MegaMullet: Dynamics of Atrocity Scripts in Conflict

Spread the love

TLDRUpFront: A MegaMullet addressing the flood of atrocity scripts on social media: What are the different kinds of scripts used in local and foreign narratives? Why are we seeing more now? Why would actors share them? What do they hope to gain? What are best practices for navigating these materials safely?


(Correction: Although Bangladeshi fighters fought alongside Palestinians as early as 1948 and fought alongside the PLO in the 1980s, an earlier version of the article incorrectly stated that PLO fighters likewise supported Bangladesh’s War of Liberation. This was incorrect, as only political support was provided by the PLO to Bangladesh. This has been corrected with additional sources added in the ‘Glorify vs. Horrify’ section.)


Cultural scripts are units of meta-language transmitted across all mediums: written, video, audio, and conversation; conveying information that forms narratives. In the Hamas invasion of Israel, we’ve seen the resurgence of atrocity cultural scripts; alleged and actual footage of heinous acts. There’s a host of questions related to atrocity scripts that I’m going to get into in this MegaMullet. I studied cultural scripts, including atrocity scripts, as part of my research on radicalization. Although the timing was not intentional, a peer-reviewed research article six-years in the making was just published in a journal and I leverage extracts of this below (1). Although the article is behind a paywall as the author I can provide a copy if requested, just hit me up on InfoMullet on Facebook or @InfoMullet on Xitter. 

In this MegaMullet I’ll begin by parsing through the different kinds of cultural scripts used to convey atrocity: actual incidents, actual propaganda, disinformation, and misinformation. I’ll describe how algorithmic sorting creates different channels within which these four types interact to create narratives that anchor themselves in perception, including a fascinating case study of how neighboring countries of India and Bangladesh have adopted diametrically opposed narratives of a conflict half a world away. How local purpose appropriation of the Hamas v. Israel conflict gets processed through their local views, and then dual-purposed returning through virality to US social media.

I’ll then explain some of the macro dynamics of why atrocity scripts seem to be increasing in volume, including the emergence of para-professional and amateur OSINT communities and how the collapse of Twitter’s safety mechanisms under Musk makes the most recent conflict atrocities seem “more” than past conflicts.

What follows may be a bit academic, but pulls from my research on how a system of radicalization can be exploited by non-state actors using atrocity scripts that can counterintuitively aid their cause. And how the dynamics of provocation and reaction between state and non-state actors result in something called the “accidental guerilla” phenomena that we are watching play out in Israel.

Finally, I’ll end with some thoughts on vicarious trauma, where the nature of social media and the spread of atrocity scripts can subject people to intentional or inadvertent trauma that combines with a feeling of helplessness creating stressors that, although not comparable to those living through the events, can be harmful nonetheless. As my work requires a lot of research efforts into traumatic materials, I’ll share some best practices for those who want to engage in para-professional or amateur OSINT communities, as well as how to share this information to avoid vicarious traumatization.

As this is the second MegaMullet in under a week I again reserve the right to edit for clarity, grammar, and minor revisions ongoing. I’ll add a Discussion section to the bottom where the FAQ I receive will be generalized and answered.

Parsing Script Types

In the image above I’ve included four types of cultural script imagery and narrative related to the airborne aspect of the Hamas invasion of Israel. I selected the airborne aspect because it provides a useful framework for understanding the different kinds of threading which applies also to atrocities, without having to spread actually violent imagery. Consider the below as a proxy for how different types of atrocity scripts can co-exist and spread as well.

Actual Incident Scripts


Actual incident footage comes from the incident itself; either from victims and survivors,  unedited perpetrator footage, or unedited surveillance or police footage.   The key distinction here is it is released either as it happens or unedited soon thereafter.

Because this footage is taken, by definition, in the midst of an incident it’s often the least engaging from a content standpoint. It tends to capture only a piece of the action and rarely has the best vantage. That’s because it was taken in the moment. Camera views are based on where they were placed or where someone was standing, and that’s often not staged in advance for cinematic viewing. Without editing, one often can’t gain a complete narrative of the event based on the source. Unless it’s been grabbed by an amateur OSINT aggregator, it shouldn’t be set to music.

In the case of the paraglider incursion, it means that the best images we have are grainy ‘blobs’ in the distance that have to be circled to be spotted. Why? Because that’s how it really was. Unlike in Hollywood movies the paragliders, as far as we know, didn’t dramatically swoop in a low altitude with the Flight of the Valkyries playing from speakers. Unlike parachutes, paragliders are maneuverable, and so Hamas militants were able to maneuver and pick their landing locations. They landed some distance from the festival or kibbutz, linked up with oncoming technical and motorbike forces, and moved as a force into those areas. This is in contrast with the disinformation scripts portrayed below.

Actual Propaganda Scripts

Actual propaganda scripts are still “real” footage, unlike the two categories below. But they have been ‘processed’ to create a narrative. This can be done in advance or after the fact. The images above, for example, are from a slickly produced video by Hamas developed in advance of the invasion and released concurrently. It amounts to a commercial of military capability and competence, highlighting paraglider operational capabilities, concluding with militants storming a mock settlement they had created and all set to a thematically engaging soundtrack.

Atrocity propaganda scripts are those that are filmed during an incident, but then packaged for distribution afterwards, and we’ll deal with these in more detail below.

But the key thing to remember with propaganda scripts is they are being shared for a reason, and that reason is to convey a message, idea, or theme. And it’s that intention that determines what gets included and what gets left out. The paraglider video is designed to convey military competence and operational capability – that Hamas like the world’s special forces is high-speed low drag capable of conducting combined air-land assault operations. They’re not going to include the time the paraglider crashed on landing or when the mechanic came out and had to work on the engine to get the thing to lift off the ground. Likewise, when Hamas, Russia, or Ukraine share drone footage, even from an actual battle, they will probably not share the videos where the drone gets neutralized or the bomb misses.

This isn’t to say propaganda videos can’t be useful. There are tons of little details to pick up from propaganda scripts, including the message they intend to convey. The paraglider training video has a different message, audience, and intended effect than the atrocity scripts released by Hamas as described below. Understanding why Hamas would shoot this video, and what they intend to convey with it, is important to understanding Hamas.

Viral Disinformation

The third type of script, and perhaps the most useless,  is unfortunately often the most viral and widely spread. This is flat-out disinformation, characterized as completely false and inaccurate presentations. The narrative emerging in the first hours of Hamas’s invasion was of an “airborne” component with some natural confusion arising from whether it was parachutists, hang gliders, or paragliders.

