Military influence via social media : the challenge of changing behavior without using force

(1)

July 3th 2014

Master’s thesis in Sociology

Track: Social Problems and Social Policy

Student: G.M.G. Nielen 10295518

Email: gwenda3dec@hotmail.com

Supervisors:

1

st

_{Drs. Thomas G. Kampen}

2

nd

_{Dr. Barbara da Roit}

U

NIVERSITEIT

VAN

A

MSTERDAM

MILITARY INFLUENCE VIA SOCIAL MEDIA

(2)

1

PART

1

(6)

5 1. INTRODUCTION

“For to win one hundred victories in one hundred battles is not the acme of skill. To subdue the enemy without fighting is the acme of skill.”

(Sun Tzu 1963: 77, originally dated around the 6th_{century BC)}

In the times of Sun Tzu armies were deployed to expand or defend territories. While professional defence forces of non-NATO countries and more irregular armed factions are still involved in armed conflicts on their own territory, NATO members’ armies are in general deployed outside their home country. This military deployment to resolve worldwide security issues is in most cases mandated by either NATO (North Atlantic Treaty Organization) or the UN (United Nations). These two

transnational organizations are concerned with global security issues and can and will call for military intervention when considered necessary.1_{The foundation on which both organizations base}

their legitimacy consist of a list of articles which are internationally acknowledged. The third article from the universal declaration of human rights (UN 1948) states that ‘Everyone has the right to life, liberty and security of person’ and the fifth article reads that ‘No one shall be subjected to torture or to cruel, inhuman or degrading treatment or punishment’. In article 5 of the North Atlantic Treaty (NATO 1949) it is stated that ‘…action as it deems necessary, including the use of armed force, to restore and maintain the security of the North Atlantic area… are taken forthwith when an armed attack on one or more member states occurs.’ These articles emphasize the role for military organizations in providing worldwide security for people, seen as an important precondition for living. The deployment of military troops to provide security is based on their legitimacy to use lethal weapons and physical force. But the quote above though, from the classic work ‘The Art of War’ (Sun Tzu 1963) on military strategy, already made clear that a military organization must be capable of doing more than using lethal weapons and physical force to be ultimately successful. This point of view is also reflected in the most recent publication of Army Doctrine (2013: paragraph 1503) in which it is said that when goals can be realized without the use of physical force, such an approach is to be preferred. Former British General Andrew Mackay and Commander Steve Tatham (Mackay & Tatham 2011) also make a clear case for other ways besides the use of firepower to engage in military operations. They propose a change in the conduct of military operations and argue that changing behavior in areas of operations without using physical force is likely to become a pre-eminent factor in securing future success of military operations. As the success of military operations

1_{The UN Security Council defines its primary responsibility as the maintenance of international peace and}

security (http://www.un.org/en/sc/) and NATO sees its essential purpose in safeguarding the freedom and security of its members through political and military means (http://www.nato.int/nato-welcome/index.html).

(7)

6 is related to worldwide security and stability it is important to explore the possibilities for new ways of conducting military activities.

1.1 MILITARY ACTIVITY TO INFLUENCE BEHAVIOR

In general, the objective of military operations is to influence the behavior of specific people in an area of operations. By influencing the behavior of people in an area of operations the military organization wants to synchronize their behavior with the goals of the military operation. To link this behavior with military goals a general categorization will be used in which a differentiation is made between undesired behavior, neutral behavior and desired behavior in relation to objectives of the military organization. This means that behavior that is negatively influencing the accomplishment of the objectives of military operations can be labelled as ‘undesired’, behavior that contributes the accomplishment of military goals can be labelled as desired and behavior that is neither

counteracting nor contributing in respect to the military objectives can be labelled as ‘neutral’. To influence behavior the idea of target audiences is derived from social marketing strategies (Kotler & Zaltman 1971). This means that by the identification of Target Audiences (TA’s) communication to influence behavior can be directed on a specific person or a specific group of people. This points out that to be able to reach and influence a target audience communication is indispensable. Three different dimensions can be identified concerning the military process of communication to

influence behavior. The first dimension is related to the distinction between kinetic and non-kinetic communications (Mackay & Tatham 2011). Military activities that are influencing behavior through the use of physical force (or directly threatening with it) are referred to as kinetic activities. All other military activities can be labelled as non-kinetic activities which are indirectly influencing behavior through the use of means that are not potentially lethal or physically damaging. The second dimension is related to the modus of communications. Following the classification of The

Netherlands Organization for Applied Scientific Research (TNO 2011) the modus of communication can be defined as watching, sending, asking or interacting. The third dimension that can be assigned to communication is the approach that can either be direct or indirectly aimed at people. As the basic transmission model of communications (Shannon & Weaver 1949) makes clear, this

communication between a sender and a receiver must be facilitated by a channel. This channel or medium can have many appearances. Examples are a flyer, a computer, a bullet or money, but there are many more different media that can be identified to facilitate communication. So the sender can be seen as an actor using a medium to send information to a receiver. This means that a flyer must be spread, a message on Facebook must be posted and a bullet must be shot. In Figure 1 the

(8)

7 are summarized to provide a simplified overview of how a military organization uses

communications to influence the behavior of a target audience.

1.2 RELEVANCE

The time that the military changed behavior only by the use of physical force has past like Sun Tzu has already advised and influencing behavior by non-kinetic means is preferred to the use of physical force. A nonphysical environment in which non-kinetic communication is conducted is the internet and more specifically social media. That social media is already used by organizations to influence the behavior of people becomes clear when surfing the internet and being confronted with the numerous marketing efforts that are present on almost all webpages. This also became clear from the experiences of expert respondents that were consulted during the first explorations of the field. Besides two experts from commercial companies, one expert from the communication branch of the Dutch National Police and one expert from a NGO also report that they have had fruitful experiences with influencing behavior via social media. The question is though, if it works in a specific context for a commercial company, the Dutch National Police or a NGO, can it work for the military organization as well? It seems worthwhile to at least investigate the applicability of social media for the military organization of influencing behavior in areas of operations.

