De-Radicalisation and Humanitarianism in Indonesia

: Since 9/11, de-radicalisation programs have become central to every country that deems itself at risk from terrorist attacks from global terror groups such as ISIS. Consequently, many countries have implemented programs to “inoculate their Muslim populations” and de-radicalise and disengage those Muslims deemed radicalised through securitisation and “moderate Islam”. Such programs aim to persuade individuals to renounce extremist ideas and violence and adopt moderate Islam, often state-orientated Islam, as is the case in Indonesia. The Indonesian government and civil society organisations have attempted to address radicalisation by setting up counter-radicalisation and de-radicalisation programs, with various degrees of success. The central thesis of this paper is that de-radicalisation is not achievable, and the current programs in Indonesia are not effective. The paper will ﬁrst critically discuss the concepts of radicalisation and de-radicalisation/disengagement. Then, the paper will critically assess the Indonesian de-radicalisation programs by focusing on their shortcomings and unintended consequences, which result in the labelling and stigmatisation of former detainees and their families and hinder their successful de-radicalisation and reintegration. In the ﬁnal section, the author will suggest that the Indonesian government and civil society organisations need to introduce humanitarian activities to improve their programs and reduce the chances of recidivism.


Introduction
Terrorism has become a significant policy concern for some Southeast Asian countries, especially those with a Muslim majority and minority populations, such as Singapore, Thailand, Malaysia and Indonesia. The concern over terrorism reached a new high after the 9/11 attacks and the Bali bombings in 2002. The US viewed Southeast Asia as the second front in the war against terrorism (Abuza 2009;Abuza 2007). Since that proclamation in 2002, the terror threat in Southeast Asia has increased exponentially and become multidimensional. Numerous terrorist attacks have taken place in the Philippines, Thailand and Indonesia over the past decade and a half (Agastia et al. 2020;Agnew 2010). Apart from Southeast Asia experiencing political violence, terrorism and ethnic violence, over 900 foreign fighters from the region had travelled to Syria and Iraq to either relocate or engage in its ongoing civil war (The Soufan Group 2015). This number does not account for the women and children who had travelled to Syria and Iraq seeking marriage or a new life, in the self-proclaimed Islamic State.
Although terrorism scholars such as Hegghammer and Nesser (2015) have argued that the risk posed by returnees is low, this conclusion does not reflect the resources, political stability and porous borders of many non-Western countries. Returnees to Western countries may pose a low risk, but this is due to them having stable governments and resources to carry out surveillance and secure borders effectively. As such, Western countries have increased possibility to identify, track and arrest the returnees, which decreases the chances of terrorist attacks. Unlike Western countries, Indonesia lacks the resources to carry out effective surveillance, has local terrorist groups such as Jemaah Islamiyah (JI) that have transnational links and ethnic tensions, which in the past have led to a lot of violence. The combination of such conditions means that some of the returnees who may have acquired ISIS ideology do not only have a safe haven but also may have come back to Indonesia to carry out attacks or become agents of radicalisation (Aragon 2001;Aslam et al. 2016;Singh 2016). An example of such a risk is exemplified by the 2018 suicide bombings of three churches in Surabaya by a family of six who had returned from Syria, highlighting the danger that returnees can pose to their own countries (Schulze 2018). This terror attack also adds another dimension to the already multi-dimensional threat of terrorism, as it was the first time that suicide bombings involved an entire family (BBC 2018). The multi-dimensional threat of terrorism is reflected by the recent conflict in Marawi City in Mindanao, where local and foreign fighters from the Southeast Asia region who had fought in Syria and Iraq seized control over various parts of the city (BBC 2017). As the family suicide bombing and Marawi conflict attests, the majority of Muslim Southeast Asian countries cannot afford to take the threat posed by returnees lightly. Not doing so risks terrorist attacks and the reignition of ethnic conflicts, especially in Indonesia.
As a consequence of local, regional and global terrorist threats, Southeast Asian governments have introduced several counterterrorism measures to curb the threat. However, one of the critical areas of focus has been establishing de-radicalisation programs, which are deemed as solutions, because these programs assumed to be able to surgically de-program radicalised individuals. The programs were first introduced to address JI radicalisation, but now cater to returnees from Syria, including women and families. These initial programs primarily aimed to remove the radical ideology from an individual deemed radicalised (Borum and Fein 2017). However, unlike de-radicalisation programs in Western countries that are mostly orientated towards ideological transformation and securitisation, programs in Muslim majority countries like Indonesia, Malaysia, Pakistan and Saudi Arabia tend to have a holistic approach. They include vocational training and entrepreneurship schemes, as a bid to de-radicalise and reintegrate extremists into society. In Indonesia's case, the Indonesian government, the security services and civil society organisations (CSOs) have realised that stand-alone de-radicalisation programs that focus on ideological transformation are not effective and need to be complemented with vocational training and entrepreneurship schemes to increase the chance of success.
Although terrorism is a concern for many Southeast Asian countries, this paper will only focus on Indonesia, because it is the most populous Muslim majority country in the world and has experienced terrorism. First, it will briefly discuss the terrorism industry and the coloniality of knowledge. Second, it will discuss the conceptual issues surrounding the terms "radicalisation" and "de-radicalisation"/"disengagement", because they not only influence how radicalisation is understood, but what de-radicalisation programs entail. Third, it will conduct a critical review of the effectiveness of government and CSOs de-radicalisation programs in Indonesia. Finally, it will suggest that for Indonesian de-radicalisation programs to be more effective, they need to include humanitarianism as part of de-radicalisation programs. The author recognises that radicalisation is complex and cannot be reduced to one factor or pathway. Instead, there are multiple reasons for radicalisation, and no one-size-fits-all de-radicalisation program will be successful. Instead, the author advocates a more tailored approach to de-radicalisation that involves humanitarianism and centred on Zakat or charity work, which is not only a core tenant of the Islamic faith but will also have an immediate impact on the lives of Muslims in Indonesia and outside the country. Such an approach is more likely to convince detainees to disengage from terrorist groups because it has a high personal impact and addresses some of the core reasons why some Muslims become radicalised, which is due to the suffering of other Muslims.

