Using Journalism for Self-Protection: Profession-Specific and Journalistic Measures and Strategies for Countering Violence and Impunity in Mexico and Honduras

ABSTRACT In Mexico and Honduras, journalists face violence from state and non-state actors and almost complete impunity. Given a lack of effective state protection, some resort to alternative means of (self-)protection and justice-seeking. Via analysis of 67 interviews with journalists and protection actors, this article shows how many reporters use journalistic and profession-specific forms of self-protection and analyses their benefits and challenges. It adopts a novel analytical approach building on insights from diverse areas of research on people’s responses to violence and insecurity. Scholarship on civilian responses to armed conflict offers a relevant framework for understanding self-protection measures. Although measures such as “avoidance” and “accommodation” of violent actors are a means of survival and short-term physical protection, they bring significant new risks. Above all, they can undermine trust from the public and within the profession and the usefulness of journalism for society. Drawing on literature on the protection of human rights defenders, the analytical framework is extended to show how some journalists also engage in broader self-protection strategies around transforming their work. These strategies combine protection and professionalisation and aim to boost journalists’ internal and external support and credibility.


Introduction
In Mexico and Honduras, journalists face violence from state and non-state actors and almost complete impunity for attacks. Violence is defined here as physical or psychological attacks, threats or harassment (possibly) related to journalists' work and impunity as an absence of legal consequences for the perpetrators. Violence does not affect all journalists equally but mainly targets independent and investigative reporters covering sensitive topics such as human rights violations, corruption and organised crime groups (OCGs) CONADEH 2017b;Waisbord 2002). Such journalists may also be considered human rights defenders (HRDs), or individuals who promote human rights. Violence and impunity, combined with a lack of effective state protection and the limitations of support from employers, civil society and the international community, leave some journalists feeling that they are responsible for their own security.
This article explores how such journalists turn to themselves to find alternative means of protecting themselves and seeking justice-or at least the truth. The research questions it seeks to answer are: How do journalists exposed to violence and impunity employ self-protection measures and strategies? What effect(s) do they perceive these to have? Self-protection is defined as practices on which individuals/groups rely to defend themselves from violence and the effects of impunity.
Drawing on data from in-depth interviews with journalists in Mexico and Honduras, I describe how they incorporate profession-specific and journalistic self-protection measures and strategies into their work as a complement or substitute to state protection, and assess their varying impact. I argue that while self-protection measures are often problematic, self-protection strategies-more widespread in Mexico-represent a crucial form of resistance and source of hope for journalists and are likely their best option for contributing to change. The article's main contribution is to show the need for more scholarly focus on self-protection as central to how journalists survive, defend themselves and their profession and continue and improve their work in contexts of violence and/or repression.

Literature Review
To date, there is no clearly defined body of literature or integrated theoretical framework on journalists' responses to violence and impunity. Hence, I review diverse areas of research on people's responses' to armed conflict and societal violence and the protection and coping tactics of journalists and HRDs at risk. While acknowledging significant differences between situations of armed conflict and generally violent and repressive states, scholarship on civilians' responses to the former is relevant for several reasons. First, although wars are now rare in Latin America, it has been argued (albeit controversially) that Mexico is involved in a non-international armed conflict, and that other countries in the region face levels of armed gang violence close to that of armed conflict (Bellal 2018). Second, from the viewpoint of civilians/citizens on the ground, at least, there are parallels between the two in terms of pervasive violence, sometimes from multiple, state and nonstate, sources, and insecurity; lack of state protection; and, frequently, absence of international protection. Third, as I will show, journalists, HRDs and citizens in violent/repressive regimes employ self-protection measures that are comparable to and fall into similar categories as those used by civilians in armed conflict.
Similarly, the self-protection measures and strategies of journalists and HRDs fell into similar categories whether they came from Colombia, Mexico or China, three countries that feature prominently in the narrow strand of literature reviewed on their protection/coping tactics. However, I acknowledge that the findings from the first two countries are likely more directly transferrable to the present study's cases, since all are Latin American partial democracies with sociocultural similarities, than those from a highly authoritarian Asian state.