It’s this initial confusion around the incident where disinformation thrives. In the image above are multiple stills from two videos. Both capitalize on the theme of “airborne” assault, theatrical presentations including dramatic captioning and soundtracks, and the anchoring of the viewer’s perception that they are looking at “actual footage” and “breaking news.” This can blind the viewer to clear details of the footage that mark it as false. In one clip, for example, the viewer is led to believe they are watching the actual Hamas militant descend by parachutes into the Re’mi Music Festival. And that would be dramatic if it were true.

However the actual clip takes in a heavily urbanized area with skyscrapers in the background, the crowd rather than being a rave-set is mixed of all ages, many seem to be wearing matching shirts with sports logos and kids play soccer in the foreground. This is because the clip is of a demonstration parachuting event held at a major soccer game in Egypt. (2) (3) I’ve seen plenty of other disinformation clips shared, that I’m not sharing here because they are violent. But they are recycling past conflict imagery, either from past conflicts in the region or appropriating combat footage from ISIS and Taliban playing on a Westerner’s general inability to distinguish any geography or people east of Greece or west of  China.

To clear the field, I’m going to answer from the premise that we’re excluding non-Hamas materials being recycled from past conflicts (e.g. I’ve seen ISIS and Taliban videos recycled.) There’s certainly more than enough material from the Saturday invasion and we’ll just begin with the premise that’s what we’re talking about.

Disinformation like this is the hallmark of ideological advocacy, whether intentional from the start or appropriated.

Sanitized Glorified Misinformation

The final type of script is misinformation, and the example I use here comes from the Bangladeshi feeds of InfoMullet viewers. I use this for two reasons. First misinformation is different than disinformation. Whereas disinformation is real footage falsely used, misinformation is false imagery used to convey a specific narrative. Second, it demonstrates the importance of algorithmic sorting in viral spreads (described further below) to create channels of awareness that differ based on where one is.

In the misinformation stills, which were clearly generated by an AI imaging program, Hamas militants parachute in formation down into the al-Aqsa Mosque to be received by cheering throngs of Palestinians.  These images are clearly a form of magical realism, no one’s going to mistake them as “actual.”  But even within the fantasy of the depiction, there are key things wrong.

First, al-Aqsa Mosque is about 130km from the location of the Re’im musical festival where the actual air assault took place. Not only is the location off, but the configuration of the al-Aqsa Mosque compound is inaccurate. Second, there are no paragliders, only parachutists, which is perhaps picking up on the early spread of disinformation mentioned above as input into the AI image generation. But third – what the hell? AI’s normal hallucinations create some truly bizarre scenes, like the Hamas militant in the first still who isn’t even strapped into a parachute. He’s one handing the straps while what appears to be a luggage cart (?) is strapped into the parachute. More importantly, these scripts are sanitized and glorified. The assault is portrayed as a heroic jump to seize the al-Aqsa Mosque received by Palestinians. Gone are the massacre victims of Re’im or any indication that a massacre took place.

It’s clear the narrative of these scripts, given their popularity in viral channels far distant from the disinformation viral scripts spread in the US mentioned above, is targeted at Bangladesh’s Islamic population. Few with any internet access could’ve missed the spread of attention on 7 October – Hamas had struck a major blow against Israel in a dramatic invasion.  But if your feed is filled with scripts like this, it’s easy to understand why Bangladeshi may take a generalized outrage by the West against Hamas with the specific imagery presented here and adopt a narrative of the heroic resistance of Muslim’s third holiest site in almost what amounts to a ‘rescue’ of oppressed Palestinian.

Glorify vs. Horrify:  Bangladesh Misinformation vs India Disinformation in Portraying the Hamas v. Israel War

These channeled narratives of viral misinformation and disinformation often are received through a local context but end up having a dual use. Original material from the Hamas v. Israel war is obtained and reconfigured to serve local narratives. Those local narratives, however, where they overlap with third-party foreign interests, then become viral on their own. One useful case that exemplifies this is the different ways Indian and Bangladeshi social media are portraying the Hamas v. Israel War, including its atrocities. These neighboring countries have a complicated relationship, overviewed previously in this NYE AMA 2019 Forecast Response, summarized below:

Muslim-Hindu relations has always been the tectonic fault-line in India, even before the partition. But the partition by the British of a Muslim majority Pakistan and a Hindu majority India made this fault line visible. And the wars that have been fought since then have led to at times, a jingoistic nationalism and machismo in Hindutva opinions of Muslims, and in some cases vice versa.  There is no love lost between Hindu nationalist and Muslim populations – and if Modi keeps pushing on this fault line he may just get an earthquake.

This fault line also runs through Bangladesh, which lies on the east of India but joined Pakistan during the partition as a Muslim-majority country. It was known as “East Pakistan” then. But the Bengali subsequently fought a War of Independence and left the Pakistan Dominion becoming an independent country in the 1970’s. Given its large westerly neighbor Bangladesh and India have always been integrated in many ways not dissimilar from the US and Mexico. Like US/Mexico relations, there is a large illegal immigrant Bengali population in India for economic reasons.

A country with a majority of Muslims and a history of living under perceived occupation and oppression by a more powerful different religious aligned state; Bangladesh’s antipathy for Israel runs deep. After the first partition, what were then called ‘East Pakistanis’ traveled to Israel to join Arab states in the 1948 Israeli-Arab war and later, fought as members of the Palestinian Liberation Organization (PLO) in Lebanon in the 1980’s (19). Shortly afterward Bangladesh allowed the PLO to open a political office in the capital city of Dhaka (20). In contrast, although Israel recognized Bangladesh’s independence after the liberation war, the new Bangladeshi government took the time out of forming a new state to formally reject this recognition. Even today, Bangladesh does not recognize Israel as a state and prohibits both trade and travel to the country. Those who break this ban face arrest and charges of sedition (5)(6). In contrast Bangladesh consistently supports the cause of an independent Palestinian state based on 1967 borders and hosts numerous bilateral exchanges, scholarships, and trade.

A majority Hindu country that perceives a restive minority Muslim within its borders and a neighboring majority Muslim countries as an existential threat, India tends to side with Israel(7). I’ve summarized the Hinduvata ideology and nationalist tendencies of India’s Prime Minister Modi in suppressing non-Hindu populations both in the article above and in this YouTube covering the 2021 Farmer Protests.

It’s no surprise these countries have opposite takes on the Hamas v. Israel war. Bangladesh takes the “sanitize & glorify” approach of misinformation described above, portraying Hamas as rebellious heroes striking a blow to liberate al-Aqsa that if you squint your eyes might begin to look like an aspirational fantasy of similar actions against India.

