Anarchy is What Students Make of It: Playing Out Wendt’s Three Cultures of Anarchy

Abstract This article explores the hidden educational potential in the board game Diplomacy. While commonly recognized as a good low-cost negotiation simulation and a useful teaching platform, the original game version over-emphasizes the conflictual nature of international relations and presents an image of international relations that is not reflective of the state of world politics today. The article reports experiences from developing and teaching a modified version of the game. It suggests a concrete way to integrate International Relations (IR) theories directly into the game. By making three common theoretical approaches an integrated part of the game, as well as expanding game strategies, win conditions and associated payoff tables, Diplomacy can be turned from a tool for teaching one variant of Realism into a flexible platform for teaching a broader range of IR theories, including Liberal-Institutionalism and Constructivism. In this way, the gap between abstract theory and simulated practice can be bridged in a way that students much appreciate as a learning mode. The article thus suggests a partial solution to one of the key challenges of using games and simulations in political science teaching: how to connect the game or simulation with theory.


Introduction
In academic disciplines that are closely related to a particular profession, such as medicine and doctors, legal studies and lawyers, or pedagogy and teachers, theory and practice tend to be closely intertwined. Academic teaching in these disciplines also, unsurprisingly, makes much use of simulations, where theoretical knowledge of medicine, law or pedagogy can be applied. While the academic discipline of political science has some relation to the politician, and the study of International Relations (IR) bears a connection to the practice of diplomacy, the connection between theory and practice is more tenuous. Students interested in learning how to actually "apply" their teaching in practice, may find that there is not much offered in their department's teaching program, beyond Model United Nations or European Union negotiation simulations, although there are notable innovative exceptions, such as the wargaming network at King's College London, or the Zombie simulation in Roskilde University (Horn et al. 2015).
Foreign policy decision-making is inherently complex, as it typically involves processing much information in a short span of time and under time-pressure (Hudson 2014). One strength of good simulations is that they approximate the complexity of real-world decisionmaking (Loggins 2009). However, one of the central challenges of using simulations and games in political science teaching, is how to connect the game/simulation with theory. In typical IR teaching simulations, such as Model UN or EU negotiations, there tends to be a disconnect between abstract theory and the simulation, as simulations commonly are focused on understanding the basics of how the UN works, the negotiation process, country positions etc., rather than the theories and meta-theories that fill IR journals and the heads of academic IR experts (Kirschner 2020;Loggins 2009;Horn et al. 2015).
Earlier studies have shown improved learning results associated with simulation games using experiments or quasi-experiments (Galatas 2006;Stover 2007;Lee and Shirkey 2017). Asal et al. (2019) have also recently demonstrated that short simulations can effectively be used to illustrate specific theoretical concepts in IR theories, e.g. levels-of-analysis. Scholars have also applied simulations to illustrate one theory with each simulation, e.g. Neo-Liberal Institutionalism with Prisoner's Dilemma and Neo-Realism with Diplomacy (Asal 2005), or critical theory with Monopoly (Paino and Chin 2011), and to inspire students to think theoretically by playing the game outside of class (Rittinger 2020). However, teaching a broader range of IR theories through one simulation in class is still uncommon (see Horn et al. 2015 for an exception). I propose a way to integrate high-level theory directly into the simulation.
This article builds on a previous article (Mattlin 2018) that reported on a teaching concept in which I adapted the board game Diplomacy to make it pedagogically more useful for IR teaching purposes. The earlier article introduced the basic teaching concept, as well as teaching experiences and student feedback from three years of teaching the course at the BA-level in the local language in the University of Turku . In this article, I take the process one step further by directly integrating IR theories into the game and adopting three theory-derived ways to play and win the game. This modified version of the course concept was taught in English in the University of Helsinki (2018) and in the University of Turku (2019-20). Transferability of the course concept was tested by teaching the course in the last three years under different circumstances: to students with different backgrounds, including international students and diverse majors, also at the MAlevel, and in a different teaching language. Course evaluations and feedback indicate that the further adaptation of the course concept was similarly well received by students. 1 The article argues that the adapted version of the game allows students to play Diplomacy in subversive ways that seamlessly assimilates the game into class teaching, and better integrates "high-level" theory into the simulation, thus partially overcoming the challenge of connecting the game (simulation) with abstract theory. The article thus makes a unique contribution to the burgeoning literature on using simulations, such as Diplomacy, also for theory teaching in political science.