The Terrorism Industry
This part of the paper briefly discusses the terrorism industry as a product of the West, which raises several issues connected to the coloniality of knowledge (Ilyas 2021;Herman and O'Sullivan 1989;Stampnitzky 2014). This is important because the theories, concepts and models used by Western and non-Western terrorism scholars, governments, CSOs and media are a product of this industry. The terrorism industry has been around since the 1970s and was set up by Western countries to counter the threat posed by the Soviet Union and prevent local conflicts that were taking place in non-Western countries from becoming global and threatening their geopolitical interests. Since the 1970s, and as a consequence of terrorist attacks such as 9/11, the literature on terrorism produced by think tanks, research centres and published in specialist journals has rapidly increased in Western countries. The literature has published a lot of research that details new theories, concepts and models that have been developed by terrorism scholars with a Western heritage based in Western institutions to detect and prevent radicalisation and carry out de-radicalisation (Borum 2011;Butcher and Luxen 2019;Copeland and Marsden 2020) That said, the use of concepts like radical, extremist and radicalisation models in non-Western countries raises several issues connected to the coloniality of knowledge (Ilyas 2021), but a comprehensive discussion on this is outside the remit of this paper, but a few comments will be instructive for the reader. There are several problems with Eurocentrism, which include European history and social sciences being privileged, and the social sciences being a force in constituting and comprehending the modern world. As such, their deficiencies are found when they are applied in the non-West because they are not adequate to understand the non-Western world (Seth 2014, p. 312).
By this, Seth means that when categories of social sciences are employed outside the West, they reproduce the West to understand the non-West and may not be suitable. In the context of the terrorism industry, his argument is important, because it suggests that when terrorism theories, concepts and models are used in non-Western countries, they reproduce the Western terrorism imaginary, which may not reflect the non-Western reality (Ilyas 2021). Examples of this are discussed later.

Problems with Radicalisation
This section of the paper will discuss some of the conceptual problems connected to radicalisation. Concepts like radicalisation and de-radicalisation are popular among the public, governments, think tanks and the media. They are complex, but their popularity masks their conceptual problems and impacts the understanding and solutions to radicalisation. Since the late 1970s, radicalisation has been connected to Muslims (Lynch 2013) under the framing of "new terrorism", which deems religion as the cause of radicalisation. However, over the past few years, radicalisation has also been used to describe why and how people with Western heritage and living in Western countries join far-right groups, including terrorist ones (BBC 2019; Butcher and Luxen 2019). This shift in the application does not mean that radicalisation is understood as being the same for both groups. In Muslims' case, radicalisation is deemed as being caused by "extremist" Islamic ideology, especially in light of the new terrorism thesis (Gofas 2012;Mazlan et al. 2017;Tackling Extremism 2013). The logical consequence of this understanding is that the state security apparatus, geopolitical security apparatus, international NGOs, and strategies emanating from such an understanding, focus on disseminating a "moderate", often state-orientated understanding of Islam to prevent radicalisation and carry out de-radicalisation (Rabasa et al. 2007;Mazlan et al. 2017). This understanding is then superimposed on other possible causes for why an individual identifying as a Muslim may become radicalised or engage in violence such as local socio-political issues and international conflicts that involve fellow Muslims being victims of violence. The superimposing is problematic because it overemphasises Islam as an ideological cause at the expense of non-ideological causes for radicalisation. However, in the case of far-right radicalisation, the opposite is suggested; Christian ideology seems to be de-emphasised, and non-ideological, often personal circumstances are emphasised as being the causes of radicalisation (Copeland and Marsden 2020). This situation is rather surprising given that far-right extremists often see their actions and use of Christian symbols as defending Christian Europe and Christians more generally. This is evidenced by Crusader symbolism and vocabulary interwoven with far-right ideologies used by European Christian foreign fighters fighting in the Syrian conflict and on the various Facebook pages of far-right groups (Koch 2017). The same type of material was also found in Brenton Tarrant's manifesto, who killed 51 Muslims in 2019 at a Mosque in Christ Church, New Zealand. Similar material was found in two documents prepared by a Singaporean Christian of Indian ethnicity, who wanted to attack two Mosques in the country, as a call for "war against Muslims" (Zhang 2021). The latter case is very disturbing because it demonstrates that Western far-right ideologies have mutated and camouflaged the hidden racism, with Islamophobia to gain global appeal, even appealing to ethnicities that would ordinarily be victims of such ideologies.
However, the aim of the theories, concepts and models, as Neumann (2008, p. 4) argues, is to identify what goes on before the bomb goes off. As a concept and process, radicalisation is not straightforward or linear, and violence is not always the end result, even though governments and the media often portray it as such. For example, Borum (2011) and Schmid (2013) have identified several definitions employed by scholars and governments. However, the two main ways in which radicalisation is defined are ideological and psychological (Sageman 2017;Neumann 2013;Kundnani 2014). Ideological radicalisation refers to an ideological shift: acquiring extremist ideas.