In contrast, Indian right-wing social media accounts have taken a “blacken & horrify” approach to disinformation, heightening alleged atrocities as part of a broader Islamaphobic narrative aligned with Modi’s Hinduvata agenda to suppress Muslim, and other non-Hindu populations, in India. Although much of this material is created for local consumption by Indians, over 70% of the engagement comes from within the country, it travels abroad along Islamaphobic and OSINT lines (mentioned below) magnified by the breakdown of Twitter’s safety mechanisms (also below).

The Indian approach to “blacken & horrify” is to amplify not just an atrocity that occurred, such as the killings and kidnappings of young children, but to make it even worse by implication of the most barbaric activities in which the atrocity occurred.

Disinformation includes photographs of a Hamas militant playing with an alleged kidnapped Israeli girl,  implying pedophilia. But the image is from early September and just appears to show an adult playing with a child. The horrific context of the image can only occur when it is framed as an outgrowth of the Hamas attack. In another video, young girls are again highlighted as being captured ‘sex slaves’ of Hamas but in the video itself, the girls appear to be part of a trip and are using their cellphones happily  (7).

One of the most egregious examples of this Indian “blacken & horrify” approach to disinformation, that may have gone viral into the US space, is a video of a 12-year-old boy being beheaded in the back of a pickup truck. The video however has been dated to an atrocity that occurred in 2016 during the Syrian Civil War (8).

Given the allegations of Hamas beheading children in the Kfar Aza attack disinformation such as this is particularly sensitive. Although many earlier reports of beheadings have since been walked back by IDF military, Israeli, and US political authorities, others have maintained it happened. This includes the original reporter, Nicole Zedeck, who has stood by her original reporting (9), and both the Israeli Foreign Minister and later the Defense Secretary Austin who implied some of Hamas’s atrocities were worse than ISIS (10).   Images have been released by Israel of murdered children under other means. So I’m willing to take their word on it for the beheadings and we’ll eventually find out when reports are released. But as far as I could find, no imagery of this specifically horrific act was released by any source. Both out of caution and respect for the dead and the family survivors.

But in today’s amateur OSINT environment (see below) people began scouring the internet for horrible things to find. Either to prove or disprove particularly horrific details. Some may have happened across the Syrian video and taken that as ‘proof’ while some may have just seen the images released by Israel. Some may not have found anything related to this, but they probably found something because there is a lot of uncomfortable content out there.

And that’s the problem, and challenge, where actual incident atrocity reporting gets mixed with disinformation into an amateur OSINT community.  To me – killing children is bad enough, amateurs don’t need to go scouring the internet for confirmation or sourcing when no one’s disputing that children were killed.  And finding Indian disinformation doesn’t ‘disprove’ the atrocity itself, just proves that you can’t trust Indian Hinduvata bloggers with an agenda to push.

This represents the dual-use nature of propaganda, disinformation, and misinformation scripts.  Indian “blacken & horrify” disinformation serves a local purpose for Modi of portraying Muslims as an ‘unfit’ people, a common trope of Islamaphobia. It’s also used locally by conservatives to attack liberals by forcing them to defend the undefendable. But Islamaphobia and conservative v. liberal culture wars are not confined to India.

Through Twitter’s breakdown amplifying viral replication and its own OSINT communities, these Indian-manufactured atrocity disinformation scripts reached US  social media spaces, often from pro-Israeli Republican lawmakers such as Marc Zell.  From there the amateur OSINT community looking for proof found them and often began sharing them outside of a framed context that prepares bystanders for the content they’re about to receive.

The beheadings became a kind of idiomatically structured meme. A form of repetitive phrasing simplified to a single line, that got inserted over and over that Hamas “raped, killed, and beheaded children.” We know those are likely true, but the repetitive idiomatic phrasing anchors the cultural script in the minds of the viewer, regardless of whether or not they see an image.

This kind of memetic spread of scripts has a third often unintended purpose. The Greeks knew that the mind was the worst artist of atrocity, which is why they kept the horrific acts off-stage letting the audience fill in the images. But what works in fantasy, such as Oberon’s drumbeat of accusations against the Mountain in the infamous Game of Thrones duel; can create trauma when taken out of its fantasy context frame. Both for the people saying it, as they condition their brain to habituate around imagery they have created for the purpose of illustrating a text, and in onlookers. I cover this a bit more in the final section about how to avoid experiencing or inflicting, vicarious trauma on someone through sharing atrocity scripts.


Macro Dynamics in Current Volume of Atrocity Scripts

Trend Towards Para-Professional & Amateur OSINT

Open-source intelligence (OSINT) is a traditional method of intelligence collection and analysis that focuses on source material already in the public domain. What used to be a field exclusive to classified intelligence positions has since broadened to commercial analysts and most recently expanded to include large para-professional and amateur communities. Para-professionals are often current or former analysts with experienced backgrounds, and local domain expertise in the subject matter they cover, and are willing to put the time and investment into sourcing. Whether because of professional obligations to engage in social media, or they are picking up alternate careers, they create persistent presences on Twitter, YouTube, and Discord becoming recognized as a credible source of record. @RALee85 is an example of this on Twitter. With prior service experience as a Marine, an Alfa fellowship tour in Russia, and serving as a current fellow at the respected Foreign Policy Research Institute (FPRI) he has both the bonafide and experience to parse imagery and his analysis of events in Ukraine’s battlefields has been a boon. As in actual intelligence analysis, the para-professional community operates by a maxim that could be described as “it’s more important to be accurate” than right or first.

Being ‘right’ or ‘first’, or both seems however to be the motivating force of a large swath of the amateur OSINT community. Although many try hard, often with no background, no experience, and ping-ponging back and forth over global hotspots they prioritize salacious, titillating, engaging, and being first with breaking news. Within this group, I also place the ideological aggregators for whom OSINT is as much advocacy of a previously held and determined position as it is. Starting with an ideological outcome in mind, any event is a backward fit within that position, and content is shared in that narrative. These amateur groups aggregate like-minded thinkers who want a steady serving of confirmation bias for breakfast and pre-chewed righteous indignation along a comforting line of thought second breakfast. With low editorial standards, compromised ethics when it comes to objectivity, and poor records of accuracy these include popularized Facebook pages, ideological YouTube channels, and the three major cable news networks.