Exploring the theoretical potential of Diplomacy Diplomacy (Avalon Hill/Hasbro) is a board game developed by Allan B. Calhamer in the late 1950s. The game has been used in the teaching of international relations, diplomacy and dynamic decision-making (Asal 2005;Gattie 2004;Mattlin 2018;Rittinger 2020). Diplomacy is a relatively close approximation of international relations' anarchical world. However, in the game's original version, the logic resembles a Hobbesian world, or in terms of current IR theories, Offensive Realism (Mearsheimer 2001). Offensive Realists, like Mearsheimer, argue that fundamentally international relations have not changed from what they have been since Thucydides. Under conditions of anarchy, as understood by Offensive Realists, states can rely but on themselves, promises made can never be fully trusted and security considerations are primary. Great powers are ceaselessly jostling to gain power over each other, i.e., they are more concerned with relative than absolute gains, and seek to grow their relative power when possible. In other words, Mearsheimer claims that Great Powers are revisionists (Mattlin 2018). As a consequence, realpolitik, i.e., power political behavior, is a prudent course of action (Johnston 1995).
The original game version underlines international relations' conflictual nature. Players strive to gain an advantage by occupying supply centers on the board through movement of military units (armies and navies), engaging in negotiations and signaling (both verbal and non-verbal). The system is anarchic, as there is no higher authority that is able to enforce deals that players negotiate among themselves. Lies are made and promises broken and the threat of war is hanging over players like Damocles' sword. Weak actors will be eliminated, while stronger players engage in powerbalancing. Playing the original Diplomacy tends to show students how quickly they become engulfed by the self-help world's logic, which can be educational in itself (see Mattlin 2018).
One of the educational benefits of the game is that students are able to reflect on the rules of the game, the world as described by IR theories, and their own worldviews. Unsurprisingly, class debriefings tend to lead to debates, sometimes even heated, about Realism (Asal 2005). However, this focus on realpolitik and Realism can also be a limitation, especially if the educational aim is to teach also other theories than (Offensive) Realism. Recently, there has been an effort by scholars to broaden the theoretical uses of Diplomacy in IR teaching (Bridge and Radford 2014;Mattlin 2018;Rittinger 2020). Some scholars have noted that the game can be useful also for what it leaves out (Bridge and Radford 2014;Rittinger 2020). Rittinger (2020) also sees the game as a way of modeling world politics, rather than replicating it. He suggests allowing students greater freedom to challenge the objectives and rules of the game, rather than just reflect on how the game logic links up with existing IR theories.
This article builds on the research insight that participation in simulated international negotiations makes actors adapt to the rules and norms of the interaction environment. Therefore, changing the game rules can also be expected to produce a game outcome that follows a different theoretical logic (Niv-Solomon et al. 2011;Tamai et al. 2016). As constructivist Alexander Wendt persuasively has shown, there is not just one timeless and universal logic of anarchy in world politics. The logic and rules of anarchy in world politics depend on the interactions between states and other actors, and the intersubjective understandings thus created. Or, as Wendt famously noted, "anarchy is what states makes of it" (Wendt 1992). Why then, we may ask, should the Diplomacy game force students to play out just one logic, especially if this logic is no longer the dominant logic in international relations? 2 Why not allow students to "make their own anarchy"?
In the further developed version of the game that this article reports on, students get to "play out" theory in practice, by choosing between three game strategies that approximate the three basic cultures of anarchy: the Hobbesian, Lockean and Kantian, as identified by Alexander Wendt (Wendt 1999(Wendt [2010; cf. Wight 1991;Bull 1977Bull [1995). Instead of regarding the international system's structure only in materialist terms (Mearsheimer 2001;Waltz 1979Waltz [2010), Wendt argues that social structures form the system's culture, which can take several forms and be internalized by system participants to various degrees (based on force, self-interest or legitimacy). In the Hobbesian culture, interaction is based on enmity, in the Lockean culture it is based on rivalry, while in the Kantian culture interaction is based on friendship (Wendt 1999(Wendt [2010). Among contemporary IR theories, the Hobbesian culture resembles Offensive Realism, while the Lockean culture underpins Liberal-Institutional IR theories. The Kantian culture can be seen in the genealogy of theories on security communities and political integration (see e.g. Deutsch et al. 1957). By this teaching concept, high-level theory comes alive and can conceivably be internalized to a deeper extent than by just reading about the theories in text-books, i.e., through active learning.

