Rampage violence has become a social problem of increasing concern in the 21st century, especially in the United States where more public mass shootings occur than anywhere else on the planet (Lankford, 2016, 2019; Lemieux, 2014; Madfis & Lankford, 2022; Silva, 2022). Further, public mass shootings have become more frequent (Peterson & Densley, 2020) and more deadly (Lankford & Silver, 2020) since the turn of the 21st century, and both public and private mass shootings have increased over the last few years during the COVID-19 pandemic (Schildkraut & Turanovic, 2022). As the problem has worsened and gained more public awareness over the last 20 years, the research on mass/rampage violence – public mass shootings, in particular – has also increased exponentially during this time (Duwe et al., 2021).

Traditionally, mass shootings have been defined as the killing of four or more people with firearms during a single episode at one or more closely related locations (Fox et al., 2018). However, this definition remains contested (Huff-Corzine & Corzine, 2020), and numerous similar terms can be found throughout the literature, such as mass murder (Duwe, 2007; Holmes & Holmes, 2001; Levin & Fox, 1985; Meloy & O’Toole, 2011), active shooting (Kissner, 2016; Rusho et al., 2021), mass violence (Daly, 2018; Huff-Corzine & Corzine, 2020), rampage (Madfis, 2014, 2020; Newman et al., 2004), and massacre (Harrison & Bowers, 2010; Schildkraut & Muschert, 2014). There are many more location-specific terms such as mass public shooting (Silva & Green-Colozzi, 2022), workplace massacre (Duwe, 2007), familicide (Liem at al., 2013), and mass school shooting (Curran et al., 2020; Paez et al., 2021). When considering terminology regarding how far into the act of mass violence the perpetrator was when it ceased, the list of terms grows further. Scholars have written about averted school rampage (Madfis, 2020), averted school violence (Cowan et al., 2022), averted school shooting (Daniels & Page, 2013), averted mass violence (Cowan et al., 2022), thwarted mass homicide (Sarteschi, 2016), and mass shooting with the designations of completed, attempted, failed, and foiled (Silva, 2021a, 2021b; Silva & Green-Colozzi, 2022), and school shootings as either completed or averted (Winch, 2021).

While some of the variations in terminology may be substantive (and entail important differences in offender profiles, attack behaviors, motivations, opportunities for intervention and prevention, etc.), far more research is needed to compare across these categories (Madfis, 2020). That said, we do know that many common factors and patterns among perpetrators exist (such as cumulative strains, extensive fantasizing about and planning for the rampage attack, desire for lasting fame, leakage of information about the threat to others, etc.; see Lankford et al., 2019; Levin & Madfis, 2009; Madfis, 2017) and persist despite, for example, variations in the site of the mass violence (see, for example, Silver et al., 2018).

As empirical knowledge of mass violence expands it would be helpful to utilize universal language to explore this violent phenomenon with a more standardized set of terminology. The increasing frequency of mass violence in the United States and its damaging effects on society suggest the dire need for empirical research which is more easily understandable and less contradictory in the use of language so that research may better inform public debate and policy. Scholars’ use of more universal and uniform language is one method to potentially address this need. Though some scholars (such as Huff-Corzine & Corzine, 2020) discourage the implementation of any standardized definitions around the phenomenon due to the desire to keep various forms of mass violence distinct from one another, we contend that more universal language and terminology for these acts can benefit this area of study if developed in a systematic way with holistic intent.

In this article, we summarize how previous studies have defined and operationalized mass violence (and associated terms such as mass shootings, rampage, mass murder, etc.) across the spectrum of attempts and completed acts and argue in favor of adopting terminology which is as broad as possible, yet retains the most crucial identifying features so as to maintain distinctions between single-incident mass casualty events and other multiple-incident mass casualty events like serial killings and state-sponsored mass violence such as genocides. To that end, we suggest adopting the broad term rampage to which specificity can be added. For example, the location and weapon of rampage violence may be further specified by referring to a school rampage stabbing or a workplace rampage shooting. Further, the completion status of an attack may be characterized as a completed, attempted, interrupted, arrived, prepared, planned, or researched rampage to clarify how far along the offender was able to carry out their deadly plot (these new terms will be defined and explained later in this piece). Thus, instances could be referred to as, for example, a completed school rampage stabbing or an attempted workplace rampage shooting.