Self-Protection Measures
Recent scholarship on civilian responses to armed conflict suggests that people largely depend on themselves and each other for survival and protection since the state fails to protect or even targets them and the response from the international community is frequently delayed or non-existent (Suarez 2017;Jose and Medie 2015;Baines and Paddon 2012;Bellamy and Williams 2009;Mégret 2009;Bonwick 2006). Civilians commonly employ a range of self-protection measures which vary between countries and contexts but tend to represent similar tactics (Jose and Medie 2015, 528).
Scholars have proposed various typologies of self-protection measures (e.g., Jose and Medie 2015;Baines and Paddon 2012;Bellamy and Williams 2009). Common to most are the following three non-violent ways of dealing with violent actors (both state and nonstate) 1 : avoidance (or non-engagement), remaining neutral (or active non-engagement) and accommodation (or non-violent engagement). The literature on journalists' and HRDs' responses suggests that they have evolved comparable measures, as I will show. Avoidance is used to refer to civilians eschewing interaction with armed actors, for example by keeping quiet, hiding, using warning calls, changing routines or fleeing (Jose and Medie 2015;Baines and Paddon 2012;Bonwick 2006). Similarly, in Colombia and Mexico, citizens respond to insecurity by avoiding public spaces or public transport (Berents and Ten Have 2018). In Mexico, HRDs employ practices including risk-assessment, changing routines or "avoiding certain places and people" (Nah et al. 2017, 8-9). Many HRDs depend on improvised tactics as well as formal NGO security training, for example not reporting attacks for fear of reprisals from the authorities (Amir 2013;Nah et al. 2017). As a last resort, some HRDs relocate overseas via NGO or university programmes or by claiming asylum (Gready 2019;Jones 2015).
The "avoidance" measures of journalists tend to be bound up with the practice of journalism as well as survival. The main such measure is self-censorship, defined as avoiding investigating or publishing on certain issues due to fear of physical or other reprisals (Fadnes, Krøvel, and Larsen 2020). Self-censorship is widespread in Mexico (Harrison and Pukallus 2018; and China (Repnikova 2014;Tong 2007). In China, journalists sometimes wait months or years for the right moment to investigate or report on sensitive issues to minimise pressures (Repnikova 2014).
Remaining neutral involve civilians overtly adopting a neutral stance in relation to different armed actors in a conflict, for example "community peace initiatives" in Colombia that allow civilians to work together to peacefully resist the conflict (e.g., Burnyeat 2013). For journalists, preserving their neutrality is an important part of their protection which has increasingly come under threat in recent decades; Sambrook (2016a) identifies the "loss of neutral status" as a "key driver[…] of the increase in violence towards journalists" worldwide (loc. 415-416).
Accommodation involves civilians interacting with armed actors, often under duress, for example by paying for protection, acting as informants, using connections with government or armed groups to obtain "privileges", "survival sex", etc. (Jose and Medie 2015;Baines and Paddon 2012;Bonwick 2006). Some HRDs in Mexico engage with the state strategically, requesting state protection merely to ensure that attacks are officially documented (Nah et al. 2017). For journalists, again, "accommodation" tends to be bound up with their work. In China, investigative journalists are careful to quote officials and other sources acceptable to the authorities for security reasons (Repnikova 2014;Tong 2007). In northern Mexico, journalists often report only the official version of crime or corruption (Relly and González de Bustamante 2013). Journalists in Mexico are commonly co-opted by both the government and OCGs (IACHR-UN 2018, para. 6; Relly and González de Bustamante 2013, 119-120).

Self-Protection Strategies
Journalists and HRDs have also evolved complementary self-protection strategies to build internal and external support and professionalise. These strategies go beyond survival and physical safety to address other aspects of their longer-term security and ability to continue and develop their work.

Building Internal Support
In Colombia and northern Mexico, journalists have employed mutual protection practices such as travelling together, collaborative reporting and publishing sensitive stories in different media outlets simultaneously to reduce the chances of reprisals González de Bustamante and Relly 2016). In China, journalists use "cross-regional media supervision" to circumvent pre-publication checks by local authorities (Repnikova 2014, 121;Tong 2011, 55-57). Such collective practices have parallels in some war reporting; for example, see Sambrook (2016b, loc. 3496) on the "Sarajevo pool".
In China, Colombia and Mexico, journalists have formed informal networks and formal associations, putting aside competition and the desire for scoops to work together for security, solidarity and journalistic proposes (IACHR-UN 2018;González de Bustamante and Relly 2016;Repnikova 2014). Similarly, HRDs commonly rely on national or local peer networks for protection, solidarity and emotional support (Bennett et al. 2015;Amir 2013).

Building External Support
Some journalists and HRDs also try to extend their support/solidarity networks to include local civil society organisations (CSOs), international allies and the public. The literature on HRDs emphasises the importance of connections to multiple, multi-level stakeholders, from CSOs to states and multilateral bodies, to harness support and deter perpetrators (e.g., Bennett et al. 2015). While this theme is less clearly developed for journalists, reporters in China and Mexico have similarly found they need support from local civil society, including lawyers and NGOs, international allies and the public, and to use social media, to strengthen their security and work González de Bustamante 2013, 2017;Repnikova 2014).
International support is key for journalists, particularly from foreign journalists and media outlets. Chinese journalists have forged networks with journalists internationally, including via global consortia (Repnikova 2014). National-level collaborative journalism projects in Latin America like the League Against Silence cooperative in Colombia increasingly work with foreign journalists and rely on international donors (Janetsky 2019).
The need for public support is another important emerging theme. Cottle, Sambrook, and Mosdell (2016) argue that civil society has a duty to protect journalists given the essential role of independent journalism, but that this is undermined by inadequate social awareness and understanding of this role. Journalists in Mexico and China use social media to circumvent censorship and gain public trust (González de Bustamante and Relly 2014; Repnikova 2014). Gready (2019) highlights the need for HRDs to find ways of countering negative public perceptions, due in part to smears by the state and mass media, and secure popular support. Alternative and social media and the internet are vital when HRDs seek to publicise their activities and threats against them for self-protection (Nah et al. 2017).

Professionalisation
Journalists in Mexico and China have also moved towards professionalism and professionalisation (González de Bustamante and Relly 2016; Repnikova 2014). I define "professionalism" as journalists' ideas about what constitutes "good journalism" including observing standards such as autonomy, accuracy, fairness, confidentiality of sources, ethics, public interest and, in some cases, objectivity. By "professionalisation" I mean the process by which journalists achieve "professionalism". In Latin America, professionalisation often includes a focus on bolstering professional autonomy ) and developing investigative journalism (IJ), frequently curtailed in repressive/violent contexts (Harrison and Pukallus 2018). For example, the Colombian League Against Silence was set up in 2017 to carry out collaborative IJ into stories not being covered due to corruption or fear (Janetsky 2019).
Scholars have also noted growing professionalisation of HRDs' work, including developing ethical codes. O'Flaherty and Ulrich (2016) argue that professionalisation is important to ensure that the work of human rights field officers working for inter-governmental organisations (IGOs) and NGOs is rigorous and accountable. However, the professionalisation of HRDs is contentious, as discussed below.