In the case of Muslims, this is adopting a literalist understanding of Islam. Psychological radicalisation refers to a further shift: acting on the ideas in violent ways. In the case of Muslims, this means using violence to forward the individual or group cause, such as Al Qaeda and ISIS or affiliate groups. Although these definitions appear to be separate, some terrorism scholars suggest that they are connected because ideological radicalisation leads to psychological radicalisation (Neumann 2013;Sageman 2017). These definitions have enabled governments to use the conveyor belt theory of radicalisation to assess whether an individual is at risk from radicalisation or is radicalised. For example, the British government has adopted the conveyor belt theory of radicalisation as part of its counterterrorism strategy because it sees a literalist understanding of Islam leading to violence (Prevent UK 2011).
Despite the distinction between ideological and psychological radicalisation, there remains vagueness and difference around what constitutes ideological radicalisation and what scholars, and governments should focus on to further understanding and develop policies. The ambiguity around what constitutes ideological radicalisation rests on what counts as extremist ideas, which differs depending on the context (Neumann 2013). Hence, what is deemed extremist within one context and between contexts differs, resulting in a different configuration of ideas that a community, a society and a government deem as dangerous and constituting radicalisation. This difference is due to socio-political and cultural contexts and whether radicalisation is understood through the lens of old terrorism or new terrorism (Crenshaw cited in Karawan et al. 2008). Therefore, definitions of radicalisation developed in the West may not be suitable for Muslim majority countries because the history, socio-political environments and culture, and corresponding values and norms are different. Applying such definitions is likely to capture phenomena that reflect Western understandings of radicalisation, rather than grasping the reality on the ground in Muslim majority countries. Two examples will clearly illustrate the point being made. Firstly, radicalisation models such as the "Model of Jihadization and the Typical Radicalisation Pattern" are designed by Western scholars based on Western experiences to identify pre-radicalised and radicalised persons. These models have become popular in many countries as a sure and uncomplicated way, including Muslim majority countries like Indonesia, to detect pre-radicalised and radicalised individuals. In the case of Muslims, these models are used to identify signs that will determine if an individual is pre-radicalised or radicalised. The signs include religious symbols, religious vocabulary and changes in religiosity. In a Muslim majority context like Indonesia, the models are problematic because the signs of pre-radicalisation and radicalisation are part of everyday cultural life. Using these signs as indicators to identify pre-radicalised and radicalised individuals means that most of the Indonesian Muslim population would either be pre-radicalised or radicalised, needing preventative intervention. Secondly, if Indonesian scholars use surveys developed by terrorism scholars with Western heritage and based in Western countries to identify Islamic fundamentalism in the country, then most Indonesian Muslims would be identified as being fundamentalists and, therefore, vulnerable to radicalisation or being radicalised and necessitating preventative intervention. As Roy (2008, p. 2) argues, the process of violent radicalisation, as he calls it, has little to do with religious practice, while radical theology such as Salafism from the perspective of non-Salafis, does not necessarily lead to violence. Furthermore, Kundnani (2012, p. 21) has argued that Salafi registers are a way for Salafis to articulate their worldviews, but like Roy, he adds that this in itself is no evidence that Salafism is causing violence, but merely that the theological register that Salafis appeal to is being used to secure legitimacy within a given milieu. This suggests that a better understanding of radicalisation is needed. One which does not appeal to the logic of new terrorism and is rooted in sociopolitical conditions that act as a trigger for an individual to acquire and act on extremist ideas.
Although the two definitions of radicalisation focus on two ends of the process, some scholars deem psychological radicalisation more important than ideological because of its connection to violence (Sageman 2017). They argue that many people may hold extremist ideas, but very few act on them in violent ways, and therefore, it is more important to focus on psychological radicalisation. However, scholars who advocate a state-centric approach and adhere to the conveyor belt theory contend that ideology is a precursor to violence and should be prioritised (Neumann 2013). That said, both views are correct depending on the positionality of the observer (new terrorism, state-centric or critical approach to terrorism). Perhaps the best way out of this quagmire is to pay close attention to the advice Professor Mark Sedgewick's gives in his 2010 article on radicalisation to governments and scholars. He advises that it is best to see radical and radicalisation as relative concepts and always refer to the continuum and location of radical and moderate on the continuum. Terrorism scholars should be aware that concepts like radicalisation are politicised and part of a broader political agenda, which is connected to the risk of radicalisation and governments' ideological agendas. Although he refers to Western countries, but the same advice is pertinent for non-Western scholars working on terrorism because they also need to refer to the location of "radical and moderate" on their local continuum when they research into terrorism, advise their governments and talking in the media.
However, these definitions are not without criticism. Some scholars have called radicalisation a myth (Hoskins and O'Loughlin 2009). Others, like Mark Sedgewick (2010), argue that the concept of radicalisation over-emphasises ideology and de-emphasises the non-ideological circumstances that may lead some to become radicalised. Numerous authors have argued that non-ideological factors can also cause radicalisation. These factors include the foreign policies of Western countries, the unilateral invasions of Afghanistan and Iraq under the guise of the War on Terror, the Palestinian-Israeli conflict, and support for authoritarian regimes in Muslim majority countries by Western countries (Demant and de Graaf 2010;El-Said 2015;Ganor 2002;Hegghammer and Wagemakers 2013;Kuhle and Lindekilde 2012;Kundnani 2014;Lutz and Lutz 2015;Ramakrishna 2014;Sageman 2014;Schmid 2013;Sedgewick 2010). Other scholars have mentioned that radicalisation is motivated by thrill-seeking, injustice experienced by an identifiable and relatable other (religion, ethnic and nation), identity seeking/belonging, fidelity, humanitarianism, economic and political reasons and the desire for revenge (Schmid 2013;Borum 2011;Agnew 2010;Borum and Fein 2017). Other terrorism scholars such as Kundnani (2014) have argued that the concept of radicalisation from the outset was designed to serve the state's interests. Schuurman and Taylor (2018) contend that critics have deemed radicalisation subjective, lacking in an agreed definition, linear and deterministic. Silva (2018) argues that governments ignore critical research favouring research that is in line with their political and ideological agendas. Nasser-Eddine et al. (2011) and Neumann (2013) correctly highlight that radicalisation experts only agree that radicalisation is a process.