Prior Debates Over Terminology

In scholarship on homicide, the first term used to refer to single-incident mass casualty events was “mass murder” (Fox & Levin, 1998), and the traditional definition limits the phenomenon to those events wherein at least four victims were killed during a single episode at one or more closely related locations (Duwe, 2007; Fox et al., 2018; Holmes & Holmes, 2001). The term mass murder is deliberately broad so as to encompass all types of weapons used, including knives, axes, guns, bombs, and any other potentially deadly weapon. In recent years, terms such as “mass shooting” and “active shooting” have become far more prevalent (Fox & Levin, 2015). The aforementioned terms focus only on gun violence, and many studies specifically limit their qualifying criteria to only include incidents of mass violence in which firearms were the primary weapon (see Newman et al., 2004 for an example). As Madfis (2020) notes, however, in his study of averted incidents of school rampage plots, numerous deadly school violence plans were formed by students which involved knife attacks and the use of various types of bombs, explosives, and incendiary weapons. Likewise, bombs were a vital component of the Columbine High School killer duo’s plan though most were not detonated (Larkin, 2007). Further, the deadliest mass murder at a school in American history (the 1927 Bath School Disaster in Michigan) involved dynamite and hundreds of pounds of pyrotol, but no firearms (Bernstein, 2009), and the deadliest mass casualty event committed by a single perpetrator (the 2011 Norway attacks) utilized both bombs and guns. Amman and colleagues (2022) recently found 138 mass stabbing attacks across the world between 2004 and 2017.

Accordingly, we suggest using the term rampage in line with Madfis (2020) so that all types of weaponry may be included in future operationalizations and subsequent datasets. Certainly, guns enable far more deaths in a shorter timespan than less lethal weapons like knives and axes (Madfis & Levin, 2013), but, until such time as there is evidence to suggest that substantial differences exist in terms of motivation or other characteristics between those who utilize guns and those who utilize other weapons to commit their attacks, the broader term rampage should be adopted to allow for the inclusion of more incidents, larger samples that are more representative, and thus more comprehensive analyses.

Other important definitional debates in the mass violence literature have included excluding cases based on the location of the attack (e.g., removing cases of familicides in private homes), offender’s connection to the attack site (e.g., limiting cases to school attacks only by former and current students), the relationship between offenders and victims (as specific or random targets), and by the number of victims killed or injured (Daniels et al., 2007, 2010; Levin & Madfis, 2009; Moore et al., 2003; Muschert, 2007; Newman & Fox, 2009; Newman et al., 2004; Vossekuil et al., 2002). We view all of these exclusion criteria as problematic (and as having the potential to create less representative samples), unless done for specific methodological, conceptual, or pragmatic reasons.

In particular, we take issue with the practice of requiring specific death counts as an inclusion criterion. Such a focused definition needlessly omits myriad cases which have not resulted in multiple deaths (Madfis, 2020). Cases in which the perpetrator has the desire and intent to kill multiple victims yet fails to do so when their plans come to the attention of authorities before being effectuated and those who severely injure many people but fail to do so fatally are excluded in conventional definitions of mass murders and mass shootings. As Madfis (2020) states, “These less-successful perpetrators may be distinguishable by their overall incompetence with weaponry and/or their inability to maintain secrecy about their future actions, but not necessarily by their original motivations and goals” (p. 169). Therefore, there is much use in adopting the broader notion of rampage violence and including all cases which involve an attempt to kill multiple people in a single event (White-Hamon, 2000; Madfis, 2020).

One of the newest and most interesting areas of emergent scholarship in this area is the issue of completion status, or how far along the perpetrator was able to carry out their deadly plot to kill multiple victims. The first studies to explore cases of averted incidents (such as Daniels & Page, 2013; Daniels et al., 2007, 2010; Larkin, 2009; Madfis, 2014) exclusively did so in the context of school settings, but knowledge and terminology in this area has rapidly expanded in recent years.