Advantages, Risks, Gaps
The literatures reviewed focus mainly on "avoidance" measures; they provide fewer insights into whether and how journalists and HRDs use "accommodation" and "remaining neutral". The literatures on journalists and HRDs lack a sustained discussion of the advantages of self-protection measures generally. The disadvantages of self-protection measures have been studied in more detail. Some measures employed by journalists and HRDs can bring risks as well as benefits (for similar observations on civilians, see: Jose and Medie 2015, 529;Bellamy and Williams 2009, 22-23). For example, "avoidance" measures like self-censorship can have a negative effect on journalism and its public image (Hughes and Márquez-Ramírez 2017, 514;González de Bustamante and Relly 2016, 693-693); exile can affect HRDs' ability to continue their work (Jones 2015).
There has been relatively little scholarly discussion of the self-protection strategies of journalists and HRDs or their potential advantages and drawbacks. Relevant here is a debate which questions the desirability and unintended consequences of applying a "professional" label to HRDs. O'Flaherty and Ulrich (2009, 8) acknowledge that professionalisation can potentially "have the effect of disqualifying, and thus excluding, some groups of [HRDs]", particularly those working on a local and/or voluntary basis. Critics maintain that professionalisation can lead to discrimination whereby "professional" HRDs enjoy greater legitimacy and access to support, protection mechanisms and resources than more grassroots HRDs, even though the latter may be in greater need (Malkova 2018;Jones 2015). The literature on HRDs also contains some scattered references to the risks of seeking international support, including funding, resulting in increased repression (Bakke, Mitchell, and Smidt 2019;Bennett et al. 2015). It is not clear whether professionalisation and international support can also be problematic for journalists.

Background
Mexico and Honduras differ in many respects: Mexico is 13 times more populous, four times wealthier and is considered more geopolitically significant, while the countries' political systems are distinct (federal vs unitary states). However, they also have important similarities. Both countries are considered partial democracies. Both are parties to the major UN and regional human rights treaties and many relevant rights are embedded in domestic legislation. Yet both have a history of widespread human rights violations, high levels of societal violence, increasing militarisation, weak rule of law, growing presence of OCGs which have infiltrated politics and state institutions, rampant corruption, and almost complete impunity. Both experienced a critical juncture in the 2000s after which human rights violations, including against journalists, escalated: for Mexico, the government's militarised "war" on drug cartels from 2006; for Honduras, a coup d'état in 2009 (AI 2019; IACHR 2019; Palifka 2020; UN 2019; Berg and Carranza 2018;Frank 2018;IACHR-UN 2018;Keck and Sikkink 1998).
Impunity for such attacks is almost total: in Mexico over 99% remain unpunished (Article 19 2019, 166), while in Honduras the impunity rate is around 90% (IACHR 2019, para. 110). This makes it difficult to prove who is responsible-and easy for governments to blame generalised violence and OCGs (Article 19 2018, 37-38). However, official figures and NGO research suggest that state agents are key, if not the main, perpetrators, responsible for up to 50% of all attacks or more (SEGOB 2019, 10; Article 19 2018, 32, 78; C-Libre 2020, 38-42, 114; C-Libre 2016, 26); collusion between state actors and OCGs is also common (Article 19 2018, 78; IACHR 2019, para. 52). Hence the state is directly responsible for a substantial proportion of the violence as well as for impunity, raising questions about its willingness to implement its international human rights (IHR) commitments.
In both countries, violence and impunity have persisted and increased despite legal reforms and the creation of dedicated state institutions in the last 10-15 years, following international and domestic civil society pressure. At the time of the interviews, the main such institutions in Mexico were the federal-level Special Prosecutor for Crimes Against Freedom of Expression (FEADLE) and the federal Protection Mechanism for Journalists and Human Rights Defenders, both based in Mexico City; there were also local-level prosecutors and protection mechanisms in Mexico City and Veracruz state, the two locales where interviews were carried out. In Honduras, the key institutions were the national Public Prosecutor's Office