The Larger the Event – the More Material to Spread

A second macro dynamic is based on the size of the event itself. In a given public mass killing or terror event of the kind I study, you may have at must only a handful of primary sources for OSINT, often one of which is the perpetrator who increasingly livestreams their events. If the event is over in minutes, only those on the scene who have the wherewithal to film while attempting to survive, and maybe security camera footage, is going to be the sum total of all sources. With the Hamas and PJI invasion of Israel however, there were 21 separate locations of conflict some of which were contested areas for over 24-48 hours.  With an invasion force of over 1,000; there’s more than enough personnel to assign specialized tasks of harvesting footage for propoganda and spread (see below in the Actor section.)

Twitter’s Role in Amplifying Atrocity Scripts

But before getting to the Actor’s role, I think it is important to recognize another major macro factor that is new to this particular unfolding event. And that is that we are watching the first major breaking conflict in a post Musk-takeover of Twitter.

In the past, the sum total of all materials created in an event was not equal to what the public ended up seeing. First, scripts had to run the gamut of safety teams and content moderation technologies on social media platforms which often prohibit violent content. This content might circulate around and on low-moderation platforms like 4Chan/8Chan, Telegram, and Signal; but that means to access that content you had to be able to tolerate everything else associated with those platforms. And even then the audience of these areas pales in comparison to the size of the major content aggregators of TikTok, Facebook, Instagram, YouTube, and Twitter.  That makes content also harder to share. If you want OSINT material on Discord or Telegram, you have to know the specific Discord server or Telegram channel to join.

Some of this content eventually was released by investigations, government reports, or researchers, but well after the Amateur OSINT community had already moved onto the next hotspot. Because fewer saw it in the immediate aftermath, fewer circulated it, and it faded from view. It’s always been out there, but the time delay between the incident and its acquisition was such that it wasn’t in the public view as much as the Amateur OSINT community.

What’s changed is Musk’s takeover of Twitter. Despite its low subscription base, Twitter has always served a key role in ‘breaking news’ and providing immediate information. Part of Musk’s takeover of Twitter was an intentional dismantling of safety moderation technologies and teams under the premise they were stifling free speech.  Unlike the niche providers above, Twitter’s material is globally accessible as soon as it is published. And constant tweaking of the algorithm with a reduction of user content management tools has resulted in more extreme content reaching more people, even if that’s not their wish. This has made the instant sharing of atrocity scripts, be they actual incidents, actual propaganda, misinformation, or disinformation as described above lightning-fast and globally accessible (11).

Speaking from personal experience, there’s never been a lack of disturbing “atrocity” content if you knew where to look or were patient in finding it as it circulated and bubbled up. But I do think with Twitter’s safety mechanisms dismantled and its algorithm pumped to maximum engagement, we’re seeing more content on a platform where it makes it easier for that to circulate, resulting in pieces getting more exposure and picked up for recirculation.

Why Actors Spread Atrocity Scripts in Conflict

Broadcasting, Narrowcasting & Algorithmic Sorting

We’ve cleared a lot: the nature of scripts, the macro trends amplifying – but the question remains, why would any actor even want an atrocity script broadcast? It seems counterintuitive at first. But part of the reason is the feedback dynamics of violent radicalization, and how non-state actors in “ungoverned” spaces can use them through broadcasting and narrowcasting to further radicalization. In this case ‘ungoverned’ is from the perspective of the country targeted, and it means that normal day-to-day law enforcement investigations and judicial proceedings are unavailable, either because the safe haven is in a physical space that requires military intervention to reach (such as Hamas and Gaza) or is digitally manifested and hosted in that gray area of the internet (such as 8Chan.)


The feedback structure and extracted text are taken from the soon-to-be-published peer-reviewed article on the Root Causes of Radicalizations. Although formulated years ago, the academic publication takes a long time, hopefully, the implications for the Hamas v. Israel war are clear.

Broadcasting capacity, [A], is the ability of the non-state actor to mass-disseminate cultural scripts from the ungoverned space into the governed space, satisfying the first criteria of a cultural script contagion. These cultural scripts fuel the green cheese theory of our thought experiment above, facilitating the early development of radicalization. Shared examples of ungoverned space suffering of an in-group at the outgroup’s hands increase perceived grievance and moral outrage while contextualizing the grievance into the narrative. Without concern for state actor surveillance or interference, the non-state actor can provide a continuous stream of new content and facilitate fixation (Hoffman, 2017, p. 5830).

Narrowcasting capacity [B] represents the non-state actor’s ability to target tailored content to small populations, sometimes even individuals, who share similarity and notoriety bias (Hoffman, 2017, p. 5920).  Narrowcasting accelerates fixation, identification, and activation, as shown in [C], by tapping into these biases supporting a cultural-script contagion.

Narrowcasting capacity is increased based on the severity of a conflict zone the ungoverned space overlaps with, depicted in [D]. We define a conflict zone as an area of active violence which could include violent mass protests, irregular, or conventional military conflict. The non-state actor in this zone may participate in the conflict or exploit state weakness. Safe havens and conflict zones can coexist from a non-state actor’s perspective. Despite local military conflict and risk, the hindrance of state actor intervention protects against disruption of non-state actor efforts. The cultural scripts created in a conflict zone carry a higher notoriety bias because they depict real conflict…Through social media and content platforms, narrowcasting can reach a larger audience quickly. Individual ISIS foreign fighters in Syria and Iraq acted as celebrity influencers, collecting audiences in the thousands and providing daily updates. These fighters ensured “immediate accounts of heroic battles and more mundane daily activities,” activating the warrior mentality in their media followers (Hoffman, 2017, p. 5929). Insurrectionists in both Washington DC on January 6th, 2021, and Brasilia, Brazil on January 8th, 2023, narrowcasted live streams of their participation in those events. Activation efforts within these narrowcasts may encourage travel to the ungoverned space or conflict zone; the commission of predatory mass violence in the governed space; or other actions supporting the cause.

In analogy, if cultural scripts are the ‘power’ for radicalization, non-state actors are the power generators operating river dams or coal plants to mass produce cultural scripts. Broadcasting and narrowcasting are the means of distributing and transmitting power from the source of its generation to its consumption. The algorithmic sorting of social media and content platforms adapts transmission pathways in real-time to seek out and find consumers who may have already knowingly or unknowingly indicated preferences for such content.  Last-mile distribution occurs when the scripts enter governed space networks described above and are shared through self-selected echo chambers and bolstered by the illusion of numbers. High-risk populations can then identify their favored green cheese theory generators and connect directly to their content, facilitating future narrowcasting. (1)

The algorithmic sorting mentioned in the article is key to understanding how seemingly opposite narratives can mutually coexist in Bangladesh and India, and then how only parts of those get received back into the US. This is key to understanding why some areas don’t seem as horrified, or jubilant, as others. Using the case study above, the ‘sanitized & glorified’ cultural scripts I’m seeing in Bangladesh are only apparent because I have a global audience and can filter to a region. If all my social media engagement was from the US, I probably wouldn’t have seen it. Likewise, most of the Bangladeshi on my feed are probably not seeing as much of the actual incident footage, or even the Indian “blacken & horrify” disinformation coming out of India.  They aren’t unaware of outrage over Hamas’s action, but they may assume the sanitized narrative of striking a blow against Israel and for liberation is causing that outrage. These can mutually coexist on the internet.