The teaching concept
The Diplomacy board game has several features that make it suitable as a negotiation simulation in teaching. The basic rules of Diplomacy are abstract, strategic and simple, and contain no chance elements, beyond the draw of the played country. Despite this, the game brings home the complexity of foreign policy decision-making. The game's negotiations, social interaction, psychological and game-theoretical aspects make it useful for teaching IR and foreign policy from the perspective of an actor (decision-maker). In Diplomacy, game orders are simultaneously resolved. This simulates actual real-time behavior and turns the game into a good simulation of dynamic decision-making (Gattie 2004). However, there are also some limitations that can be improved on. To address these limitations, I introduced novel game elements that include team play, employing dedicated peace mediators, and debriefing discussions. The most important change concerns altering win rules that enables game outcomes that are mediated and negotiated (see Mattlin 2018).
The game is fully integrated with the other teaching (lectures and game sessions alternate), and played by students during class hours, always following a regular lecture that incorporates small group discussions in fixed groups (to lower the threshold for participation). Each game session is in turn followed by an online discussion (in a learning environment) related to the theoretical topics covered in class that week. Typically, the game has ended after four two-hour game sessions.
The game is played in pairs of two but can also be played individually or in threeperson teams. In addition, there are 1-2 peace mediators. One practical limitation is that, as there are seven countries on the game board, preferably student numbers need to match this (plus the peace mediators), i.e., students playing the countries should ideally be 7, 14, 21 etc. The course is designed primarily for two-person teams, i.e.
14-16 students. The main benefit of team-play is that it introduces a "domestic politics" aspect into negotiations. Practically, it also enables more negotiations, interactions and observations during the game, as well as emotional support from team members. Roles are designated by the teacher semi-randomly, e.g. based on the participant name list. The teacher has a very active role throughout the course, as lecturer, game master and "mentor"-always ready to respond to students' questions.
In order to embed IR theory into the game, I expanded the ways that the game can be played to allow for three different ways to end the game: Domination, Rank Order and creating a Zone of Peace. The three ways to play the game are designed to roughly correspond to, and highlight, three prominent IR approaches: Offensive Realism, Liberal-Institutionalism and Constructivism. Granted, this does not nearly exhaust the IR theory universe, but provides students insight into three theoretical approaches that differ in their basic assumptions. Players are free to choose any strategy aimed at achieving one of these goals, and they may also alter course mid-game. Players do not need to disclose to other players which strategy they are following. To facilitate this choice, students will read the game rules, as well as the section of Wendt's seminal book Social Theory of International Politics that discusses the three cultures of anarchy, before the game begins. Students generally also read this and other compulsory readings. I see several reasons for this. Firstly, the importance of reading is emphasized throughout the course. Secondly, course syllabi are not excessively long (quality over quantity). Finally, and most importantly, each week's literature is actively discussed in class (together and in fixed small groups), as well as in the online discussions and learning diaries. Students quickly discover that if they don't read, it will be difficult to follow the course (and they will not look very good).
Game rules and cultures of anarchy are also discussed in class before the game. Differential game "payoffs" in terms of partial course grades (25% of the final grade) were also introduced (Table 1). Game performance will thus have some, albeit limited, impact on students' final grade.

Winning by domination
The first win condition is essentially the original way to play Diplomacy, where victory goes to the player that reaches at least 50% of the available centers on the game map. The game strategy approximates the Hobbesian logic of anarchy, where players do not respect each other's sovereignty or even right to exist, but instead seek every opportunity to take advantage of each other, if need be by eliminating weaker opponents from the game. In terms of current IR theories, the Domination strategy is similar to Offensive Realism. If the game ends in this outcome, only the dominating player/team wins (the top grade 5 for the game part of the evaluation). Other players still left in the game get a 3, while eliminated players and the peace negotiator(s) lose and only get a 1 (the lowest pass grade).