Methodology
The findings are primarily based on 67 in-depth semi-structured interviews with journalists (44) and protection actors (23)-state and non-state actors that support and protect journalists at risk-conducted in Honduras and Mexico in March and September 2018 respectively (31 interviews in Honduras and 36 in Mexico). Given the importance of studying violence against journalists and impunity at subnational level (e.g., Hughes 2017, 167), I carried out interviews in two locales per country: in Honduras, the capital Tegucigalpa and San Pedro Sula/El Progreso; in Mexico, Mexico City and Veracruz state. 2 These were chosen because of their persistently high rates of violence and impunity (Article 19 2018, 74-75, 83; CONADEH 2017a) as well as their accessibility and relative security for a sole researcher.
The primary interviewees were journalists who reported on human rights or other sensitive/controversial issues such as corruption and OCGs and/or were involved in activism around violence against journalists and impunity. Such journalists have been identified as a sub-group of HRDs most at risk globally (UN 2011). The secondary interviewees were protection actors, including representatives of government protection schemes, CSOs, NGOs international NGOs (INGOs) and IGOs. Protection actors provided a critical view of journalists' self-protection measures and sometimes directly supported self-protection strategies such as collectives/networks and professionalisation; however, these interviews are largely used as background information in this paper. Relevant academic and grey literature and media reports were used for triangulation but also as a data source in themselves.
I selected interviewees on the basis on my knowledge and contacts from my previous employment for an INGO 3 and via snowball/chain-referral sampling. In practice there was overlap between the two groups of interviewees since a significant minority of journalists interviewed in both countries had worked for (I)NGOs as well as media outlets. Moreover, the journalists' profile varied between countries: in Mexico most journalists worked for mainstream outlets (14/24), often on a freelance basis, or a combination of mainstream and alternative outlets (5/24), whereas in Honduras most journalists (15/21) were part of the alternative media, with one-third of these (5/15) working for outlets established/ run by NGOs. This is likely due to selection bias inherent to snowballing as well as a reflection of the countries' differing media sectors and their openness to coverage of human rights/sensitive issues. The sample of journalists was not intended to be representative of the general media landscape in the two countries: it targeted a specific subset of journalists who are HRDs.
Prior ethics approval was obtained. Given the vulnerability of the journalist-interviewees and the "high risk" nature of Mexico and Honduras, I followed a strict ethical and security protocol, including using secure/encrypted communications and safe meeting places, and ensuring anonymity where desired. 4 Most interviews were conducted in Spanish, lasted 60-90 min and were audio-recorded. Questions centred around the alternative (non-state-centric and/or non-legal) norms and strategies that journalists used to counter violence and impunity, their usefulness, obstacles encountered and whether journalists saw themselves as HRDs. I found most interviewees friendly and relatively open, likely due in part to my INGO background/experience and Spanish language skills. Many journalists seemed to embrace the opportunity to discuss their experiences and some seemed to find it cathartic, particularly in Honduras, albeit a few found it somewhat upsetting at times. I transcribed the interviews and conducted a thematic analysis, triangulating the data with primary and secondary written sources.

Findings
In this section, I discuss the use and effects of the self-protection measures and strategies used by journalists interviewed in Mexico and Honduras, employing the categories identified in the literature review.

Self-Protection Measures
"Avoidance": Self-Censorship In both countries, a third of journalists interviewed said they had employed self-censorship as a temporary or semi-permanent measure. In Mexico, unlike Honduras, it was a practice mainly mentioned by journalists from outside the capital (despite Hughes and Márquez-Ramírez 2017), likely indicating higher risk levels in other regions. Seventy-eight per cent of the journalists in Veracruz state (7/9) and two reporters in Honduras emphasised that they avoided investigating/covering OCGs, especially alleged links with state authorities, even if they had hard evidence because they believed it would not lead to arrests and was unacceptably risky.
Self-censorship is usually seen as negative, an acceptance of fear, and the journalists who employed it did so because they saw no alternative. As noted, it can have a negative effect on the quality of their journalism and its usefulness for society. However, a few journalists also saw self-censorship as a form of positive risk management. In Honduras, 14% (3/21) said self-censorship was the only effective form of protection, along with leaving the country. Mexican photojournalist Félix Márquez said he had avoided covering topics such as homicides since his return to Veracruz after fleeing overseas after the 2015 murder of his colleague Rubén Espinosa: Lots of people criticize me […], especially foreign journalists [or those from Mexico City]: "You're self-censoring" […]. [I say] "Yes, because I can't [… just] arrive for two days, report and go elsewhere. I live here, this is the way I stay alive at the end of the day and yes, it's self-censorship, definitely, but it's also a method of protection.
"Accommodation" As mentioned, "accommodation" involves interacting with (potentially) violent actors, including official and/or OCGs, often under duress, for example using connections to obtain privileges or strategic engagement with institutions. In Mexico and Honduras, some journalists engaged with the authorities without expectation of investigation or protection but merely to document or deter attacks (see also Nah et al. 2017, 8) while others used their professional contacts to circumnavigate official protection channels. For example, two Mexican journalists said they had persuaded high-level officials to investigate threats against them unofficially to assess their credibility; one had also secured bodyguards from the federal police via their then employer rather than directly from the federal Protection Mechanism, which they did not trust. While such tactics are creative responses that give vulnerable journalists some sense of control, they seem likely to benefit well connected and more experienced journalists and do nothing to press for the proper functioning of official protection mechanisms.
The form of "accommodation" most frequently mentioned in both countries was the perceived co-optation of journalists via threats and/or bribery. Fifty-eight per cent of journalists in Mexico (14/24) and 62% of journalists in Honduras (13/21), as well as some protection actors in both countries, believed corruption was widespread in the profession; some journalists believed that a sizeable proportion of colleagues (Mexico) or even the majority (Honduras) were implicated in some form of unethical or criminal practice. Not surprisingly, none of the journalists interviewed admitted to personal implication, although several mentioned having been offered bribes or "gifts", particularly in Mexico. While corruption in journalism, as in other areas, is challenging to document and prove, its existence in both countries is relatively well established by academic and grey literature, backing up these journalists' perceptions. However, it must also be acknowledged that in such contexts it is possible that some of the journalists interviewed who criticised corruption among their colleagues/profession had themselves been involved in practices that others might see as corrupt or unethical.
The most common corrupt practice alleged in both countries, though particularly emphasised in Honduras, was taking official or unofficial payments from political or economic actors, usually government departments or officials, via government advertising contracts or payroll, sinecures and other perks, etc. In both countries it was common knowledge that some journalists were also bribed and/or coerced into working for drug cartels. Journalists from Veracruz and other Mexican states such as Guerrero, Sinaloa and Tamaulipas in particular spoke of this practice (though again not in relation to themselves). According to their accounts, one of the cartels' favoured methods was to get journalists work for them as "enlaces" (liaisons) within media outlets, dictating to colleagues which stories to publicise or suppress (see also Relly and González de Bustamante 2013). According to Félix Márquez and freelance journalist Israel Hernández, although effectively criminals, "enlaces" had become perversely indispensable to other journalists' security in certain areas of Veracruz state, enabling them to self-censor where necessary to avoid violent reprisals. However, it is important to note that in Veracruz, at least, journalists believed that colleagues were co-opted by the state government far more often than by cartels.
Such perceived co-optation greatly undermines journalism's ability to remain independent and serve the public interest. It also impairs journalists' capacity to protect themselves by fomenting distrust within the profession, corroding public trust in journalism and deterring public support for journalists at risk (discussed below).