How a government understands radicalisation will determine its prevention strategy and how it deals with those deemed radicalised or engaged in terrorist violence. Most governments such as the UK adopt a precautionary prevention approach by employing drain the swamp strategy with its roots in Iraq to remove extremist groups from the country (Piazza 2007;Wintour 2013). This strategy is twofold, and firstly, it aims to remove conditions that governments deem as causing radicalisation, such as illiberal ideas and values and politics that the government consider as being against the country's social fabric and political interest. Secondly, the government aims to fill the swamp with liberal ideas, values, civil rights and politics that are in line with its interests. This approach is not necessarily concerned with the now but with the future. It aims to change the future by responsibilising citizens and to utilise citizens and institutions as frontline informers to drain the swamp and simultaneously roll out community initiatives that advocate liberal ideas and values, civil rights and government orientated politics to inoculate communities that are deemed at risk from radicalisation.
In the case of Muslims, the goal in most Muslim majority and minority countries is to drain the swamp of what the government considers extremist ideas and values and politics. To inoculate Muslims, depending on the country, governments may employ moderate or a state-centric interpretation of Islam or combine Islam with certain liberal ideas and values and civil rights and government-sanctioned politics.

Problems with De-Radicalisation
This section of the paper discusses the conceptual problems connected to de-radicalisation. De-radicalisation is a complicated concept and process because it is often understood by governments, including what terrorism scholars call disengagement, which means renouncing violence and leaving terrorist groups. Although de-radicalisation and disengagement are different processes, they are often combined to develop de-radicalisation programs because it is assumed that one would lead to the other. In the case of Muslims, deradicalisation means ideological transformation, by which an extremist understanding of Islam is renounced, and a moderate one is adopted, which in some instances is state-centric (Agastia et al. 2020). Advocates of de-radicalisation hope that radicalised individuals will be entirely and surgically inoculated by the end of the program and pose no risk to society.
The other part of de-radicalisation is disengagement, which is concerned with getting radicalised individuals to psychologically and physically disengage from the use of violence and terrorist groups (Silva 2018). Like radicalisation, disengagement may unfold differently for different radicalised individuals. For some individuals, small incidents may lead them to start the disengagement journey, but for others, it could be big events (Horgan and Bjorge 2009). Other factors, such as duration of membership of the group, family connections with other group members, fatigue and change in leadership and failure to achieve goals also play a role in a radicalised individual's disengagement journey. The disengagement journey involves two connected processes; one is psychological and the other physical. Negative experiences characterise psychological disengagement, change in personal priorities and disillusionment with the group not achieving its goals. Physically disengagement is perhaps straightforward, and it can occur because of imprisonment, change of role in the group, or being removed from the group.
Although de-radicalisation and disengagement are often used interchangeably to mean the same thing as mentioned earlier, they are different in terms of their aims (Schmid 2013;Horgan 2014;Pettinger 2017). Raets (2017) has argued that there is much confusion over their meanings and aims and what both processes mean stand on shaky ground. This situation is logical because radicalisation is ill-defined; as such, it will impact how de-radicalisation and disengagement are also defined. In reality, they are two quite distinct processes and have different meanings and goals, and they should be seen as such. However, in the imaginations of governments, who are looking for a silver bullet solution, they are seen as the same, because one is conceived as leading to the other (Silke and Veldhuis 2017; Horgan 2008; Anindya 2019). However, de-radicalisation does not always lead to disengagement, and disengagement does not always lead to de-radicalisation, and any program that tries to achieve both, is probably not going to succeed.
Despite being different, de-radicalisation and disengagement are often combined to make one program called the de-radicalisation program, which is the case with the Indonesian programs that are discussed later sections. These programs are often simple quick fix tools to deal with a range of complex intersubjective socio-political phenomena and psychological processes, which cannot be resolved over a few weeks or months by "laboratory-style" interventions. Combining and relegating both processes to simple laboratory-style quick fix tools without considering contextual factors may be useful for electioneering and media publicity, but it runs a high risk of being ineffective and increasing recidivism. Perhaps the biggest problem with the de-radicalisation programs is that they are plagued with a lack of clear empirical criteria to evaluate their success, which has in the past led government officials to grossly exaggerate the success of their programs (Koehler 2017a). Success or failure of any de-radicalisation program, at least in the governments' and public's eyes, depends on recidivism rates.
However, terrorism scholars have argued that using recidivism rates is not the best measure to determine the success of de-radicalisation programs, because the base rate of recidivism is unknown, and therefore, it is problematic to use recidivism as proof of success (Koehler 2017b;Rabasa et al. 2010). Terrorism scholars have argued that other factors need to be taken into account to determine the success of a de-radicalisation program, including governments providing accurate information; identity of who has been de-radicalised and disengaged (low-level operative, opportunistic operative and sympathisers); tracking of de-radicalised individuals over a long period of time; resources available; and on the ground political dynamics.