A more general trend that motivates the investigation of social media is the increasing time people spend communicating through the internet, the increasing amount of devices connected to the internet and the rising use of social media. It might even be concluded that face-to-face social Figure 1. Military Activity to Influence Behavior

(9)

8 networks are being extended or even partly replaced by cyber social networks (Wellman 2004). This implies that if the military organization wants to reach people to influence their behavior with non-kinetic means they will have to be on social media. From a report published by Cisco (2013) about the worldwide internet connectedness it can be concluded that while in 2012 there were 2.3 billion internet users, it is estimated that by 2017 about 3.6 billion people are connected to the internet (which is almost half of the projected population of 7.6 billion in 2017). Furthermore it is predicted that by 2017, there will be more than 19 billion global network connections up from about 12 billion connections in 2012. Also by 2017, more traffic will traverse global networks than all prior ‘internet years combined. From 1984 until 2012 the traffic was 1.2 zettabytes while the 2017 forecast is 1.4 zettabytes just in one year. 68% Of this internet traffic will be generated through Wi-Fi and mobile-connected devices. Lastly it is important to mention that The Middle East and Africa, as current and potential areas of operation, will continue to be the fastest growing IP traffic region from 2012 – 2017 with a 5-fold growth. The telecommunications research site Buddecom (2014) has published a more detailed report on the key developments in Africa and the Middle-East. It states that as a result of the introduction and the current rapid expansion of third- and fourth-generation (3G/4G) mobile technology in Africa, broadband internet access for the mass market has become a reality.

Combined with the arrival of low-cost smartphones, locally manufactured in for example Ethiopia, this has lowered the barrier of entry to the internet for African people. Besides a reported grow in smartphones all over the continent it is also noted that the average smartphone data usage is rapidly increasing. While the image of Africa and to some extent also the Middle-east is often one of

underdevelopment both companies predict quickly growing opportunities to reach the population via internet and social media.

Another reason to investigate the use of social media for military activities is the potential power it has to influence behavior. Some examples that show this power are provided by Castells (2013A) in his book ‘Networks of Outrage and Hope: social Movements in the internet age’. It is explained how in Tunisia and Egypt as in most of the Arab revolutions, the uprisings started on the internet and were then taken to the streets showing the important role that social media played in the Arab revolutions. A similar conclusion is drawn by Tufekci and Wilson (2012) based on their analysis of the events on Tahrir Square. They show that controlling for other factors, social media use greatly increased the odds that a respondent attended protests on the first day. Additionally, more in general, Zangh et al. (2010) have examined the influence of social networking sites on political attitudes an behavior. They conclude as well that people can be influenced in their political behavior by communicating on social media. These publications indicate how powerful social media

(10)

9 can be in influencing people’s behavior and it point out that social media could indeed provide an environment to conduct non-kinetic military activities.

The military activities that have been most thoroughly explored and described in Dutch doctrines are those that are related to the more traditional kinetic warfare. One indicator that implies a growing importance of non-kinetic activities though is the observation that in times of drastic financial cutbacks and the dismantling of kinetic military assets in the Dutch army (Ministry of Defence 2011), the field of non-kinetic warfare is expanding. Within the Dutch military organization different functions can be identified that are involved in the non-kinetic influencing of people and also different organizational elements can be pointed out that have non-kinetic influence as main operational task. The functions can generally be divided in communications (Com), information operations (InfoOps), psychological operations (PsyOps), civil military co-operation (CIMIC) and to some extent also cyber operations (CyberOps) (NATO 2003, 2012). A brief description of these different functions is provided in Annex 1 (Non-kinetic influencers in the military organization). One unit that is completely focused on non-kinetic activities in the Dutch army is the Civil Military Interaction Commando (CMICo), officially established in October 2013 after the major cutbacks. Within this unit the Target Audience Analysis Procedure (TAAP, CMICo 2014, Annex 2) is used. This tool consisting of 10 practical steps provides a contextual understanding of the cultural, historic and social composition of target audiences. As becomes clear from Annex 2, the 10-step analysis is quite extensive. Remarkable is though, that in the 6th_{step it seems if ‘the internet’ as a medium has been}

added later. When trying to find a document that provided guidelines for the use of internet or social media to influence behavior it had to be concluded that it didn’t exist. This leads to the

conclusion that for the use of internet or social media by the Dutch military organization to influence target audiences there is no structural approach which provides another reason to investigate the subject.

But there are indicators that the Dutch military is interested in social media. One indication is that from a conference on social media (NISA 2014) it became clear that social media as a source of information is currently being evaluated and researched by many international parties varying from universities, intelligence services, military intelligence units and non-governmental

organizations, including the Dutch military as well. But also other indicators can be found of the Dutch defence Force’s interest in the opportunities and the threats of social media. One example is a report published by The Netherlands Organization for Applied Scientific Research (TNO 2014A) published in cooperation with the Dutch 43th Army Brigade. Even though in this report the use of social media is mainly explored for communications during a field exercise (as part of public affairs) it indicates that there are initiatives to investigate the use of social media. A broader exploration is

(11)

10 presented on a social media map for the military organization also designed by TNO (2014B), this time in cooperation with a more divers panel of interested parties and experts. The map provides an extensive and structured overview of 26 ‘threats’, arguments against the use of social media and 27 opportunities, arguments in support of the use of social media or opportunities. These indicators that point out the first explorations of social media by the Dutch military show that a further investigation is worthwhile.

Military operations are taking place now and will take place in the future. For this reason it is important to figure out how military organizations can bring security and stability with as limited physical force as necessary. Therefore it is relevant to encourage military thinking about non-kinetic approaches in general and more specifically on the use of social media to influence behavior. So the main questions is how the Dutch military organization can use social media to influence the behavior of people in areas where they are conducting military operations. Unfortunately, on first sight it can already be concluded that the scope of the answer to this question is far beyond the reach of this single paper.

Therefor the general aim of the study is to open a window that shows a view on a new landscape with numerous possibilities, some theoretical road signs to bring structure to the environment and a few practical examples to show how things can actual come together.

1.3 THEORY, METHODS AND FINDINGS

Because of the unusual construction of the study concerning the connections between theory, methods and findings I will shortly explain the evolvement of the study. To clarify the structure, the coherence of the different elements of the study is depicted in a model. Figure 2 shows how the different theoretical and methodological elements are related.