Winning by a mediated and negotiated rank order
The second way to win the game is by achieving a negotiated and signed agreement on a final rank order among teams still remaining in the game, i.e., establishing a pecking order with corresponding benefits. Peace mediators are essential to this outcome, as their role is to help players negotiate and structure an acceptable agreement. The rank order also needs to be reflected on the game board, i.e., the agreement cannot be in conflict with the actual situation on the board, which may require orderly withdrawals. The game strategy follows the Lockean logic of anarchy, where interests conflict and players are rivals, but recognize each other's right to exist, as well as to some possessions. This strategy then approximates Liberal-Institutionalism. Players engage in giveand-take negotiations based on the strength of their negotiating position. Teams share unequally in the victory: the winning player/team gets the most (grade 5), the 2nd ranked player/team gets a grade 4, the third a grade 3, other players still left in the game a 2, while eliminated players only get a 1. Achieving this outcome is also a win for the peace negotiators (5).

Winning by erecting and maintaining a Zone of Peace (ZoP)
The final way to play and win the game is by establishing a Zone of Peace that covers > 50% of game board supply centers. The further conditions for a Zone of Peace (cf. Kacowicz 1998, 10) are: 1. That a Zone of Peace has been declared in public diplomacy (in the closed online course area) by a minimum of two and a maximum of four teams, and all participating countries have declared their participation in it. 2. That there has been a full game year of no attacks by teams participating in the zone (no moves against the units or centers of another country). 3. That players in the ZoP defend jointly against external attack. 3 The ZoP game strategy is designed to broadly follow the Kantian logic of anarchy, whereby players can establish amicable relations and expectations of peaceful interactions over a large area. This also roughly approximates the notion of a "pluralistic security community" (Deutsch et al. 1957;Wendt 1999Wendt [2010, 302-307) or "nascent security community" (Adler and Barnett 1998). In fact, the ZoP resolution of the game is quite close to how Kacowicz (1998) defined a "zone of stable peace": that territorial changes are removed from the agenda (except by mutual agreement or peaceful means); that there is a minimum of nonmilitary intervention in others internal affairs; and that countries of the region sustain an economic rather than romantic/heroic attitude toward their nation-state. These can be regarded as preconditions for a pluralistic security community, that may or may not develop. Kacowicz (1998, 9-10) differentiates between three gradations of zones of peace: negative or precarious peace (mere absence of war), a zone of stable peace (no expectations of violence), and a pluralistic security community.
If a Zone of Peace is successfully maintained, players participating in the ZoP get a grade 4. Players not participating, but still in the game, get a 3, while eliminated players only get a 1. The peace negotiators' grade depends on how actively they were involved in negotiations to erect the ZoP. Whether this accurately reflects a Kantian logic of anarchy was critically noted by one student when reflecting on how their game ended. In that game, the ZoP was attacked by a team left outside of it, that therefore tried to break the zone. The seeming inevitability of this attack on the ZoP in the last game year (following from the second and third conditions above), was seen as problematic by the student. Admittedly, it is not ideal if the declaration of a ZoP leads to an incentive for other players to attack. If we strive to convey Constructivist notions (shared identities), then ZoP should not be so attractive that players are tempted to opt for it out of purely self-interested motivations, although Wendt himself reminds us that weakly internalized pluralistic security communities may also emerge from self-interest, or even coercion (Wendt 1999(Wendt [2010, 302-304).
The correct balance needs to be struck between making the ZoP a viable option, but not too tempting and easy to achieve. The critical student suggested that perhaps lowering the grades for achieving a ZoP (from 4 to 3) and staying in the game, but not being party to the ZoP (from 3 to 2), would make it more realistic. This would also reduce the calculative aspect of ZoP participants, as a ZoP would not be just an easy route to a good partial grade. Table 2 recaps the win conditions and related theories.