However, some countries have claimed that they have very low recidivism. For example, Singaporean and Saudi Arabian officials have claimed that their de-radicalisation programs are successful (Horgan and Braddock 2010;Rabasa et al. 2010;El-Said 2015). That said, questions remain over what constitutes success for these programs. For example, have hardcore members of terrorist groups been de-radicalised and disengaged or only low-level operatives, opportunistic operatives and sympathisers? Furthermore, are those deemed as de-radicalised and disengaged monitored after release. If this is the case, this raises questions over whether the individuals have really been de-radicalised and disengaged (El-Said 2015). However, other countries have been less successful, such as Indonesia, Malaysia, England and Wales, Northern Ireland, United States, Spain and the Netherlands (Renard 2020;Ismail and Sim 2016). In this situation, it is probably best for governments to just focus on disengagement, rather than hoping for de-radicalisation, because there is no guarantee or sure way to identify if a radicalised individual has truly been de-radicalised and poses no threat to society. However, there is a better chance of a radicalised individual to disengage (Schmid 2013). As Silke (2011) argues, prioritising de-radicalisation over disengagement is a dangerous gamble. In the following section, the paper assesses whether the de-radicalisation programs developed and implemented in Indonesia are successful.

De-Radicalisation in Indonesia
This section of the paper will critically review de-radicalisation programs in Indonesia, but before doing this it is important to mention that radicalisation in Indonesia is deemed to be caused by Salafi Islam, which is considered the antithesis of Pancasila (national philosophy) and Islam Nusantara (traditional Islam in the Malay and Indonesian archipelago), as discussed below. Such an understanding is understandable because detainees convicted of terrorism offences see the Indonesian government as un-Islamic and therefore illegitimate (Hilmy 2013). The aim of terrorist groups that these detainees belong to, especially those allied to transnational terrorist groups such as Al Qaeda and ISIS, are to transform Indonesia into an Islamic state, despite the country being a Muslim majority. However, as mentioned earlier, such an understanding overlooks how local socio-political and international conflicts lead some individuals to acquire extremist ideas, become radicalised, and engage in violence.
The best way to assess Indonesia's de-radicalisation programs is to start chronologically with the government's programs and then CSOs' programs. The Indonesian de-radicalisation programs combine de-radicalisation and disengagement but are often referred to as de-radicalisation programs. These programs not only aim to de-radicalise detainees by persuading them to adopt Islam Nusantara, but also create conditions for disengagement by offering them life coaching, help with starting a family, financial assistance for their families, upskilling and even financial help to set up a business (Silke and Veldhuis 2017;Irwin 2015).
The Indonesian government's first program was aimed at JI members who had been convicted of terrorism offences. The program was presented as being focussed on deradicalisation and disengagement, but in reality, its goal was to gather intelligence on extremists and prevent future terrorist attacks (Osman 2014). If de-radicalisation or disengagement did occur, it was a bonus. For the detainees' de-radicalisation involved offering them a contextualised reading of Islam to persuade them to adhere to Islam Nusantara and Pancasila. The program was based in prison and set up by Suryadharma, the head of Densus 88 (the police counterterrorism team) and continued under his successor, Tito Karnavian (Hwang 2018). The approach was dubbed the soft approach because of its humane outlook (Rabasa et al. 2010) and centred on Densus 88 officers building relationships with the detainees (Istiqomah 2011;Sukabdi 2015). This situation often resulted in the officers praying alongside the detainees and sharing iftar (break of fast), and in some cases, facilitated the marriages of the detainees and support for their families (Abuza 2009;Rabasa et al. 2010;Hwang 2018). In some cases, officers rewarded compliant detainees with perks, such as better detention conditions and reduced sentences (Osman 2014). This approach was adopted because it was deemed that by treating the detainees humanely, they would be more malleable for intelligence gathering on other extremists (Hwang 2018).
The de-radicalisation part of the program was conducted by former terrorists Nasir bin Abas and Ali Imron because it was believed that they would have more influence over the detainees than moderate scholars. Abas is a former JI member and fought in the Afghan "War against the Soviet Union" during the 1980s, while Imron was involved in the ethnic conflict in Ambon during the early 2000s (Abuza 2009;Hwang 2018). Involving former extremists, especially those who have had combat experience, is deemed vital because they have credibility and garner respect among the detainees due to their past exploits (Hwang 2018;Abuza 2009). Former extremists often have a more profound impact on detainees than moderate figures and can clarify and contextualise specific Islamic ideas often used by extremist groups to foster disunity and engender violence, such as those mentioned earlier (Osman 2014). Abas and Imron's involvement in the program resulted in some success because they managed to convince detainees to renounce violence to achieve political goals (Aslam et al. 2016). That said, the Densus 88 program's success is difficult to gauge because it depends on what counts as success; is gathering intelligence success, is detainees renouncing extremist ideas success or is disengaging from groups and violence success? In the Indonesian government's case, success meant gathering lots of intelligence because the program was set up to gather intelligence. Although some program graduates have not returned to violence, they have not renounced their extremist ideas either (Osman 2014). Others have re-joined terrorist groups, often citing their hatred for the authorities (Istiqomah 2011;Taufiqurrohman 2015). The program at one level can be viewed as a success because some radicalised individuals have disengaged but another level for some scholars and neighbouring countries it is a failure and a "bomb waiting to explode".
Despite the perceived success and the humane outlook of the de-radicalisation program, it was beset with problems. Firstly, the program was never set up for de-radicalisation because the Indonesian government set it up to gather intelligence rather than de-radicalise or disengage, and if both took place, it was a bonus (Rabasa et al. 2010;Hwang 2018). Secondly, detainees complained about being tortured by Densus 88 officers, which is likely to reinforce their commitment to extremist ideas and terrorist groups. Moreover, Densus 88 have also been accused of targeting Islamist activists and extrajudicial killings and being a foreign intervention into domestic affairs because of the funding Densus 88 has received from the US and Australia post 9/11 (Hilmy 2013;Lamchek 2018). Thirdly, the program was ad hoc and underfunded. It is not clear why the Indonesian government underfunded the program. There are two possible reasons for this. One, it may have been seen as expensive. Two, the government may not have had the patience to wait for results because de-radicalisation and disengagement are long term processes; even then, success is not guaranteed.