(12)

11 The use of social media in political processes has been studied quite extensively by for example Shirky (2011) and Zangh et al. (2010) but the use of social media by the military organizations to influence behavior in the context of a military operation has not been studied yet. So before an empirical research could be conducted it was necessary to explore the field to identify the most interesting theoretical starting points concerning the topic. In Figure 2 it is shown that after exploring the field some theoretical concepts and themes were identified as the foundation of the first survey. The outcomes of the first survey provided interesting insights that gave reason to change the method and specify theoretical framework. As becomes clear from the model, the final theoretical framework got shaped after the first survey was conducted just as the final research questions. As the outcomes of the first survey have contributed to the content of the final

theoretical framework and the design of the second survey it is necessary to already refer to some empirical findings in the parts describing the theoretical framework and the methods. The detailed analysis of the response on the first survey however will be described in Part 4 in which all the empirical findings will be addressed. In short it can be said that the central research component of the paper is the second survey and that the first survey provided the foundation for the second. In the following parts it is attempted to separate the theoretical framework, the description of the methods and the findings as much as possible but it can’t be prevented that at some points the theory, methods and findings are overlapping.

(13)

12

PART

2

(14)

13 2. THEORETICAL FRAMEWORK

As explained earlier the theoretical framework got shaped throughout the study and the theory that was used to guide the first survey is somewhat different than the theory that was used for the second. Because the second survey is placed in a central position in the study this part will describe the final theoretical framework as it was shaped after the first survey. Because the outcomes of the first survey did make a contribution there will be some references to the first survey, its theoretical funding and its outcomes.

As mentioned in Part 1 by means of a window metaphor, the goal of this paper is twofold. In the first place the goal is to shed a theoretical light on the subject of influencing behavior of a target audience via social media by a military organization in an area of operations. Secondly the goal is to provide some practical tools for the Dutch military on how to influence the behavior of a target audience via social media in an area of operations. Therefore the theoretical part of this study is quite extensive and will start with first describing a general theoretical context of influencing behavior via social media by a military organization. Based on this general overview three

mechanism are described in more detail which provide three specific explanations of how behavior can be influenced through social media communications. The literature study can therefore be seen as a theoretical research in itself as well as a preconditional part of the empirical research.

2.1 INFLUENCING BEHAVIOR THROUGH SOCIAL MEDIA

There are different definitions of social media to be found. A commonly accepted definition is provided by Kaplan and Haenlein (2010) in which they define social media as a group of internet-based applications that build on the ideological and technological foundations of Web 2.0, and that allow the creation and exchange of User Generated Content. For a layperson this definition doesn’t really make clear what social media is and while most people can think of Facebook or LinkedIn, social media is more than that. To get a clearer and concrete image of what social media is, an overview is presented in Figure 3. When considering that the presented overview concerns the most popular social media applications of 2013, its size will be magnified when introducing all worldwide used social media applications. Every continent, country or region can have specific social media platforms. Besides geographically preferred social media, the specific applications that are popular differ over time as well. So when considering influencing behavior through social media it is evident that firstly there should be an analysis of the specific social media reachability of target audiences in the specific area of operations.

(15)

14 Figure 3. Social media landscape 2013 (FredCavazza.net 2013)

As is done in the picture above it is possible to make a more generalized distinction of different social media applications based on their function. A comparable division is made by TNO (2011) in which they describe four possible activities that organizations can deploy concerning social media being: watching, sending, asking and interacting. It is evident that watching is essential in the process of target audience analysis but that by watching itself, it is impossible to influence behavior. By using social media to send, to ask or to interact can however influence the behavior of target audiences. To see social media in this way, it becomes detached from geographical locations, the dimension of time and specific lay-out. To structure the social media landscape in such a way makes the complexity become easier to comprehend when trying to make sense of (using) social media.

Because the main goal of this study is to make the use of social media in an area of operations practical it is necessary to further specify the subject. A useful theoretical insight is provided by Kitzen (2012) through his investigation of how obtaining collaboration with the local population is practiced in the reality of contemporary warfare. A multitude of possibilities to influence the behavior of target audiences in an area of operations can be distracted from his research. These are not linked to a specific medium and must be seen as general approaches of collaboration (voluntarily and forced). He identifies several ways to instigate a behavioral change, varying from providing people with information, increasing their physical mobility and bribing them to blackmailing people, reducing people’s earnings and even killing people as the most drastic way of influencing their behavior. In total he identified over 40 different approaches to realize collaboration that were based on four years of Dutch military employment in the province of Uruzgan in the south

(16)

15 of Afghanistan. The complete overview is presented in Annex 3. Many of the presented

engagements between the military organization and a target audience can be translated to influencing behavior via social media. A possible spectrum for influencing behavior through social media based on Kitzen’s (2012) model is depicted in Figure 4.

Figure 4. Possible military communication activities via social media to influence the behavior of a target audience in an area of operations

The dimensions of the spectrum presented in Figure 4 are a somewhat different than those in Kitzen’s (2012) model. While Kitzen (2012) uses the more general terms of ‘Stimulate behavior’ and ‘Change behavior’, in Figure 4 it is specified a bit further to ‘Stimulate desired behavior’ and ‘Correct undesired behavior’. This further specification points out two ends of a behavioral spectrum

between desired and undesired behavior. The missing link in the middle can be categorized as neutral behavior. Seen from the military perspective there are in general three different types of behavior that can be identified. Behavior of a target audience can be counterproductive to the achievement of military goals, it can neither support nor hinder the achievement of military goals or it can support the military goals. These three different modes of behavior will be referred to as undesired, neutral and desired behavior respectively. This means the categorization of the behavior

(17)

16 of a target audience is related to the goals of the military operation. So by influencing the behavior of target audiences in areas of military operations the military organization wants to synchronize the target audience behavior with the goals of the military operation. The effect that the military

organization strives for is on the one hand that undesired behavior will no longer be counteracting in respect to the specified goals. Undesired behavior must then be corrected to become less

undesirable but preferably become neutral or even contribute to the military objectives. Behavior that is already contributing to the military operation must be stimulated to amplify the positive effects. Last, neutral target audience should then be encouraged to show desired behavior. Because it is impossible to depict clear boundaries between undesired and neutral and between neutral en desired behavior, only the two outer behavioral dimensions are inserted in the model. Another difference is that Kitzen’s (2012) model is mainly applicable to individuals. Because social media is inherently social this aspect needed to be taken into account. As communication relations can vary from one-to-one to one-to-many to many-to-one and many-to-many (Castells 2013B) the size of the target audience was taken as one of the axes. The last dimension of risk was chosen because of the precarious environments that the military organization in general operates in. Also because the use of social media to influence behavior of target audiences is new, a risk assessment provides

important guidance for the future implementers. The previously mentioned variable ‘modus’ that contains the categories sending, asking and interacting is left out of the model. The reason for this is that the size of a target audience, the modus and the specific approach are highly interdependent.