Reporting
Perhaps not surprisingly, journalists also often favoured publicising and reporting on violence against journalists and impunity, an important and potentially powerful response (see Cottle, Sambrook, and Mosdell 2016). This diverges from the framework of self-protection measures outlined above and, as an attempt to secure public support, could be seen as a form of self-protection strategy. However, I argue that the limited nature and impact of most such coverage means that it is better understood as a professionspecific form of self-protection measure.
In both Mexico and Honduras, several journalists said they had publicised threats and attacks against themselves via the media or social media. Additionally, a large proportion had reported on incidents involving colleagues and related official investigations as part of their work, particularly in Honduras (67% or 14/21 journalists interviewed versus 46% or 11/24 in Mexico). Mexican journalists used a wider variety of media to highlight the issue, for example books, documentary films, animations, photography exhibitions and festivals, as well as media reports and public events. In both countries, the main motivation was to raise the public's awareness of the issue. Some journalists also hoped that by documenting and denouncing violence and impunity their coverage would act as a form of protection, deterring further attacks and triggering governmental action (particularly in Honduras); others wanted to express solidarity, outrage and remembrance (the last particularly in Veracruz).
Although such coverage was a common response, a small minority of journalists in both countries (11% or 4/45) thought that its influence over the public or government was negligible. For example, Mexican journalist Luis Cardona said that his newssite Diario 19 covered the issue but people's "capacity to be astonished has already been exceeded; more than anything I do it so there's a record, but I can't really see any impact." Neither had publicising his own kidnapping and torture in Chihuahua state via a short animation led to any progress in the official investigation into his case, he said, despite it attracting a wide audience and critical acclaim.
The limited impact of coverage of violence against journalists and impunity can sometimes be linked to its lack of depth. In Mexico, three reporters specialising in investigations and human rights noted that although certain outlets covered the issue fairly often, they tended to do so in a superficial, formulaic way, reproducing official statements or lists of killed journalists. Few Mexican journalists were really investigating to find out why their colleagues were attacked and why state institutions charged with investigating and prosecuting did not work, they said.
This dearth of IJ on violence against journalists and impunity was in turn attributed to five key factors. First, the time-consuming and resource-intensive nature of IJ, with resources particularly scarce in Honduras. Second, a reluctance among media outlets and some journalists to dedicate too much time, space and limited resources to the issue due to its perceived lack of newsworthiness and some journalists' reluctance to "become the story". Third, decreasing access to official investigation files and interviews with officials or victims' families or lawyers.
Fourth, in both countries reporting on and, especially, investigating violence against journalists and impunity was itself dangerous and could lead to further attacks. In Mexico, Marcela Turati of Proceso said the outlet had to reduce its coverage of the 2012 killing of Veracruz correspondent Regina Martínez after another of its journalists was threatened by local government officials for investigating the crime. Turati herself had to flee the country after being threatened by a Veracruz official for investigating the 2014 abduction and murder of local journalist Gregorio Jiménez de la Cruz. Anabel Hernández (Mexico) stopped investigating the 2017 killing of renowned journalist Javier Valdez for fear of putting his family at risk. In Honduras, Lesly Banegas and other Radio Progreso reporters had to abandon broadcasting an interview with one of the alleged killers of a well-known journalist when the other perpetrators began to threaten them.
Fifth, perceived corruption in the profession acted as a further deterrent to journalists investigating attacks on colleagues in case it revealed that the motive was the victim's unethical or criminal practices rather than their journalism. For example, according to Jennifer Ávila (Honduras), she and her colleagues at Contra Corriente had often discussed covering the issue but had been put off by the fact that in Honduras pervasive impunity meant that the reasons for journalist killings were unknown, while corruption in journalism was a fact. She did not believe that the majority were targeted for their journalism. Similarly, in Veracruz (Mexico), a third of the journalists (3/9) were keen to highlight that, although many colleagues killed in the state were clearly targeted for their journalism, a significant minority in recent years had allegedly worked for or had connections with drug cartels.
Hence, the impact of coverage of violence against journalists and impunity is undermined by its poor quality, particularly lack of investigation, due to several factors including journalists' use of "accommodation". Perceived corruption leads to a lack of trust among journalists themselves-but, crucially, also a lack of public trust and support. In Mexico, 46% of the journalists (11/24) and a third of the protection actors (4/12) noted a lack of public support for journalists at risk; two Honduran journalists made similar observations (on Mexico, see also Relly and González de Bustamante 2017;Harrison and Pukallus 2018). As a quarter of Mexican journalists (6/24-all from Mexico City) put it there was a "distance between journalists and society" which left people indifferent to journalists' plight. In both countries, a significant minority (17% or 4/24 in Mexico and 24% or 5/21 in Honduras) said that journalists were commonly seen as "sell-outs" or in the pay of the highest bidder, particularly the government, rather than independent and unbiased.
Allegations of corrupt practices are controversial for the profession and their advocates because they have the potential to undermine the stance that journalists are attacked because of their professional activities and to reinforce, even if unintentionally, the government narrative that it is because they are "up to no good". They are also taboo within the profession-against the mantra of "perro no come perro" ("dog doesn't eat dog", i.e., you don't attack your colleagues). But for a small but growing minority of journalists in both countries, the solution to corruption in journalism is to acknowledge and confront it and professionalise (discussed below).