Intervention and the Accidental Guerrilla Syndrome

Of course, cultural scripts don’t exist in isolation from physical actions. When a mass violent event occurs, societal reaction can take many forms, one of which is an intervention into the ungoverned space to destroy the safe haven the non-state actor is operating from.  Again, this was developed years ago but well explains exactly what’s about to happen with Israel’s rumored ground-invasion of Gaza imminent. Again, an extract from Root Causes of Radicalization explains how these dynamics contribute to radicalization:

Another form of connection, [A], occurs when the societal response to terrorism in the governed space results in a state actor intervening in the ungoverned space to deny a safe haven and disrupt operations. Strength in these terms is measured along a continuum of intervention methods beginning with non-violent developmental efforts and scaling up to include retaliatory strikes or sustained full-scale military intervention.

Limiting the effectiveness of any intervention is the number of accidental guerrillas created in response to the intervention, shown in [B]. Any intervention of foreign entities into local spaces creates some natural pushback. However, strong violent interventions into the ungoverned space can cause occupants of that area to violently oppose the actors from the governed space they perceive as foreign (Kilcullen, 2009, p. 38).  Accidental guerrillas emerge not in support of a radical ideology but in support of local interests or opposition to foreign aims relative to the strength of the intervention (Kilcullen, 2009, p. 38).

Non-state actors anticipate this effect by integrating themselves into ungoverned space communities. Poorly conceived foreign interventions into the ungoverned space can cause more harm to the locals than the presence of non-state actors (Kilcullen, 2009, p. 274). Radical non-state actors exploit these interventions to portray themselves as local defenders against external influence (Kilcullen, 2009, p. 38). Turning a safe haven, ungoverned space into a conflict zone [C] also benefits radical actors by improving their notoriety bias and potentially fueling vicarious emotional exploitation by self-similar sympathizers in the governed space. Reducing the strength of intervention through “slower, less violent, more locally based, or lower in profile” methods lessens the impact on local populations, reduces the risk of accidental guerillas, and weakens the non-state actor’s ability to exploit the intervention (Kilcullen, 2009, pp. 37–39).(1)

The mechanism of the accidental guerilla dynamic is a key focus here because it’s what’s about to happen in Gaza. As originally depicted by Kilcullen, Accidental Guerilla is a cycle of four stages, just replace “AQ” with “Hamas” and “conflicted area” with “Gaza” and see if it sounds familiar(12):

Kilcullen calls this the accidental guerilla hypothesis, an extremist group embeds itself within a given ethnographic population exploiting their grievance, and conducts brazen attacks on their enemy, which draws retaliation onto the population they have embedded within. The hope is to spark overreaction, to get the enemy to abandon proportionality. When retaliation lacks proportion or causes suffering as collateral damage on the population, the extremists hide within, which generates further grievance against the enemy, strengthening by relative comparison the militants’ hand.

This goes a long way to answering the “why” of “why” Hamas launched in the invasion.  It can end up being, in the long term, a successful asymmetric way of degrading a far superior conventional foe. Consider the classic case of accidental guerilla, which was the al-Qaeda attack on the United States in 9/11. That pulled the United States into a 20+ year war with not just AQ, and not just the Taliban, but a big chunk of the Afghanistan population who became accidental guerillas against a foreign enemy. And in the end, the US lost.

There hasn’t been an election in Gaza since 2006. Hamas held no referendum or plebiscite on these attacks to assess popular support. But if you’re a Palestinian in Gaza, whatever sympathy and horror you may have had at the Saturday attack was tempered by the immediate need to survive the retaliatory bombardment.

Given that Hamas chose the terms of this engagement, they had plenty of time to prepare Gaza for what was sure to be a guaranteed Israeli reaction.  I’m not sure Hamas intended to achieve all that it did, as that was as much a massive failure on Israel’s part as military acumen on Hamas (which I’m covering in another InfoMullet which I’ll link here when complete.) But they knew at least if they even partially succeeded, Israel was coming for Gaza again.

Hamas’s Agency in Conducting and Spreading Atrocity Scripts

Let’s pause for a breath here. I’ve spent a lot of time parsing the kinds of cultural scripts, discussing macro factors, and going over non-state actor dynamics in broadcasting and narrow-casting atrocity scripts. But don’t mistake that for overlooking the agency of Hamas in all of this. Hamas is a violently radicalized Islamist militant group with a current worldview where, in the context of the principles of warfare mentioned previously, they do not believe in distinction and proportionality when it comes to civilian casualties. They are seeking to liberate not only Palestine from Israel but form it into an Islamic state. In pursuit of that goal, indiscriminate killings of civilians, even with excessive force, in their worldview, is simply not morally wrong but a legitimate act of resistance.

On 7 October Hamas committed atrocities in the intentional targeting, kidnapping, and killing of non-combatant civilians, including the elderly and the young. Regardless of any discussion on the “blacken & horrify” scripts, there’s no dispute that the massacre of children did occur, and that’s bad enough on its own.

It didn’t necessarily start out that way however, and understanding the evolution of Hamas’s violence is important to understand how it got to a place where murdering children was seen as a net positive thing to do. This is not a validation or defense of Hamas’s actions, but a dive into understanding the pathway of how cycles of violence can harden the approaches of violence, shifting further and further away from the principles of warfare of distinction, humanity, proportionality matter, and especially “honor” which should not be mistaken as a quaint outdated notion but the simple belief that those who take upon themselves an authority to act with violence are bound by a corresponding obligation to temper it.

Originally Hamas focused its violence on Palestinian collaborators with Israel and what in the article I refer to in the previous MegaMullet as legitimate military targets of occupation  (e.g. military units and checkpoints). The specifics can be debated, but this followed a pattern of distinction and proportionality. But the Hebron massacre, also known as the Cave of Patriarchs Massacre, referenced in the past MegaMullet changed that. In 1993, an Israeli illegal settler,  Baruch Goldstein opened fire on 800 Muslims in prayer killing dozens and wounding twice that amount(13).