Finally, 17 different approaches or mechanisms can be identified in Figure 4. Even though the spectrum is not exhaustive nor complete it still provides an indication of the range of

possibilities. Its main function here is exactly that. Namely to make clear that the opportunities to influence behavior via social media are multiple. All 17 approaches are based on literature combined with information from expert conversations. Because the spectrum concerns the hypothetical spectrum of influencing behavior via social media within an area of operations, these approaches will need a structured evaluation of applicability before they can be applied in practice. To fully complete an ex ante evaluation of all the proposed mechanisms in Figure 4 is a complex and time consuming exercise. As it is too time consuming to investigate all suggested approaches in detail a selection has to be made based on the decision of what approaches seemed most fruitful to

investigate in greater detail. The three approaches that are expected to be of most practical use will be investigated in detail. This means that it was decided to focus on approaches from the top right quarter of the model as these approaches are aimed at smaller target audiences and have the lowest risk. The approaches that were included in the first survey were partnering, connecting and crowd management as in some of the first expert conversations these were referred to as being effective

(18)

17 based on personal experiences. The outcomes of the first survey, combined with evolving theoretical insights and additional information from expert conversations instigated a subject change so that for the second survey the approach of crowd management was replaced with the approach of

confronting. This means that the continuing theoretical framework will concentrate on partnering, connecting and confronting as the three central approaches of the study. To further specify these approaches the three defined types of behavior of a target audience will be taken as the starting point for the specification of three different mechanisms. The following chapter (2.2 Three

mechanisms) will be focused on describing a mechanism for correcting undesired behavior by using the approach of online confronting, a mechanism that activates neutral behavior through the social media partnering approach and a mechanism that encourages desired behavior by connecting people on the internet. In Figure 5 an overview is given of the general theoretical outline of the three mechanisms on which the further study will concentrate. In the next chapter a detailed theoretical elaboration will be provided on these three mechanisms.

Figure 5. Integration of behavior, approach and effect towards three mechanisms

2.2 THREE MECHANISMS

Since the early work of Elster (1989) in which he describes social mechanisms as the nuts and bolts processes by which cause and effect relationships in the social world come about, the social mechanism approach has become more prominent within the field of social sciences over time. Because a mechanism approach will lead to clear explanations of why and how people might behave in a certain matter within a certain context it will result in concrete conclusions. As the main goal of this study is to provide concrete solutions the focus on mechanisms seems to speak for itself. It will ensure that the outcomes will be useful for military personal involved in non-kinetic influencing activities via social media. While the mechanisms approach is often used to explain causal relations in empirical data in this study social mechanisms are used in another way. Based on literature, expert knowledge and expert experience three hypothetical theoretical mechanisms are

(19)

18 identified and described. When searching for and identifying three potentially useful mechanisms there are a few guiding principles that were taken from a review of mechanism-related literature done by Hedström and Ylikoski (2010). First, a mechanism can only be identified by the kind of effect it produces and secondly, it needs to refer to a causal process. In the third place the mechanism has a structure that explains the how and why of a phenomena that also includes contextual factors that influence the outcomes. In the following four chapters the search and identification of mechanisms is described. These mechanisms will explain the change from undesired behavior to neutral

behavior, from neutral behavior to desired behavior and from desired behavior to sustained desired behavior respectively. The last chapter is devoted to the integration of contextual factors in the model. Based on the findings of the first survey the focus of this chapter is on preconditions as these are an essential aspect of the context in which mechanisms come about.

2.2.1 CONFRONT TO CORRECT UNDESIRED BEHAVIOR

In this section an explanation is provided as to how undesired behavior on social media can be changed into neutral behavior. The mechanism is inspired by experiments of the Dutch National Police with directly approaching misbehaving individuals via social media. This approach was referred to during the expert meetings by a TNO expert and also by a police expert having a high position within the communications branch. For this seemingly effective approach a theoretical starting point can be found in routine activity theory (RAT) (Cohen & Felson 1979). The exhibitional approach (Hogan 2010) and the multiplicity in online and offline selves (Turkle 1995) are introduced to specify RAT to fit the specific context of this section.

The routine activity approach is first introduced by Cohen and Felson (1979) and provides a theoretical model for the explanation of crime. Instead of emphasizing the characteristics of offenders they focus on the circumstances under which crime comes about. They state that most criminal acts require a convergence of likely offenders, suitable targets and the absence of capable

guardians against crime and they empirically show that victim risk rates for rape, robbery, assault

and personal larceny (e.g. theft) with contact in or near the home are only a fraction of those elsewhere. Therefore they state that ‘College enrollment, female labor force participation,

urbanization, suburbanization, vacations and new electronic durables provide various opportunities to escape the confines of the household while they increase the risk of predatory

victimization.’(Cohen & Felson 1979: 605). Even though their theory is situated in the pre- internet era, the link to social media is easy to imagine as internet seems to fit in perfectly in the list as a new niche in which criminal offenses can take place. Yar (2005) explores the extent to which RAT’s concepts can be transposed to crimes committed in a ‘virtual’ environment. The examination finally

(20)