Since the initial program, the Indonesian government has set up another program, and like the previous one, it is a combination of de-radicalisation and disengagement. It caters for Indonesians who have been deported from various countries for being radicalised or have aspirations to join a terrorist group or fight in the Syrian civil war (Sumpter 2018;Anindya 2019). The deportees and returnees include men, women and children. These individuals are deemed a significant security risk because some have had military training and have frontline fighting experience and could carry out attacks in Indonesia (Mcbeth 2019).
The government's de-radicalisation program is very short and lasts for one month and involves a psychological assessment to ascertain the level of radicalisation, which is based on using Islam; religious re-education by using Islam Nusantara; life coaching (counselling and life skills); signing a contract to abide by the Pancasila; and after-care (Anindya 2019). Like the previous program, the government deems radicalisation as being caused by a literalist understanding of Islam, which is clear from psychologists' indicators and the use of imams to persuade detainees to adopt Islam Nusantara. The program seems not to be entirely successful because some deportees upon release still deem the Indonesian government un-Islamic (Anindya 2019). The program also employs forceful persuasion tactics to convince the deportees to sign a contract that states that they will abide by the Pancasila, which perhaps is an indication of why the program is not successful (Sumpter 2018). The after-care part of the program is organised and delivered by local CSOs, focusing on training deportees with life skills and finding a sustainable form of employment or establishing a business (Anindya 2019). The ultimate aim of such programs is to ensure that the detainees do not return to the old networks and engage in violence. However, finding employment for detainees is not easy because they face several problems, which are discussed later in this paper.
However, like the Indonesian government's first de-radicalisation program, the current program is also beset with similar problems. These problems include the length of the program (one month is not adequate time for any meaningful change to take place, especially for committed extremists); organisational concerns such as lack of resources; disorganisation between stakeholders; competition between stakeholders; lack of training for employees; use of force and no clear guidelines for or understanding of different aspects of the program (Anindya 2019; Agastia et al. 2020). Finally, the program does not address the political motivation of detainees and only focuses on Islam. As mentioned earlier, there are many reasons why someone becomes radicalised (Borum and Fein 2017;Agnew 2010;Nilsson 2015;Malet 2015;Osman 2014). Consequently, only providing Islamic re-education may not be the appropriate method to de-radicalise the detainees. Instead, what is needed is a tailored approach that also considers the political motivations of the detainees. This situation will require more resources and a well-thought-out tailored approach, which increases the chances of long-term disengagement and may foster conditions that could lead to detainees abandoning extremist ideas and reducing recidivism.
Aside from the Indonesian government's de-radicalisation programs, some CSOs have set up programs that mainly focus on disengagement. Such a set up may sound bizarre because radicalisation is deemed as being caused by a literalist understanding of Islam, but there is a good reason for this. According to Osman (2014, pp. 220-23), there is a tacit recognition in Indonesia that de-radicalisation is not possible in most cases and is shaped by the country's socio-political history and the detainees being well versed in Islam. In some cases, the former detainees' Islamic knowledge exceeds that of the CSOs and discussing/disagreeing on Islam could also build resentment among detainees and therefore, make the job of CSOs much harder. In other cases, the former detainees consider moderate or state-appointed Islamic scholars as enemies of Islam and therefore, do not trust them (Rabasa et al. 2010). Additionally, some former detainees reject de-radicalisation because what it entails contradicts what they believe to be the correct Islam (Hilmy 2013). Given this situation, CSOs have to walk a delicate path and decide if de-radicalisation or disengagement will achieve their desired goal. If de-radicalisation is pursued from the start, it may mean that the detainees do not cooperate. However, if disengagement is chosen, then there is a possibility that disengagement could foster conditions that may lead former detainees to abandon extremist ideas.
Despite this recognition, the de-radicalisation programs run by CSOs do have a deradicalisation component, but it is not overtly obvious. The programs are structured around three key areas-life skills training, business entrepreneurship and de-radicalisation. For example, the best-known de-radicalisation program, which received global acclaim, is set up by Yayasan Prasasti Perdamaian (YPP) to disengage former detainees (Power 2018). The program focuses on providing life skills and opportunities for business ventures for former detainees who have decided to disengage from terrorist groups and violence. Through this approach, YPP aims to foster economic conditions favourable for successful disengagement and the ultimate renunciation of extremist ideas. The business ventures include cafes, restaurants and t-shirt printing (Hwang 2018). Although the program does not explicitly focus on religious re-education, it is introduced during the latter part of the program after trust has been built with the former detainees, which is understandable because having trust can facilitate discussion on sensitive issues. The program's success seems to rest on three factors-trust/loyalty-building, assistance with personal matters and help to establish a sustainable business.