After that attack, Hamas collectivized the Israeli population to the acts of Goldstein and saw “Israel”  as not following the laws of warfare and indiscriminately targeting civilians. This justified a retaliatory stance of their own indiscriminate and excessively violent, methods of killing civilians as ‘leveling the field’ with how they perceived Israel as acting.  Whether this belief of Goldstein was sincere, or a cynical exploitation to become more violent is beside the point. They acted on it.

In the sincere or cynical debate we can through reports that, at times, Hamas has offered a moratorium on civilian attacks to Israeli officials but has been rejected. Whether these were sincere or cynical overtures, Hamas can justify its continuing violent radicalization of its members with “hey we tried” without really putting much effort into the hard work of changing their method and altering their behavior.

If anything, their behavior escalated with the 7 OCT Invasion. Analysts note that in the past, Hamas attempted to prioritize targeting adult males under the premise they served as a ready conscription force for Israel, but avoided killing women and children unless it was collateral damage. When they conducted kidnappings, it was of Israeli soldiers for the most part, and as far as I know, did not kidnap women (14).

Finally, it’s possible that an intentional desire to create atrocities and publicize them was magnified by the extent of Israel’s operational failures. Soldiers who achieve a “breakout” against an entrenched enemy they’ve spent months, years, or even decades fighting against are at high risk of losing discipline and turning to the worst acts of savagery. Israel’s military responses, unfortunately, took hours to get to some of the impacted areas due to the redeployment of crucial IDF battalions away from Gaza to the West Bank in the months leading up to the invasion.


Without having to worry about their own survival or maintaining military order, some militants have turned to the kinds of acts that are all too common in military history when soldiers run amok within a civilian population they are viewed as enemies, or worse, not even human. I don’t want to try and wash this entirely away as a loss of discipline. I think there are some strategic calculations for Hamas and we should not let this slide as a moment of “loose discipline.” But it’s not a lie that this is the most significant attack Hamas, or any non-state or emerging-state actor, has ever conducted against Israel and the combatants involved knew it. With the failure to respond taking up to hours and so no obvious pressure, that creates really bad conditions on the ground where we regularly see time and time again the worst qualities of humanity come out.

ISIS Rewrote the Book on Atrocity Scripts

Another practical reason for Hamas to both engage in atrocity and then publicize it has to do with ISIS. Prior to the rise of ISIS, the so-called Islamic Caliphate in 2013-2014, the general consensus was that atrocity didn’t sell. That the pushback against horrific imagery of the worst behaviors of humankind in combat would serve as a deterrent to sharing them broadly. However, ISIS proved that Web 3.0 was a big place with a long tail and many dark alleyways. The atrocity scripts ISIS released were slickly produced, set to music, and abhorrently violent. Spread along social media pathways via algorithmic sorting in both broadcasted and narrowcasted forms, they wandered through the general population to reach the most extreme niches of their target audience.

Even if the overwhelming majority of the planet’s population, of all faiths, ideologies, and regional dynamics were repulsed by what they saw, enough individuals found self-similarity and notoriety in the atrocity scripts and ISIS’s overall messaging; that up to 30,000 foreign fighters would travel to Syria and Iraq to join ISIS. Those who couldn’t travel sent funds. And those who couldn’t travel and still wanted to fight – ended up fighting at home. This isn’t unique to ISIS. The first four chapters of my dissertation study this phenomenon in the “terror contagion hypothesis” which describes violent radicalization spread through public mass killing terrorism as a form of social contagion. Indeed this concept of social contagion spread through extreme acts is the basis of the terror contagion hypothesis for public mass killings which was the motivation for Root Causes of Radicalization (1).

And no one needs reminding that in the state of the world today hatred of Israel from a policy standpoint or more generalized anti-Semitism, among extremist populations this resonates.

That’s what makes these cycles extremely difficult to counter. When you have entrenched grievances and cycles of violence, the worst of the worst tend to rise in those systems as they can appeal to that cycle of grievance and violence. Collective punishment of the whole for acts of the individual becomes not only contemplated but accepted as a regular course of action under a theory of “they did it first.”

Indeed, at least one retaliatory murder has occurred in the United States as of this writing,  where a six-year-old American-Palestinian boy was stabbed 26 times and his mother injured in an attack authorities are describing as retaliation to the Hamas atrocities (15) and an attempted mass-shooting in Brussels that killed two and injured one was conducted by a self-professed supporter of ISIS (18).

Hamas Isn’t Alone in Celebrating Atrocity

Again, the characteristics we want to pin on Hamas as being unique to them as a violent movement, or even to Islamic violent non-state actors, are more common than we may like to admit. I mentioned Ben Givr in the article having a portrait of the Hebron massacre perpetrator in his living room. Likewise, the Kahan Commission found Ariel Sharon responsible for failing to take steps to prevent the Sabra and Shatila massacres during the Beirut War. Although directly perpetrated by Christian militias during the Beirut Civil War, Sharon commanded the IDF surrounding the camps which could have prevented or ended the massacres which were on a scale, of fatality and barbarity, as the 7 OCT Invasion (16).

But Sharon still ended up having a productive political career and served as Prime Minister. To many Israelis, he was a hero. And as much as we may find abhorrent the propaganda around the worst of the worst acts, it can resonate with some.

This is why I am such a stickler when it comes to the bright line of violence and take an autistic adherence to laws of war. When we start allowing subjective interpretation or failing to distinguish individuals or even groups from the larger populations they inhabit, we get into this vicious cycle that is extremely difficult to break out of and in many ways becomes self-fulfilling.

Just as the Hebron massacre may be used by Hamas as “justification” for indiscriminate killings of civilians, the Remi or kibbutz massacres may in the future be likewise used by some current or future Israeli terror group to commit acts of atrocity.


Actual Atrocities are not True Crime Podcasts

Vicarious Trauma

The problem with humans is we are well-evolved for an environment we no longer live in. Our brains are hard-wired to perceive threats and react in an effective way because it may mean life or death in the next few months. Affective threat responses involve getting big, making lots of noise, and gathering allies. The body dumps a neurochemical mix activating fight, flight, freeze, or fawn responses. These neurochemicals are not optimized for nuanced performance. Major muscle movements get amplified at the expense of fine motor control. Aggression transferrence happens when the friend or foe distinction gets clouded by adrenaline and we can attack bystanders, even friends, families, and allies we called to us in reaction to the threat. All of this is incidental to surviving the next few moments of a directly perceived threat. The dump of neurochemicals is so strong and so toxic that we can go into shock in the hours following an incident and into exhaustion a few days later. The body traded future mental and physical energy to amp every system’s dial up to 11 at the moment. If we do survive, our minds imprint dense memories of enhanced senses. Where a normal day might in analogy be a page of text, a traumatic incident we survive might be chapters in a book, or the whole book itself. This is why people often describe traumatic events as if they occurred in slow motion, or being able to “review” the incident with tremendous detail after the fact. If you were a caveman and a bear came into your cave, those ‘dense memories’ are the brain’s way of reminding you for the rest of your life what it took to survive.