19 concludes that even though differences between ‘virtual’ and ‘terrestrial’ worlds limit the theory’s usefulness on cybercrime the theory’s core concepts can be applied. This is confirmed by Yar (2005) who concludes that the differences are rather interpreted as ones of degree than ones of kind. More specifically Holt and Bossler (2014) refer to different researches in which RAT is being transposed to cyber-violence, mainly concerning harassment, cyberbullying and stalking. They also acknowledge that internet has created new opportunities for crimes and deviant behavior online. This means RAT is indeed transferable to a virtual environment especially when the concepts are adapted to the specific context. To make RAT better fit the proposed mechanism, the exhibitional approach (Hogan 2010) provides additional insights. It is based on Goffman’s ideas of presentation of self but is translated to the online artifacts people exhibit as a form of impression management. When people ‘act’ on social media they perform for an online audience. Because of the crucial role the online

audience plays in influencing the online behavior this aspect will be added to the model. The second

idea that will be incorporated into the RAT-model is the notion of online multiplicity of selves. One of the first to acknowledge a multiplicity in online and offline selves is Turkle (1995). She argues that online and offline selves are not separated but that multiple selves are interwoven. The online environment provides a space where people can experiment with activities they wouldn’t explore offline. For this reason the RAT concept of likely offenders will be redefined as experimental

offenders. In the Figure below the routine activity approach is combined with the exhibitionist

approach and the idea of online multiplicity of selves to provide a specific context to explain why a social media harassment like online threatening or bullying can come about.

(21)

20 This model provides an explanation as to why people are experimenting with undesired behavior on social media like harassment or bullying. The aspect of the model that provides a possible intervention is the perceived absence of a capable guardian. The existence of an online audience, the experimenting with online behavior and the availability of suitable targets are all three aspects that are difficult to effect directly by a military organization. Because the perceived absence of a capable guardian can enhance the experimental behavior online a possible intervention could be placed there. So the first mechanism to explain the change of undesired behavior to neutral behavior can be derived from this model. Online undesired behavior like a social media harassment could be counteracted by emphasizing an online norm through an intervention by a capable (online) guardian. This could be done by confronting the offender with the undesired or even law-breaking behavior. By emphasizing the connection between offline and online norms these two worlds can be connected to bring two previously separated selves (online and offline) closer together. In short this mechanisms can be described as correcting undesired behavior through confronting.

2.2.2 PARTNERING TO ACTIVATE NEUTRAL BEHAVIOR

This section describes a mechanism that can explain how neutral behavior of a target audience can be activated so that it contributes to reach the goals of the military organization. The main idea came to the front during the expert meeting with a marketing expert who pointed out ‘partnering’ as a method that is often used to increase sales. It refers to hiring a person who is believed to be able to influence the behavior of a specific target audience to communicate on behalf of the company. This approach is also applied on social media where a partners are used to communicate positively on a product so that the message with the (indirect) advertisement will reach a specific TA that is receptive for the message of the sender. To uncover a mechanism that can explain the partnering method and make it fit the subject of this study, this chapter will bring together Granovetter’s theory on the strength of weak ties (1973, 1983), with Cialdini & Goldstein’s (2004) views on social influence and French and Raven’s (1959) account of the bases of social power. By combining these ideas, an explanation will be provided on how neutral behavior of a target audience on social media in an area of operations can be changed into desired behavior.

The theoretical base is taken from Granovetter (1973, 1983). He sees groups of people as networks of individuals that are connected through relations that can be either strong (often labelled as a friend) or weak (often labeled as an acquaintance). These networks can furthermore have many appearances. On the one side it can consist of a small geographically isolated community in a remote African area but on the other extreme it can consist of 802 million people who are the daily active users on the social media network of Facebook (the daily average in March 2014 as

(22)

21 presented by Facebook 2014A). In both cases the bigger network can consist of multiple smaller networks like family’s or interest groups respectively. One general conclusion is that people in any network must have something in common that connects them. The connecting aspect can be divers like having shared interests, workplace, education, ideas, personal characteristics, norms or values. Also a Target Audience, by NATO (2012) referred to as a collection of people who have common characteristics and vulnerabilities, can be can be seen as a group of connected people that form a network. When networks are small and are characterized by a shared culture and a strong cohesion it is difficult for new ideas to appear because of the structural limitations of the network. But Granovetter (1983) points out that most networks are not completely closed though. Via the connecting medium of weak ties that serve as a bridge between different networks, new or

innovative ideas may penetrate these small and tightly knit networks. It means that one person from a certain network has a weak connection with a person that is part of another network. There are no other connections between the people from both networks as for these two acquaintances. The weak tie between the two persons then serves as a bridge by which information can be exchanged between the two networks. To try and change behavior the military organization must alter the network structure by establishing a weak tie that can fulfill the bridging function so that new ideas can find their way into the target audience network. This is made visible in Figure 7.

Figure 7. Bridging the Military and the Target Audience Network

In the second context it might seem strange to connect with the ‘green’ person from the target audience instead of connecting with the person that is placed central in the target audience network. To take into account the structural network characteristics Hanneman and Riddle (2005)

(23)

22 have created an online platform that provides clear guidance in the use of Social Network Analysis (SNA). Some structural characteristics of a target audience network that could be taken into account when identifying the most suitable partner for an influence campaign are network density (the proportion of relations in the network from the total of possible relations) centrality (which person is most central) and paths (how many relational steps are connecting two people). Several more structural characteristics of networks and nodes (persons) could be used but it is far beyond the reach of this section to elaborate on all the possibly relevant concepts of SNA. In a further stadium, when the use of social media to influence people is some more integrated within military

operations, the use of SNA could be a next step of understanding network dynamics.