Self-Protection Strategies: Transforming Journalism
Self-protection measures might protect individual journalists in the short-term, by avoiding further attacks via silence or complicity with corrupt or criminal actors, or deterring them through publicity. However, such measures are problematic and risky. Some, like co-optation, are journalistically unethical while co-optation and reporting both carry the threat of violent retribution. Self-censorship and co-optation also undermine journalism as a profession and its usefulness to society, by eroding mutual and public trust, as well as solidarity and support for journalists under threat.
Hence, many journalists, particularly those with experience of activism or working with NGOs, had developed broader self-protection strategies which went beyond immediate physical security and superficial reporting to focus on improving their work as well as continuing it more safely. These strategies combined protection and professionalisation, with the aim of building journalists' internal and external support and their legitimacy. Different strategies were employed by varying proportions of journalists, however most were more prevalent and successful in Mexico than in Honduras, as discussed below.
Building Internal Support: Mutual Support and Collaboration Some journalists, especially in Mexico, were increasingly working together for both protection and journalistic purposes. Twenty-one per cent (5/24) of Mexican journalists, from both Mexico City and Veracruz, mentioned that reporters often simultaneously published sensitive stories in more than one outlet, at domestic and/or international level, to spread risk. For example, according to veteran reporter José Reveles, when El Universal refused to publish an investigation by Laura Castellanos into the 2015 Apatzingán massacre by federal police, Proceso and Aristegui Noticias and US broadcaster Univisión published it at the same time; this not only got around her newspaper's censorship but also protected her from being the sole target of reprisals for a highly sensitive exposé (see also Castellanos 2015).
Journalists in Mexico were also working as part of collectives and networks, facilitated by social media, often with the aim of promoting self-protection and professionalisation. Forty-two per cent (10/24) of Mexican journalists interviewed said they were part of such groups, which operate in the capital, at state level and occasionally across states (see also Article 19 2018, 45, 88-90). They often provided training, both professional (e.g., digital journalism and IJ, how to cover sensitive issues like the environment, elections and security) and security-related (physical and digital security and emotional self-care); some carried out risk analyses and documentation of attacks. Some groups also employed mutual protection practices, often in the form of collective monitoring of members covering sensitive stories, including when travelling to unfamiliar or dangerous areas, and more generally staying in touch, sharing/corroborating information and alerting each other of security risks. In the case of the Veracruz Journalists' Network, members organised joint reporting trips when covering sensitive subjects like narcofosas (mass graves used by drug cartels).
Relatedly, in Mexico, journalists were also increasingly focusing on collaborative journalism, defined as "a cooperative arrangement (formal or informal) between two or more news and information organizations, which aims to supplement each organization's resources and maximize the impact of the content produced" (Stonbely 2017, 14). For example, specialist NGOs Mexicans Against Corruption and Impunity (MCCI) and Quinto Elemento had collaborated with domestic and international media outlets on investigations into the offshore holdings of the Mexican élite as part of the Panama Papers leaks (Chavkin 2018). As well as boosting resources and impact, such projects also offer journalists some protection by spreading risk, as with simultaneous joint publication.
Such collective initiatives also had limitations and challenges. Their success depended on journalists overcoming their egos and desire for scoops, which tends to conflict with their training and practices and the commercial nature of many outlets. A lack of trust, including due to fears of infiltration by government or OCG "spies", was also a recurring theme. Simultaneous joint publication is likely to benefit better known and well-connected journalists and to be employed for stories of national or international interest. Some collectives were stronger than others, most operated only at subnational level and placed little emphasis on counter-impunity strategies. José Reveles cautioned that joint complex investigations were still in their infancy: journalists were still learning to let go of their individualistic mentality, the pursuit of exclusives and ownership of information, he said. There was little evidence of collaborative journalism in Honduras, despite awareness of its potential and attempts by several journalists to develop this area; common obstacles to joint working were a limited pool of like-minded journalists or outlets, lack of mutual trust, perceived co-optation by the government and outlets' financial precarity.
Building Public Support via Professionalisation Some journalists, particularly in Mexico, believed that their best option for addressing violence and impunity-as well as corruption in the profession-was to focus on doing better journalism (similarly, see González de Bustamante and Relly 2016, 694-696), employing various interrelated, overlapping sub-strategies to professionalise. These included: seeking journalistic autonomy; practising socially engaged journalism; raising standards; and doing IJ.
Professionalisation in such contexts of violence and impunity can be considered a selfprotection strategy in two ways. First, it constituted a clear attempt by the journalists concerned to reject self-censorship, "accommodation"/ corruption and the security and reputational risks of such self-protection measures, and (re-)claim "neutrality" vis-à-vis the government and OCGs.
Secondly, professionalisation was seen as a means of generating credibility and public support and pressure to address violence and impunity, albeit in the longer term. The overarching aim was to better connect with the public by producing journalism that was higher quality and more relevant and useful. People would then be in a better position to make informed decisions, including whether to pressure the government for accountability and change, to become part of civil society pressure to end human rights violations and impunity-including in relation to journalists. The hope was that if people could experience journalism, and therefore journalists, as important to their lives and society, they could potentially begin to understand the dangers reporters faced in providing access to information and the need to support and protect them.
However, these sub-strategies for professionalisation were being applied with varying degrees of success, and generally considerably more effectively in Mexico than in Honduras, for several reasons. Mexico has a stronger history of independent and IJ and a less divided and politically polarised media and civil society. Journalists there managed to work more collaboratively on various levels. Mexico City provides a relative safe haven, with its generally lower levels of serious risk and, above all, more access to international donors, media, NGOs and IHR actors (discussed below), which has allowed the city to act as an incubator for independent and professional journalism. This suggests that professionalisation may not be equally viable as a self-protection strategy for journalists in all cases and is contingent on levels of risk and access to resources and support, particularly at international level, as well as internal cohesion.
In both countries, then, most of the journalists had sought greater journalistic autonomy. Autonomy matters in terms of countering violence and impunity because it allows journalists to escape the mainstream media's censorship of sensitive subjects, and work conditions that leave them unprotected, vulnerable to bribes and unable to organise. Twenty-nine per cent of journalists interviewed in Honduras (6/21) and 21% in Mexico (5/24) had set up their own media outlets, mainly online. In Mexico, 42% (10/ 24) also worked freelance, including for the mainstream media, while recent years have seen the establishment of independent outlets by journalists' collectives and two organisations specialising in collaborative journalistic investigations (MCCI, and Quinto Elemento). Such variants did not seem to exist in Honduras, limiting the possibilities for journalistic independence.
Working more independently gave journalists greater freedom to set the news agenda and pursue socially engaged journalism. Two-thirds of the Honduran journalists (14/21) and 29% of the Mexicans (7/24) said they saw their profession as having a social purpose, as being a way to help people and contribute to, develop or change society, rather than just a way to make a living, by covering issues relating to democracy, corruption, human rights violations and impunity, including against journalists. Across both countries, 13% (6/45) reporters aspired to do journalism that was "closer to society" or "side[s] with the people" (Mexico) or with "victims" (Honduras), to prioritise the interests and viewpoints of ordinary people rather than elites. As noted, however, it is journalists who pursue such reporting that are most at risk of attack.
Journalists in both countries were keen to raise standards via professional training and support, to improve work opportunities and conditions but also to counter corrupt practices and distrust in the profession, key obstacles to public support for journalists and their plight. Again such training seemed more available in Mexico, where interviewees spoke of NGOs, journalists' groups and some state bodies offering training in, for example, journalistic ethics, investigative and digital journalism, and the right to reply and access information (see also "collectives and networks" above). In Honduras, journalists identified professional training as a need, with the potential to unite colleagues and build mutual trust, but a third (7/21) spoke of a lack of such opportunities.
In both countries, journalists were particularly keen to develop skills in investigative journalism. Journalistic investigation can be used as a tool to probe problems like violence and impunity, why they exist and who they benefit, even more so in countries where the truth about what is happening is obscured by the failings or corruption of the justice system and the rule of law (Bromley 2017). While journalistic probes cannot replace official investigations they can expose crimes and, sometimes, contribute to pressure leading to prosecutions. A third of Mexican interviewees (8/24 journalists; 4/12 protection actors) believed journalists had played a key role in exposing rights violations, organised crime, corruption, collusion and impunity in Mexico in recent years (see also . Without journalistic investigations into cases such as the Apatzingán massacre (see above) and Ayotzinapa (the enforced disappearance of 43 student teachers in 2014, with the alleged involvement of federal state agents), the truth about certain key events and the Mexican state would quite simply not be known. Such stories were increasingly produced by specialist organisations like MCCI and Quinto Elemento and independent digital outlets. Again, since IJ can be inherently risky, the integration of security, collaborative and mutual support practices was essential (see above).
In Honduras, however, despite 24% of journalists (5/21) expressing strong interest in IJ, interviewees agreed that there was little tradition of the practice compared to Mexico and other Central American countries and a lack of resources. The little being done tended to be carried out by large mainstream outlets and therefore restricted by the owners' interests.