However, the program only attracts those interested in disengagement and those who have already renounced extremist ideas before enrolling on the program. Hence, the experience of being imprisoned and the discovery of ideological differences with other extremists (Osman 2014) may have already triggered what Muhanna-Matar (2017) has called self-de-radicalisation, which is based on self-questioning and which sometimes occurs as part of the disengagement process (Horgan 2008). This situation begs the question of whether the program's success is really due to the program or whether the prisoners were already disengaged, and the program just offers an opportunity to start a new life. Silke (2011) argues that most terrorists disengage without going through de-radicalisation programs because of their experience of being part of a terrorist group. If this is the case for the YPP program, then it can be argued that the program is less about de-radicalisation and disengagement and more about welfare support, upskilling and business entrepreneurship. Furthermore, like most programs that aim to either de-radicalise or disengage extremists, there is no sure way of measuring if the individual has truly been de-radicalised and will never re-engage in violence or join a terrorist group, apart from recidivism rates (Horgan and Bjorge 2009;Horgan and Braddock 2010). Moreover, abandoning extremist ideas or disengaging from violence may only be selective or conditional, depending on continued assistance or no provocative actions by the adversary (Hwang et al. 2013;Clubb 2009;Rabasa et al. 2010). Despite the trust-building and financially helping detainees, some former detainees who graduated from the program have returned to violence (Osman 2014). Additionally, the biggest obstacle for YPP in producing results is the lack of sustained funding.
Other CSOs working in the area of de-radicalisation also employ a similar approach to YPP. The CSO called the Search for common ground "conducts workshops in prisons on anger management and conflict resolution for high and low-risk detainees" (Osman 2014, p. 224;Hwang 2018). The organisation offers seed money to encourage detainees to disengage and has a dual aim. The first is to prevent the detainees from re-joining their former groups and, secondly, to train the detainees in conflict resolution, so that they can implement the training in their own lives and train others (Hwang 2018). However, the program only attracts low risk and less dangerous detainees, rather than the high-risk ones. The program can be understood as both a success and a failure. It is a success because it has diverted low-risk detainees from possibly engaging in violence in the future and provided them with the tools to negotiate their way out of conflicts. It is unsuccessful because high-risk detainees could influence low-risk detainees have not shown any interest. Other initiatives have included coaching the wives of the detainees to become self-sufficient by helping them to establish home-based businesses (Osman 2014).

Labelling and Stigmatisation
This section of the paper discusses how labelling and stigmatisation can hinder deradicalisation. Earlier in the paper, it was mentioned that some detainees post-release struggle to find employment and therefore, could not have normal lives. This situation occurs because of labelling and stigmatisation, which the Indonesian CSOs have highlighted as a significant hindrance to disengagement and reintegration. For example, CSOs have argued that using labels such as countering violent extremism and de-radicalisation are unhelpful and counterproductive because they engender a range of consequences for the government, for CSOs and for former detainees (during detention and post-release) (Sumpter 2017). This situation is especially the case if the detainees perceive the program to be politically motivated or conclude that the aim is to undermine their religiosity (Sukabdi 2015;Sumpter 2017). If former detainees think the CSOs are trying to de-radicalise them, then there is a high risk of them not taking part in the de-radicalisation programs, which is why de-radicalisation is subtle and occurs at the backend of the programs after trust has been secured.
A label can have positive or negative connotations depending on what action the label is trying to capture, who is doing the labelling, and the location (neighbourhood, city or country). In criminology, labelling theory explains how being labelled, for example, a criminal or terrorist, is experienced or lived out by the labelled (Appleby 2010). In some cases, being labelled negatively can result in being treated negatively by wider society, the media and government institutions in ways that engender stigmatisation, which can foster othering-insider vs. outsider (Cherney and Murphy 2017). Experiencing stigmatisation can result in the stigmatised community or the individual feeling resentful and unwilling to cooperate with the authorities (Tyler et al. 2010). In the case of former detainees, this could mean re-joining terrorist groups and re-engaging in violence.
The continuing use of labels such as terrorist or former terrorist to describe detainees and former detainees by the government, the media and Indonesian society, even after they have undergone de-radicalisation and disengagement, is likely to lead to more stigmatisation and obstacles to reintegrating the former detainees. The former detainees may view the government with disdain and continue to see the government as un-Islamic and an enemy. Such labels are also likely to alienate the former detainees from their own or other communities (even those willing to cooperate with the state and CSOs). Additionally, the labelling is likely to hinder the reintegration process because the community may reject them and perceive the former detainees and their families as a threat.
The impact of labelling and stigmatisation does not remain within the walls of prisons. The families of the detainees also face stigmatisation from their communities, which could push them to extremism. Stigmatisation from society in general, and specifically from communities that the former detainees originate from, is a significant obstacle that can hinder the reintegration of former detainees. This situation occurs because local media is often informed about the former detainee's return, resulting in the former detainees moving around to find a community that will accept them (Anindya 2019). In some instances, former detainees' stigmatisation is fuelled by intelligence agencies that give warnings to the neighbours of former detainees about their potential threat (Sumpter 2017). Former detainees have stated that reintegration and reduced stigmatisation are two essential criteria for what constitutes a successful rehabilitation (Sukabdi 2015). The stigmatisation by the former detainees and their family can severely undermine the hard work put in by CSOs and result in resentment and ultimately recidivism. Moreover, it is unlikely that the businesses of former detainees will work or be successful if the larger Indonesian society and the local community continue to view them with suspicion and a threat. The usefulness of any skill that is developed during their detention is predicated on the larger Indonesian society trusting them post-release. The absence of trust and the existence of broad stigmatisation will inevitably hinder any reintegration and increase the chances of resentment, a renewed commitment to extremism and recidivism. One way that could offset the labelling and stigmatisation faced by former detainees is to engage them in humanitarian work, because it can facilitate rebuilding trust between the former detainees and the larger Indonesian society and the local community that they are relocated to.

The Role of Humanitarianism
In this final section of the paper, humanitarianism is proposed, as a way to improve de-radicalisation. One way to increase long term disengagement and create conditions that encourage individuals to abandon extremist ideas, overcome problems associated with labelling and stigmatisation, promote trust between former detainees and the Indonesian society, is to develop tailored de-radicalisation programs that include humanitarian work.