But the adoption of new ideas within a network can’t purely be explained by structural considerations as networks are filled with qualitative content. Therefore it is useful to connect Granovetter’s ideas to a second important theoretical concept of social influence. Cialdini & Goldstein (2004) give an overview of literature on social influence and identify two main themes of compliance and conformity. Compliance is defined as ‘a particular kind of response –acquiescence- to a particular kind of communication –a request.’ (Goldstein 2004: 592) They explain that such a request can be explicit but that it might as well be indirect. So when a person (in this case a partner of the military organization) is communicating on social media to persuade people to act in a specific way this can also be seen as urging a target audience to respond in a desired way. An important question is why people would comply with a request. One driving force behind compliance can be found in authority as ‘Individuals are frequently rewarded for behaving in accordance with the opinions, advice and directives of authority figures.’ (Cialdini & Goldstein 2004: 595). It is important to realize that who is perceived as an authority within a network depends on the content of the network and it’s relations. In a work environment the most senior employee might be seen as an authority within the network of the company while for the 111.298 members of the Facebook group ‘Beliebers’ (Facebook, 2014B), the Canadian pop idol Justin Bieber is perceived as an authority. These two examples can be connected to the third and last theoretical building block of this section which are the bases of social power (French and Raven 1959). These can explain why someone is to be perceived to be an authority. In the previously given examples the authority of the senior

employee is based on his expertise while the second example shows that popularity, admiration and likability can also serve as a base for influence. According to French and Raven (1959) five different bases of social power can be identified. These arereferent power, expert power, reward power, coercive power, and legitimate power. Even though their model has been extended with

informational power and is further refined (Raven 1992) the basic elements are still recognized as valid today. The base of social power that seems most useful when considering using a partner to

(24)

23 persuade a target audience on social media is referent power. It means that a target audience is more likely to comply with the (indirect) request to change behavior when the target audience is admiring, liking or respecting this person or is able to identify with him/her as in the Justin Bieber example. This seems the most useful because some important features of social media are following, liking and connecting based on commonalities.

In Figure 8 it is shown how a message can spread when introducing social power into the networks. The persons that have more social power within the networks are depicted as bigger circles than the smaller and less influential ones.

Figure 8. Introducing social Influence into the network model

The Figure makes clear that when the military organization is creating a bridge to a person that has no authority, in this case referent power, the mechanism of partnering is not likely to have an effect on the behavior of the neutral TA. When the military organization can create a bridging partnership with the person that has the most social power in respect to the target audience, the mechanism will obviously be much more effective. What is also shown in context 4 is that to convince the most influential person from the target audience to cooperate the same logic applies. The more authority the liaison person from the military organization is perceived to have by the preferred partner, the more compliance can be expected. This makes clear that the creation of a bridge can have different outcomes depending on the structural and content characteristics of the network.

From the previously mentioned two main themes of social influence being compliance and conformity (Cialdini & Goldstein 2004) only compliance was worked out so far. The last theoretical aspect that will be addressed is conformity. This is defined as ‘the act of changing one’s behavior to

(25)

24 match the responses of others’ (Cialdini & Goldstein 2004: 606). Where compliance was linked to the individual influence of social power on behavior in the network, conformity emphasizes the

influence on behavior from the social context. To increase the influence on a target audience it is interesting to be aware of the multiple source effect (Lee 2004). It refers to the observation that people attach a greater importance or value to information that reaches them through multiple sources. It makes clear that the value that is accredited to information is not just dependent on the content. It means that when the people reached by a central and influential person in a network, all pass on the message to the people they are connected to, there will be people who receive the same message from different people. The more internal relations, the denser the network, the stronger the multiple source effect. As shown in Figure 9, when all the people directly reached by the influential partner will re-communicate the message, all except for one person in the target

audience will perceive multiple sources of the same message. This is visible in the network as at least two incoming arrows. With every re-communication the message will spread and starts echoing in the network. Another way to have the multiple source effect come about is to cooperate with more than one influential partner within the same target audience network.

Figure 9. Conformity in a network

The presented theoretical framework is graphically presented in Figure 10. It provides an explanation as to how a neutral target audience can be persuaded to behave in a way that contributes to the goals of the military organization. The mechanism that can explain how the behavioral change can be reached is partnering, using a third party authority that is embedded in the TA network. It means that a bridging relation needs to be established between the military

(26)

25 organization and an influential person in respect to the target audience. Because people in general are receptive for communications from authorities and because it is questionable to what extent the military organization or its employees are perceived as an authority by target audiences, seeking for a partnership could potentially be effective.

Figure 10. Identifying Authority in a network

2.2.3 CONNECT TO ENCOURAGE DESIRED BEHAVIOR

The last mechanism that needs to be addressed is one that can explain how desired behavior that already supports the military goals can be further stimulated or encouraged to increase the positive contributions. This way the extent to which desired behavior supports the military organization can be maximized. The main idea for this mechanism was inspired on the uprisings in the Middle-East that started in the spring of 2010. Personal experiences from living in Damascus, the extensive coverage in the media and expert views on the use of social media for mobilization have instigated the search for similar examples and relevant research and theory in literature. This resulted in combining Coleman’s (1988) concept of social capital with DiMaggio’s et al. (2001) ideas on the social implications of the internet. To add specific and more recently published content to the mechanism, connections were made to theory about online social networks in relation to protest behavior, political behavior and connective action.

A concept that is often referred to in the context of connecting people through social media is social capital. This concept, that originated before the embedding of internet in Western world’s daily life, is described in great detail by Coleman (1988). He states that social capital exists in relations among persons. It comes about within the structure of relations between actors and among actors and it is not encapsulated in either in the actors themselves or in physical elements.

(27)

26 Social capital facilitates actions of actors within a structure of relations (network) that wouldn’t have been possible without the dynamics of these specific relations. With an example about South Korean student radical activists he points out to a specific form of social capital in the light of activism. The network of students constitutes a form of social capital in which opposition to the political system is made possible in a country known for intolerance of dissident behavior. He acknowledged that a network of people or organization that makes activities possible that are opposed to the dominant structure is an especially potent form of social capital. Another example that he provides is that ‘a group within which there is extensive trustworthiness and extensive trust is able to accomplish much more than a comparable group without that trustworthiness and trust.’ (Coleman 1988: S101) Because social capital comes about through changes in the relations among persons it seems effective to connect people online in a network that can facilitate civic or political participation. A review of research on the social implications of the internet within five specific domains is given by DiMaggio et al. (2001). One of the domains is community and social capital, linking Coleman’s (1988) ideas with the specific context of the internet. They present the opposing views on the impact of the internet as either resulting in social isolation or in the formation of social capital. They emphasize that from their point of view the internet has no intrinsic effect on community and social capital because it depends on the context and the circumstances. An example that can further clarify their point of view is specifying the context of the different ways people can use the internet. When someone uses the internet only and intensively for online gaming that replaces all the previously existing real-life social interaction it is clear that the effect of this internet use is probably social isolation. On the other hand when the internet is used to get in contact with geographically distant people with similar interest it can be argued that the internet makes possible the formation of new social relations and social capital. DiMaggio at al. (2001) conclude that internet sustains the bonds of community and provides possibilities to create new bonds rather than resolving them. Lastly, concerning politics they state that even though the implications of internet use on social capital is subject of discussion, in general it is acknowledge in literature that web-based interaction does have unique and politically significant properties. Therefore it is useful here to move on to more specific and more recent literature concerning the internet, social media and civic and political engagement.