Building International Support
In both countries, international funding was an essential part of these strategies to transform journalism. In Mexico, independent journalistic projects, journalists' organisations, collectives/networks and (I)NGOs all relied heavily on US and European donors. Without this support, many would likely not exist, or certainly not at their current scale.
In both Mexico and Honduras, several interviewees stated that the international media had the potential to amplify concerns around certain stories or issues, including violence against journalists and impunity, and exert pressure on their government, via a "boomerang"-type effect (Keck and Sikkink 1998). These interviewees were mainly based in the capital cities, where foreign outlets and correspondents are concentrated. For example, three Tegucigalpa-based journalists who had worked for foreign outlets believed that international coverage could expose human rights violations and lead to increased scrutiny of the Honduran government in a way that local media could not, precisely because of government influence. However, this idea of a "boomerang" effect was particularly marked in Mexico City, where two journalists and three protection actors spoke of the impact of the global media as almost equivalent to that of IHR actors like the UN in terms of criticism of the country's human rights record and undermining its democratic appearance. While the Mexican government felt it could safely ignore domestic coverage of rights violations or corruption, a report by the New York Times could lead to international scrutiny, meaning it had to take the issue seriously; it also led to more interview requests and media attention at home, further increasing the pressure on the government.
Moreover, a recent trend of transnational journalistic investigations into unsolved journalist murders in Mexico suggests that some reporters there increasingly view collaborations with international media partners as a means of probing impunity that is more impactful-and safer-than doing so alone. In 2019, following a year-long investigation, a group of anonymous Mexican journalists published a series of stories on the serious irregularities in the official investigation into the 2017 murder of Miroslava Breach, along with Forbidden Stories, Bellingcat and the Latin American Centre for Journalistic Investigation (CLIP). The reports, widely re-published abroad, indicated that the Mexican authorities were failing to follow up possible links between the crime and Breach's investigations into alleged links between politicians and OCGs which had caused her to receive threats (Colectivo 23 de Marzo 2019). This was followed in 2020 by the publication of another transnational collaborative investigation into the 2012 murder of Veracruz journalist Regina Martínez (Lakhani, Priest, and Dupont de Dinechin 2020).
Such transnational collaborative investigations into violence against journalists and impunity offer a promising departure within the growing global use of transnational collaborative journalism. Global media "boomerangs" would appear to be a potentially important area for further research. However, such international strategies, although undeniably important, had their limitations. Notably, they were not equally available to all: as seen, international media outlets were more present and accessible to journalists in the capital cities, and particularly Mexico City.
More generally, international support for professionalisation depended on journalists' access to international actors and resources, which was affected by their location, profile and perceived existing level of professionalism. Support from international donors and media, (I)NGOs, etc. tended to favour journalists in Mexico over Honduras, in Mexico City over Veracruz state, those who were better known and connected nationally and internationally and those seen as more independent and "professional". Hence, the professionalisation of journalists may in fact be more divisive and potentially discriminatory than might initially appear, especially for local or self-taught journalists, in a way that echoes concerns highlighted by some scholars of the protection of HRDs.