The main reason for introducing humanitarian work is to address the experiences of collective strain, which can be experienced directly or vicariously and, in some cases, can lead to what Agnew (2010) has called linked-fate. The literature on radicalisation has shown that one of the pull factors that attract individuals to terrorist groups include grievances engendered by injustices experienced by fellow community members (Agnew 2010;Schmid 2013). The community can be religious, ethnic or the nation, and in the case of this paper, the community is the global Muslim community, the ummah. The combination of collective strain and linked fate can also lead to what has called a personal quest, which has led individuals to carry out extra-ordinary acts in defence of their community that includes the use of violence (Kruglanski et al. 2014;Swann et al. 2010). The use of violence can be argued to be motivated altruism because they act in defence of their community. For example, some Muslims and non-Muslims from different countries have travelled to conflict zones to initially provide humanitarian aid and because of the violence they witnessed first-hand to people they identified with, convinced them to remain and fight and defend their brethren (Butt 2019), while others either travelled to conflict zones or carried out attacks on home soil to defend people that they identified with because they were experiencing violence at the hands of a powerful other (Nilsson 2015;Mustapha 2013;Malet 2015). For example, Mohammad Sidique Khan, one of the suicide bombers that carried 7/7 terrorist attacks in London. In his suicide video, he mentions that "your democratically elected governments continuously perpetuate atrocities against my people and your support of them makes you directly responsible, just as I am directly responsible for protecting and avenging my Muslim brothers and sisters" (The Sunday Times 2005). Another example is of a foreign fighter who travelled to Syria to fight against the Bashar al-Assad regime. He mentions that "elderly women, the children were being killed by a brutal regime, I think that deeply affected me, it touched me, and I thought I could do something, to help, in one way or another, that's what pushed me to go" (Channel 4 2015).
From both examples, it is clear that the suffering of their brethren, injustice and linked-fate pushed them to engage in terrorist violence and become foreign fighters.
However, the decision-making calculus behind joining terrorist groups or becoming a foreign fighter is deemed by the individuals as high reward, with little attention being paid to the risks involved. In reality, joining terrorist groups or becoming a foreign fighter is high risk and low impact and reward because there is an increased likelihood of being arrested or killed before making any impact in the conflict. Herein lies the value of including humanitarian work in de-radicalisation programs because by addressing the decisionmaking calculus, there are increased chances of disengagement and fostering conditions that may lead some detainees to abandon extremist ideas and which could lead to the reduction in recidivism.
Humanitarian work can be incorporated into existing or new de-radicalisation programs. Such work can be cloaked by using the language of Islam and human rights. For example, reinterpreting terms like jihad and using Quranic verses like whoever saves oneit is as if he had saved mankind entirely could be used as the basis to develop programs that include humanitarian work. Such programs would be high impact in terms of the effectiveness and be low risk because they would be legal; high impact because some of the motivations of former detainees joining terrorist groups and becoming foreign fighters will be addressed, and they can see how their efforts are changing the lives of fellow Muslims. Such programs could also be an effective way to persuade former detainees to abandon extremist ideas because the exposure to fragility and precariousness of human life through humanitarian work would foster conditions for CSOs to engender discussions on sensitive issues about interpretations of Islam and alternative understandings and solutions to conflicts.
Examples of humanitarian work as part of a tailored de-radicalisation program could include charity work involving disaster relief, conservation (cultural and environmental), and development and fundraising for disease prevention/cure research. Ultimately, deciding what type of humanitarian work a former detainee should engage in should be negotiated between them and CSOs. However, such a de-radicalisation program will need sustained funding and must be long term and include an extensive risk analysis of former detainees, and former detainees must undergo regular evaluations to assess their rehabilitation. Additionally, the exposure that humanitarian work would bring would also help dilute the broad stigmatisation that detainees face post-release because Indonesian society would see that they are trying to make amends for their past and craft a new future. The dilution of the stigmatisation would help strengthen their sense of acceptance by the community and belongingness.

Conclusions
Indonesia, over the last few decades, has experienced many terrorist attacks from homegrown terrorists. As such, the government and CSOs have developed programs to de-radicalise detainees and help them seek employment, hoping that they would not re-join terrorist groups upon release. The success of these programs is difficult to measure because there are many ways to measure success, as discussed earlier in the paper. On the whole, the Indonesian de-radicalisation programs have had some success and have implemented some innovative ideas to encourage disengagement and reduce recidivism.
Although the Indonesian de-radicalisation programs have had some success, they have been beset with problems, such as the lack of funding, de-radicalisation programs being too short and ideologically skewed. These problems can be alleviated by taking the following measures: (1) The Indonesian government needs to fund its de-radicalisation and disengagement programs sufficiently and those developed by CSOs; (2) The Indonesian government needs to sufficiently fund its after-care programs and those set up by CSOs; (3) All the after-care programs need to be well thought out, well planned, diverse, include humanitarian work and be well funded; (4) The duration of de-radicalisation programs needs to be more than a few weeks because renouncing extremist ideas and disengaging from extremist groups and violence can take a long time; (5) The Indonesian government needs to develop and implement tools to monitor the success of the de-radicalisation programs; (6) The Indonesian government needs to develop ways to reduce the stigmatisation experienced by detainees during their detention and post-release; (7) The Indonesian government and CSOs should design de-radicalisation programs that incorporate humanitarian work. However, there is no guarantee that former detainees after going through a de-radicalisation program will not renounce extremist ideas or re-join terrorist groups because it is not possible to predict the impact of future events. However, addressing the aforementioned issues will lead to comprehensive de-radicalisation programs, which will not only increase the chances of reducing short-term but also long-term recidivism, and improving the chances of reintegration of former detainees.