But we don’t live in caves today. Our mind has a hard time distinguishing affective threats perceived online half a world away from the bear walking in our cave. That dense memory can become PTSD and if experienced repetively CPTSD.

Watch any aftermath of a shocking topic of violence played out on social media and count the cues of affective threat response: people getting ‘bigger’ and ‘louder’, rallying allies and demanding allegiance, transferring aggression to bystanders who in normal circumstances might consider friends and family. I see it all the time in my work on public mass killings. One reason I usually invoke the 72-hour Razor Rules pause. Not just to allow more accurate information to come out, but to bypass the immediate affective response period.  You can almost set a clock to incidents like a public mass killing between the start of the affective response and 3-5 days later when people complain of ‘exhaustion’ and ‘fatigue’ and just want to ‘give up’ in the face of it all. That’s the down cycle of the neurochemical drump they experienced at first.

Because we can still experience trauma indirectly. Known as vicarious trauma, it’s the experience of an affective threat response triggered not through directly lived experiences, but through intermediaries that feed our senses information our rational brain knows is not immediate or local, but that ancestral evolutionary response can’t distinguish.

Maybe in another 10,000 years evolution will catch up and adapt us to the technology we’ve created, but for now, we have to take into account that when we go looking for atrocity content, or expose others to that content, an affective threat response may be triggered.

And in areas that benefit from stability privilege, we generally don’t experience life or death affective threat violence and the aftermath that can create.  In the United States, the population hasn’t experienced a direct war on its own lands since 1865. And the ratio of civilians who have served in the military to travel abroad and experience warfare has been declining since then, just as the amount of military who themselves deploy experiencing combat also declines.

Part of the theory of false memory is that when people haven’t directly experienced a thing, they tend to insert proxies from the closest thing they do know. So when it comes to thinking we know about atrocities, overwhelmingly in the modern age that’s going to be some form of entertainment, Hollywood or otherwise. But that YA Dystopian Fiction does not actually prepare people for dystopian societies any more than that True Crime podcast or snuff-gore horror movie prepares someone for amateur OSINT work interacting with atrocities.

Personally, if you don’t need to see these for a professional reason, I don’t think you should go out looking for them. But I know telling the Internet not to do a thing isn’t a useful strategy. Also, because the Internet is a place where people are going to do what they want, you may not go out seeking atrocity scripts and yet still run across them in your feeds.

For those seeking understanding, to bear witness, engage in amateur OSINT processes, or just want to know what to do when someone shares this material on your feed understanding harm reduction with atrocity scripts is vital. There are two reasons for this. First if in the course of viewing these materials, an affective threat response gets triggered, from a pragmatic standpoint any fine-motor control of analysis goes out the window, and the risk of being taken by misinformation and disinformation increases as does the risk of subjecting oneself to vicarious trauma. Second, harm reduction processes seek to reduce provoking that inadvertent affective response and vicarious trauma in unsuspecting onlookers who may not have consented to participate in it.

For instance, I made the title of this article specific, so people not interested in atrocity can skip past it. I’m being fairly clinical in my descriptions and using images of paragliders and parachutists rather than anything more serious. At 10,000+ words it’s not going to spread virally as clickbait anywhere. This isn’t by accident. When I give presentations in mixed audiences on public mass killings, even without violent imagery, I give content warnings that we’re going to use pictures of the perpetrators, gear, and manifestos (because that’s part of the evidence in my research) and I know they may not be prepared for as a different, more specialized audience might.

Strategies for Working with Atrocity Scripts

Harm reduction doesn’t begin when you share content, but before you first encounter it. The Dart Center for Journalism and Trauma has an amazing guide which, due to its length, I’m going to extract edited summaries below, but I really recommend reading the full document and practicing its guidance (17).

Before you even begin, stage your work around these four strategies:

  • Give your best working hours to the worst material. Don’t do this work on the edge of time you have available as a side hustle. Treat it like a job that’s important. 
  • Avoid working with traumatic imagery late at night before sleep. This can reduce intrusive thoughts or sleeplessness which contributes to dangerous fatigue.  
  • Work as a team. Whether as part of a professional, para-professional, or amateur group divide and conquer. Understand each member’s strengths and weaknesses, and boundaries and work within that. 
  • Keep work and personal life separate. Plan for time to transition between work activities and personal or social activities. 

In addition to what you might expect, you’ll see a theme repeated here and in the processing section below how important mental alertness and avoiding doing this work when fatigued. Think of your mind working through these traumatic images as engaging in strenuous physical activity, if your body is already exhausted and muscles strained you’re more likely to injure yourself.

Processing Atrocity Cultural Scripts

When you begin actually working with atrocity or traumatic materials, DART provides best practices to follow below (17). Pay special heed to #5 in how you decide to share any results with coworkers, fellow researchers, or even on your social media feed.

  1. Hit Pause: Don’t click on the link if you haven’t formed a process of the steps below. If you get surprised by the content, hit pause and think through things before continuing.
  2. Disrupt & Decide: Take an action that disrupts your normal viewing habits; change position, take a deep breath, and ask yourself these questions:
    • Do I need to see this now?
    • Do I need to see all of it or just sections?
    • Do I need to see it at all?
  3. Set the Machine: Alter how you see the material, for example:
    • Changing the position of the window and making it smaller.
    • Lowering the sound, or turning it off altogether.
    • Reducing the color vibrancy by adjusting the saturation control
  4. Be Active and Analytic If you have to review material multiple times, think through what you actually need each time. Rather than watching the whole thing, be specific. 
    • Taking notes to minimize the need to go back and forth repeatedly over distressing footage you have already seen.
    • Scrubbing through, or dragging your cursor through the timeline, in order to locate aversive sections. If you don’t need to look at them in detail, then don’t.
    • Riding the audio controls toggling the the volume up and down as if tuning an old-fashioned radio to the correct station.
    • Blocking out the most distressing sections of an image so that you can work with the rest more effectively.
    • Specifics for film editors: Avoid using loop play or working with dynamic trim rollers when cutting around images of death or body parts, and consider setting clip thumbnails to images that are less intense.
    • Specifics for translators: Even if you need to listen to audio in detail for translation purposes, experiment with ways of building in some protective distance.
    • Adding a standard, unambiguous strap-line in email subject headings and log sheets.
    • Give a rough description of the material in the accompanying notes.
    • Jotting down time codes for the most adverse sections of video footage.
  6. Take breaks whenever your concentration lapses. We are more vulnerable to emotional overload when we begin to feel fatigued. Neuroscientists also suspect that when we are tired and not fully utilizing the more analytical functions of our brains, we may be more likely to “record” traumatic images in a way that leads them to come back as nightmares or intrusive images.
    • Take regular screen breaks and shift your viewing position.
    • Paying heed to basic physical needs.
    • Making time for conscious breathing.
    • Identifying how you are feeling.
    • Using a recognized grounding technique if you start to feel “spaced out”, unusually “floaty” or that you are being sucked into the content.
    • Using humor as a distancing mechanism.
  7. Putting a lid on work. Adopt a process that signifies a “break” from work on traumatic materials.
    • Distraction files. Pictures of cute puppies or landscapes, etc. – to be a useful way of transitioning out of verification work.
    • A more deliberate transition ritual. Walking around the block, washing your hands or face, or physically tidying the space that you are working in.
    • A practice that acknowledges the reality of what happened in the material you have just engaged with.