As Daniels and Page (2013) identify, studying something that may or may not have eventually happened but was stopped is complicated. To combat this, they took an approach inspired by the study of law, selecting only cases which had sufficient evidence to bring about conviction for an imminent shooting (Daniels & Page, 2013). Winch (2021) considered a case averted if there was any evidence that the perpetrator demonstrated pathway engagement (Calhoun & Weston, 2003), such as researching past school shootings, planning for an attack in various ways, or preparing for the attack by acquiring a gun. Other definitions and use of terminology regarding completion status also consider activity engagement and process. Meloy and O’Toole (2011) explore leakage, which is a type of pathway warning behavior wherein there is “communication to a third party of an intent to do harm to a target” (p. 513). These researchers found pathway warning behaviors to be a predominant theme in several types of violent crime, including mass murder and school shootings. According to Meloy and O’Toole (2011) pathway warning behaviors are any that contribute to the steps of researching (learning and gathering useful information from past attacks), planning (selection of the time, location, method of attack, and entry method), preparing for (weapons and materials acquisition), and implementation of a violent incident, and that with each step the risk of violence and harm to others increases. Incorporated into Winch’s (2021) conceptualization, each of these pathway warning behaviors up to implementation meets the criteria for an averted school shooting. Once a perpetrator reaches the implementation stage, the case was deemed a completed school shooting if there was at least one injury and the incident occurred on school grounds during school hours or at a school event after hours.

Silva (2021a, 2021b) introduced a series of terms defining the progress of a mass shooting incident based on outcome operationalization. This work defined a mass shooting as a gun incident committed by 1+ person in 1+ public or populated setting within a 24-hour period in which at least some intended victims were selected randomly and/or for symbolic value. This model expands on previously used completion status concepts, offering four casualty- and target-based outcome categories and terms from least to most severe: foiled, failed, attempted, and completed (Silva, 2021a, 2021b). Offering further clarification on stages of completion status, Silva and Green-Colozzi (2022) identify four stages of a mass shooting as preparation, arrival, event, and conclusion.  

A foiled mass shooting is an incident which the perpetrator has set into motion but was stopped before the incident began, such as never reaching their intended target (Silva, 2021a, 2021b). This definition is rooted in Sarteschi’s (2016) concept of high credibility cases, in which an offender has explicitly stated their intention to carry out a mass shooting attack, identified a target or type of target, and already has a gun, has attempted to acquire one, or has a plan to do so. An incident is considered a failed mass shooting if the plan was set into motion but stopped during the incident (Silva, 2021b). For example, in a failed mass shooting case, the perpetrator would have finished preparations, arrived at the intended target, and perhaps began shooting but is then interrupted (e.g., killed or tackled) (Capellan & Gomez, 2018; Silva & Green-Colozzi, 2022). Both foiled and failed mass shootings have no casualties (Silva, 2021b).

Conversely, attempted and completed mass shooting incidents both include casualties. The distinction between these two categories rests on the number and type of casualties associated. In an attempted mass shooting, the plan has been effectuated and ended with less than four deaths but at least one victim casualty (injury or fatality). A completed mass shooting is characterized by four or more fatalities, not including the perpetrator (Silva, 2021a, 2021b).

Silva’s (2021a, 2021b) outcome operationalization model and Meloy and O’Toole’s (2011) stages based on pathway warning behavior overlap, each adding specificity to different phases of an incident. We submit that, taken together, these approaches can offer unifying terminology and definitions that can be applied to all incidents of rampage violence, no matter the location or weaponry, and such an approach provides the most comprehensive way to provide information and analysis on the completion status of incidents. At present, Silva (2021b) created the stages of preparation, arrival, event, and conclusion, while Meloy and O’Toole (2011) specify research, planning, preparation, and implementation stages. As illustrated in Figure 1, the definitions of these stages overlap, allowing Meloy and O’Toole’s research, plan, and prepare stages to add a deeper level of detail to Silva’s preparation stage. Further, Silva’s arrival, event, and conclusion stages increase the detail and nuance of Meloy and O’Toole’s (2011) implementation stage.