The use of social media to facilitate protest behavior is investigated by Valenzuela (2013). He investigates how using social media translates into increased political activity as the positive link between the two is shown in several recent studies. He identifies three different explanations as to why frequent social media use is positively related to political participation. The availability of general information or news through social media, the possibilities for expressing political opinions and the possibilities for organizing activism can all be held accountable to motivate political

(28)

27 participation. He found a positive and significant effect of both political opinion expression and social media activism on political participation. He didn’t find an effect however of the availability of general information or news on social media on protest behavior. When information is investigated in a more specific form like being related to public issues, community problems, policy or social movements there does seem to be a relation though. Valenzuela (2013) refers to several different researches in which a relation is established between specific politically tainted information and political participation. In another article written by Zhang et al. (2010) the role of social networking sites in influencing political attitudes and engaging people in the democratic process is examined. They investigated to what extent social media have stimulated people to engage in civic and political participation, both important in building a community. They find that while blogs mostly engage people that are already politically active, social media can stimulate involvement in the political process. In particular regarding young people as they are in general the biggest represented category on social media. Even though social media are mostly used to interact and connect with people who are known off-line as well, the social media groups that are centered around specific topics provide platforms on which people can connect with a diversity of people that are new to them. The connection of people from different networks to bring about social and political change is referred to as bridging social capital. They conclude that to stimulate civic and political participation, people need to be encouraged to connect and engage in interpersonal discussion about politics. Zhang et al. (2010) also acknowledge that the information-oriented use of social media can stimulates civic and political participation.

The last concept that will be addressed in this section is connective action (Bennett & Segerberg 2012). It refers to social mobilization through Digitally Networked Action (DNA). The increasing opportunities provided by digital media have changed the core dynamics of actions-networks especially concerning the political engagement of the younger generations. They conclude that collective action based on exclusive collective identity and strongly tied networks will continue to play a role in the political landscape, but more often influence will be generated through

personalized collective action formations in which digital media are integral organizational parts. Based on this observation the article examines two different logics of organizational dynamics for action formations. On the one hand they recognize the familiar logic of the more formally organized collective action but on the other hand they point to the logic of connective action that is self-organized and based on the sharing of personalized content across media networks. ‘In between the organizationally brokered collective action networks and the more self-organizing … connective action network is the hybrid pattern’. (Bennett & Segerberg 2012: 755) This hybrid form of connective action can be used to encourage civic and political engagement within the context of a

(29)

28 military operation. When the existing organizational structure in an area of operations is a limiting factor concerning social and political change and completely self-organizing networks for connective action don’t emerge the military organization could make use of this model. It means that there is a loose organizational coordination of action in which people are approached as affiliates rather than as members. The military organization provides the technological outlays and shapes the content frames within which the individuals then communicate, offering personal options to engage and express themselves. As the military organization is the administrator of the network it can have some control on the content. For example it is possible to remove contributions that are against the ‘house rules’ of the network or that are not related to the topic. Also by self-contribution to the content the communication within the network can be influenced. A last remark concerning this hybrid from of connective action is that it is possible for self-organizing networks to sprout. Either because the military organization can encourage contributors to self-administer another network or because people will carry on independently based on personal motivations. In Figure 11 the third and last mechanism of the study is presented that can explains why the connection of people on social media can encourage them to participate in political and civil participation.

Fig. 11. Encouraging desired behavior to (re)build a community

(30)

29 As mentioned before, the outcomes of the first survey have contributed to the final theoretical framework. Therefor this chapter starts with referring to the beginning of the study and will explain how the initially identified concepts of trust and truth resulted in a more general focus on

preconditions and how these preconditions then became part of the final theoretical framework. 2.3.1 TRUST AND TRUTH

From the expert meetings that took place to get oriented on the subject of influencing behavior through the use of social media two reoccurring themes could be observed. In several of the

conversations the role of either or both trust and truth was brought up when considering influencing behavior in an area of operations through social media communications. This triggered a theoretical investigation of both concepts and resulted in including both concepts as research subjects in the first survey.

2.3.1.1 TRUST

Based on Cook, Hardin & Levi (2005) the role of trust in networks was theoretically explored. The subject of analysis in the book is cooperative behavior which can be seen as individual and collective effort to cooperate with other individuals, groups or organizations. Because cooperation is part of the three previously described mechanism, the theoretical insights from Cook et al. are useful for this study. When trying to correct the behavior of an individual, seeking a partnership with an influential person to change behavior of an target audience or encouraging people to connect in an online network to get organized, in all three mechanisms cooperation is essential. Cook et al. (2005) make the important note that cooperative behavior comes about within the context of networks and depends furthermore on social order and on technological access to people as well. Social order was already touched upon when introducing the concept of ‘capable guardian’ in the first mechanism concerning correcting undesired behavior and also when referring to the bases of social power in the second mechanism concerning activating neutral behavior. They stress that even though cyber relations might be less strong or intense, people’s networks are covering far more ground through the increase of technological communication opportunities. This means that through technological advancement people’s networks are becoming bigger. Cook et al. (2005) point out a difference in network configurations between small communities and urban settings. They state that in the first context there are more multiplex networks. This means that the same people are connected in different networks and the content of a relation between two people can be multiplex. People are closely connected and ties or relations between two people can serve different purposes or have different characteristics. This means your brother can be your friend, your colleague, your dentists and the entity from which you’ve lent money, all at the same time. In the second context of the urban setting, networks are more specified and people will be embedded in multiple networks.