Frequently Asked Questions on Atrocity Scripts

There’s a lot of confusion and guilt when working with these kinds of materials, and again DART has provided some guidance to frequently asked questions which I’ve summarized below. Note how much they describe “work”, “manager”…amplifying my earlier point that if you don’t have a professional need to do this, I’d suggest not. (17).

I feel bad about not scrutinizing every detail. Surely it is my duty to bear witness?

Becoming overwhelmed from exposure to traumatic imagery can impair your ability to work with it effectively. And that doesn’t benefit- anyone, least of all the victims. Moreover, perpetrators often circulate images of atrocity with the deliberate attention of terrorizing their opponents and pushing the news agenda in directions that suit their aims. Building in some strategic distance when viewing images is a way of pushing back against that. Self-care is a professional duty as well as an act of personal resistance.

I would feel like a coward if I didn’t look…

To look or not to look is always a personal decision, but it is probably best to avoid looking at anything that is not strictly necessary for work, especially if you are working with high volumes of material.

Is there anything wrong with wanting to look at traumatic material?

No. Some trauma researchers believe that as a species we have a natural desire to seek out knowledge of phenomena that threaten our existence as part of an attempt to develop mastery over them. People slowing down to get a better view of car crashes are often accused of voyeurism, but the desire to look is just as much an instinctive survival reaction as is the desire to turn away. Our brains need to approach danger in order to understand whether we should fight or flee.

What if I start “seeing things” that I have been working on?

Intrusive recollections – re-seeing traumatic images one has been working with – are not unusual. Our brains are designed to form vivid pictures of disturbing things, so you may experience images popping back into consciousness at unexpected moments. That again is a function of our built-in survival apparatus.

Intrusions are usually nothing to worry about of unless they are particularly persistent and are associated with a longer and more general sense of becoming unwell or functioning poorly. If you are troubled by intrusive thoughts, try not to push the images away. Forcing yourself not to think about something can paradoxically help those thoughts to boomerang back. Instead, try and use your imagination to transform the images into something else. Apply the equivalent of a matte in your mind, or even blur the images, paint over them, or scroll forward to a moment in the video where the violence is no longer happening.

I am a manager. What should I be thinking of?

As a manager, you can’t control the tide of toxic content coming into newsrooms, but you play a crucial role in mitigating impacts on staff. People are more resilient when they feel that their supervisors have their best interests at heart. You should:

  • Understand how trauma-related issues can affect the newsroom, and where your duty of care lies
  • Review newsroom workflows and file management procedures to make sure that they minimize unnecessary exposure to traumatic imagery
  • Prepare new staff for the nature of the work they will be asked to do [8]
  • Foster a supportive and cohesive working environment
  • Allow staff to take adequate breaks, and, if necessary, to rotate out of shifts that require high-intensity work with user-generated content to others that are less demanding
  • Understand how the physical environment of a workspace – access to light, presence of plants, and various design features – contributes to resilience
  • Respect people’s needs to develop their own workflows and strategies for handling traumatic content
  • Know how to identify and support journalists whose well-being is being affected by trauma exposure
  • Ensure that your organisation has a comprehensive trauma management and training policy

What are the warning signs for when someone is not doing so well?

Distress per se is not a sign of any kind of underlying emotional injury. Stories that involve human cruelty are likely to be upsetting. If you are working with such material, distress reactions – including anger, despair, bad dreams, periods of numbness, feeling agitated or wired, and difficulty concentrating – are far from unusual. Such periodic bouts of emotional “bad weather” can be disruptive and annoying – and do require active self-care – but they are not signs in themselves that one needs to seek external help.

Do be attentive, though, to any reactions that get stuck and become generalized to other situations. There is a significant difference between feeling numb the first time you watch a traumatic video and losing enthusiasm for activities outside of work or affection for people who matter to you. In terms of warning signs, be particularly alert to:

  • Marked changes in character
  • Unusual irritability, or explosive anger that fires up without apparent reason
  • Images or thoughts related to a project intruding at unwanted times, which are unusually persistent and don’t diminish over time. Particularly if they involve situations in which you imagine yourself being followed or attacked.
  • Unusual isolation or withdrawal
  • The sense that life has become meaningless or foreshortened
  • A persistent and general feeling of being numb or deadened inside
  • Increase in self-medication (alcohol, drugs, compulsive overworking, etc.)

If you have any concerns, please consult the Dart Centre’s website for more information.

What else should I do to look after myself?

Everybody whose work involves trauma needs a self-care plan. Don’t forget the importance of maintaining a balance between work and other aspects of life. Exercise and finding time for friends and family are important ways of restoring balance. Take time to reflect: the material you are working with could provoke political and moral questions, and challenge certain beliefs. You may find it helpful to talk these through with friends who have similar interests and values. Keeping a journal can also be a good way of both processing one’s reactions and reconnecting with what matters to you. If a story starts to feel all-consuming, as if nothing else matters anymore, that is an indication that you should seek better balance.


I’ll add frequently asked questions and answers arising from this work here over the next week or two so check back!



(1) T. Clancy, B. Addison, O. Pavlov, and K. Saeed, “Root Causes of Violent Radicalization: The Terror Contagion Hypothesis,” System Dynamics Review., p. 47, October 2023.











(12) D. Kilcullen, The Accidental Guerrilla: Fighting Small Wars in the Midst of a Big One, 1 edition. Oxford University Press, 2009.