Conclusion
In this article, I offer a novel comparison of journalists' responses to similar problems of violence and impunity and a lack of protection in two very different partial democracies in Latin America and the challenges they face. I propose an original, interdisciplinary analytical framework for understanding how they respond via their work and assess which responses are tactically most useful. Drawing on scholarship on civilian self-protection, I show that in the case of Mexican and Honduran journalists their self-protection measures included self-censorship and "accommodation" of (potentially) violent actors but also reporting on violence against journalists and impunity. While such measures were a means of survival, security and, sometimes, solidarity in the short-term, they brought new risks and highlighted persistent challenges, notably lack of mutual and public trust linked to perceptions of corruption in the profession.
Then, with reference to research on the protection and coping strategies of journalists and HRDs, I demonstrate how some journalists, particularly in Mexico, were attempting to address these challenges by developing broader self-protection strategies with the overall aim of transforming the profession and practice of journalism. These strategies went beyond immediate physical security and superficial reporting, combining protection and professionalisation to improve their work as well as continue it more safely. They aimed to build journalists' internal and external support and credibility, addressing "accommodation" and (re-)claiming "neutrality" from the government and OCGs.
Self-protection strategies also had limitations. They were more prevalent and impactful in Mexico than in Honduras. Even in Mexico, mutual support and collaboration initiatives were still nascent, and doing better, more independent and socially engaged journalism was still risky. The success of such strategies was dependent on access to funding and other resources, often from abroad, which were not equally accessible to all, with Mexico City at an advantage compared to Mexican states or Honduras. Self-protection strategies can make important contributions to protection and justice but cannot substitute effective state protection and compliance with IHR standards, rather are an additional, complementary way of securing these.
Despite their limitations, I believe that self-protection strategies are vital in the fight against violence against journalists and impunity for several reasons. Firstly, in an otherwise fairly bleak panorama where impunity appears intractable, such strategies are evidence that journalists are not giving in to despair. They use their work and profession to regain agency and hope and to sidestep activism, which many find uncomfortable, ineffective and risky. Secondly, the collaboration required to develop and implement self-protection strategies helps journalists develop the mutual trust, solidarity and organisation they need to overcome violence and impunity and other problems they face, such as poor labour conditions and precarity. Thirdly, by doing better journalism and forming public opinion around the root causes of human rights violations, impunity and corruption, journalists can potentially inspire citizens to join civil society pressure for political and social change; in the process, they may attract societal support for their own plight and demands. Finally, the focus on transforming journalism is, hopefully, helping build the foundations for a truly free press, considered a key part of a fully functioning democracy. This may be journalists' best hope for justice and sustainable protection until broader change occurs in their governments and state institutions, as well as their best option for contributing to that change.
Overall, the article's key contribution is to show the need for more scholarly focus on self-protection as central to how journalists in contexts of violence and/or repression survive, defend themselves and their profession and continue and improve their work. It shows that journalists, particularly in Mexico, valued the norms of professional journalism in their self-protection strategies. More broadly, this indicates the importance of alternative, non-legal norms as a source of inspiration and hope for non-NGO and nonlawyer civil society actors in the fight for protection and justice. The article also suggests the potential of free and independent journalism to promote public trust in and support for human rights values and journalists at risk.