(31)

30 People then tend to have different specified networks that are only sparsely connected. This means people are embedded in different networks that consist of different people. These networks can consist of friends, colleagues or professionals and the different networks will hardly be overlapping. Based on these network differences the authors argue that trust and trustworthiness might play a very large role in motivating behavior in urban settings but that in small communities, norms play a central role. They state that communal norms might commonly be enforced and that it is not so much trust that makes people cooperate with each other but the backing of a community that is able to sanction those who cross the norms of the community. Therefore cooperation is possible within the whole community even if people don’t personally know each other. In urban areas though cooperation is much more influenced by the extent to which people trust the other to ‘return the favor’. As becomes clear from these examples the importance of trust will be different in different situations. Therefore it will be indeed interesting to establish to what extent trust is seen as a preliminary condition of influencing a target audience via social media. This is especially interesting as the clear the point is made by Cook et al. (2005) that even though trust should be a major topic within the social science, its role or function is overestimated.

2.3.1.2 TRUTH

A guideline for the military organization concerning truth and deception concerning Psyops is provided by NATO (2012). The importance of truth can be linked to three main principles of Psyops that are truthfulness, attribution and credibility. Concerning truthfulness it is stated that operations must be based on true information because the use of false information is counter-productive to the long-term credibility and the ultimate success. Also the attribution of communication to NATO in general, a specific member or a cooperating partner or organization needs to be truthful and clear for a target audience. This means that the military organization is not expected to communicate under a false identity. The success of Psyops is furthermore seen as dependent on its credibility seen from the target audience’s perspective. The use of false information is therefore seen as damaging. Not only when false information is communicated on purpose but also in cases when it was

communicated truthfully. That the use of false information or fake identities can indeed be counter-productive and do damage to the long-term credibility becomes clear from some large scale scams that have been exposed. Some recent examples of scandals in which it was proven that

organizations have been telling lies to mislead the public are the European horsemeat scandal (European Commission 2014), the Libor interest rate scandal (Financial Times 2014) and Wiki-leaks (The Guardian 2014). These events were covered widely in the media and damage was done to credibility of the organizations involved. These scandals were exposed mainly as a result of

(32)

31 mask evidence. Because computer data can easily be copied and because that what is send via the internet or even an organizations’ intranet will be stored in (several) servers where it could be (accidentally) found or even recovered via a back door. Because of the increased likelihood that deception will be revealed and can be ‘proven’ by evidence it is understandable if the importance assigned to telling the truth is growing. The importance of telling the truth on social media is emphasized by Kaplan and Haenlein (2010). They define being honest as one of the five essential points for companies when making use of social media. While commercial companies mainly suffer financial losses the consequences for the military organization are possibly higher because mistakes can even costs lives and so the advice to be honest carries some additional weight for military organizations. This all makes it understandable that the use of deception is strictly regulated in Dutch Army Doctrine (2013). It is stated that the use of deception can only be used in situations to mislead an enemy. But the boundary between a lie and not telling the complete truth is quite vague. Therefore it is interesting to investigate to what extent telling the truth, the whole truth or only part of the truth will play a role when influencing the behavior of target audiences via social media in areas of operations.

2.3.2 FROM TRUST AND TRUTH TO 12 PRECONDITIONS

The details of the methods and the findings will be addressed later in Part 3 and Part 4 but for the sake of clarity it is necessary to already refer to some aspects here. Because of the previously described dynamics concerning the concepts of trust and truth it was decided to include these concepts in the first survey to investigate the role they would play when considering influencing the behavior of a target audience by a military organization via social media in an area of operations. Among others, one statement concerning the importance of trust and one statement concerning the importance of never using falsehood were reviewed by a panel on a Likert-scale. The response on both statements didn’t provide a consistent outcome on the perceived importance of both trust and truth. In the argumentation that the panel was asked to provide for every statement several

additional aspects or other preconditions that were perceived as more important were suggested. Besides this it was mentioned numerous times that the importance of a precondition was related to the specific context. So it was emphasized that trust might be an important precondition in one context but that in another context trust could be less important or even irrelevant. In total 10 additional preconditions were suggested that could be perceived as important when influencing the behavior of target audiences via social media in an area of military operations. This resulted in a list that contained 12 potentially important preconditions depending on the context. This resulted in the decision to shift the focus for the remaining study to the relation between preconditions and

(33)

32 2.4 THEORETICAL OVERVIEW AND RESEARCH QUESTIONS

2.4.1 THEORETICAL OVERVIEW

From Pawson and Tilley (1997) the guiding principle is derived that the explanatory strength of a mechanism is dependent of the specification of the context in which the mechanism is to be relevant. Also DiMaggio et al. (2001) state that research should focus on the circumstances under which different effects might be produced. Therefore the three previously identified theoretical mechanisms from Chapter 2.3 will need to be translated into a concrete context. These three

contexts can then in turn be linked to the preconditions that were identified in Chapter 2.4 by means of the second survey. In Part 3 that covers the methods more details will be presented. The goal of the research would then become to determine what preconditions would play an important role to assure the mechanisms would come about as described in specific contexts. In Figure 12 the

theoretical framework for the central (second) survey is summarized. In this model it becomes clear that the practical investigation of the survey is aimed at identifying the most important

preconditions in general and more specific for every context that contains a mechanism.

Figure 12. Overview of the Theoretical Framework

2.4.2 RESEARCH QUESTIONS

In the introduction of Part 2 the two-fold goal of this study was mentioned. The first goal was to shed a theoretical light on the subject of influencing behavior of a target audience via social media by a military organization in an area of operations. By translating Kitzen’s (2012) model into the spectrum depicted in Figure 4 on page 13 a broad indication of the possibilities for a military

Military influence via social media : the challenge of changing behavior without using force

July 3th 2014

Master’s thesis in Sociology

Track: Social Problems and Social Policy

Student: G.M.G. Nielen 10295518

Email: gwenda3dec@hotmail.com

Supervisors:

1

Drs. Thomas G. Kampen

2

Dr. Barbara da Roit

U

NIVERSITEIT

VAN

A

MSTERDAM

MILITARY INFLUENCE VIA SOCIAL MEDIA

TABLE OF CONTENTS

PART

1

PART

2

_{Drs. Thomas G. Kampen}

_{Dr. Barbara da Roit}