocabulary,” just as (lethal) violence/homicide research might want to discuss more vigorously its fundamental concepts and definitions of violence, independently from preconceived disciplinary paradigms, practical conveniences or political agendas. The operational implementability of a purely criminological concept and definition of (lethal) violence is a consequential, but also solvable methodological issue, and therefore no excuse for the ongoing diffusion in violence research.

Besides the apparent conceptual and definitional challenges violence/homicide research is facing, there is also the challenge of measuring violence and accounting for its varying severity and harmfulness. The question is not if violence can be measured, but rather how such measuring can be achieved objectively and by relying solely on empirical facts, rather than normative constructs or pure speculations about motive, intent, or comparable subjective elements of an offense. By blending out such constructs and speculations about the subjectivities of (lethal) violence and its teleological severity we should be able to develop an authentic criminological violence classification system. A precondition to this is figuring out a universal measure of violence that should be based in and designed out of the physics of violence itself, not its social or normative or even disciplinary perception and (re)interpretation. Here first and foremost criminology with its inherent transdisciplinary nature is called upon to determine and weigh the physical variables of violence, such as the force needed to cause a certain injury, while accounting for the fragility of the attacked body, the duration of the attack, the force-multiplying or -mitigating effect of a used tool or weapon, the pain inflicted on the victim, and the attacked body part’s vitality to mention but a few.

Ultimately, a universal measure for violence could be used to design a scale of “homicidality” that would lead us out of the current include-exclude dichotomy in homicide research, while enabling us to look at (lethal) violence more accurately, more realistically, and much more meaningfully. Clearly, this is a task involving experts and researchers from “distant” disciplines in a setting that far surpasses what we commonly consider interdisciplinary. Working on the physics of violence and in an attempt to develop such a universal measure of violence and weight its “homicidality,” in the Violence Research Lab, we spent countless hours discussing violent incidents and homicides with physicists and doctors of forensic medicine. So far we have learned that the task is extremely complex, requires transdisciplinarity, and is still often times simply too much for our own disciplinary limitations. Nevertheless, it has proven worth the effort and has already provided us with a completely new take on (lethal) violence and what violence research could and should look like, as well as what the value and implications of probable findings might be. But with this we already enter a new topic and an entirely different discussion which should be the subject of another book.

To conclude, the BHS and its first findings, as presented throughout the previous chapters, are an utmost important first step in understanding (lethal) violence in the Balkans. As such, it has clearly not solved any of the grand mysteries of homicidal violence in Southeastern Europe. However, if it has managed to raise at least some good questions, while making available original and comparative empirical data on (lethal) violence from six countries of the region, then the goal of this book has been reached.