Of the win rule modifications, the trickiest part has been to devise rules that are able to convey the constructivist position of shared identities and internalized norms as the basis for a security community. The original Deutschean understanding of a pluralistic security community built on communication (Adler and Barnett 1998, 6-7), is easier to convey. Ideally, students should choose the ZoP because they do not wish to fight or betray people with whom they are friends. Interestingly, the first two times that a game ended in a Zone of Peace, the ZoP formed around some preexisting ties between the players. The first time the nucleus of the ZoP was formed around two Chinese students, who then managed to convince a third (Chinese-language studying non-Chinese student) to join them. The second time the ZoP was formed by a peace mediator around outside-of-class friendships and others players-deemed trustworthy-were then asked to join.
Many students taking part in this course afterwards reported that they either did not adopt a very elaborate game plan beforehand or, if they had a plan, they quickly had to drop it and opportunistically go with the game flow. Helmuth von Moltke-the 19th century Prussian Field Marshal-famously remarked that "No plan survives contact with the enemy." Nevertheless, to further improve the course concept and enhance the adoption of theoretical material, it might be necessary to more actively instruct the students to make the theoretical connections, especially in teaching contexts where the theoretical material is almost entirely new to the students.

Conclusions
This article has argued that the Diplomacy game has hidden educational potential that is not fully realized in the original version of the game, or in courses that separate the game from the rest of the teaching (see e.g. Rittinger 2020). The course concept was built around modifications introduced to the game's win conditions, and closer integration of theory and practice. From the point of view of the teacher, one of the most striking outcomes of the introduced game modifications is how they have managed to transform the logic of the game from the endless backstabbing, treachery and sauve qui peut -mentality that Diplomacy is (in)famous for, into a more peaceful and amicable exercise. Previously, when a negotiated peace treaty was introduced as a possibility for ending the game, students have tended to opt for it (see Mattlin 2018). With the three ways to play and win (Domination, Rank Order, Zone of Peace) there have been fewer "backstabs" (betrayals). Each time the game ended in a Zone of Peace, and all the players seemed content with the result, even those who were not participants in the ZoP, although it meant that nobody got the highest grade for the game part. This outcome was particularly interesting the first time, given that the participating students then did not know each other beforehand and came from diverse backgrounds (different majors and nationalities) Six years of written and oral feedback on the course indicate that students have highly appreciated the teaching concept. An overwhelming majority of students have considered the course concept better than more traditional forms of learning, and numerous students have spontaneously indicated that it is the best course that they have ever attended (Mattlin 2018). The positive student feedback has been remarkably consistent over the years, despite tweaks to the course set-up and different teaching contexts. The comment below is fairly typical: "To conclude, this course has been simply amazing. The various assignments have been really helpful with their variance enriching my learning process. Amazingly there hasn't been any moments of boredom because there were so many different methods of learning used. And the ability to play a great board game with such high stakes was an opportunity that sadly won't probably happen again. This has definitely been a high point of my academic path thus far!" -Excerpt from the learning diary of a student 2019 The game provides students with a unique chance to step into the shoes of actual diplomats, with somewhat corresponding elements present as in actual real-world diplomatic negotiations, such as time-pressures and uncertainty. The course has received consistently high average points in numerical evaluations by students, who have taken the course. On a 1-5 scale, with 5 representing the highest score and four different areas evaluated, as well as an overall grade provided, the six-year unweighted gradepoint average is 4.66 (annual range 4.46-4.83), based on feedback forms filled by 77 students (response rate: 92%), and 385 individual grades given. The 2018 course was taught in a different setting, yet received very similar student feedback and appreciation by students. This adds confidence to my assessment that the teaching concept is transferable also to different teaching contexts.
One of the key challenges of using games and simulations in political science teaching is how to connect the game and theory, especially when teaching abstract theory. In this regard, the teaching concept discussed in this article shows promise. Based on learning diaries and teaching experience, students are able to better make theory/practice connections, as they have continuous opportunities to actively apply their theoretical insights into the game situations, and to reflect on their game experiences in light of theories that are read and discussed throughout the course. Students are also able to reflect on the game's rules and underlying assumptions, as well as their relations to theory and the real world (cf. Asal 2005). Finally, students often spontaneously draw creative links to other theories (not covered during class) in their learning diaries. To conclude, this article has suggested a partial solution to one of the key challenges of using games and simulations in teaching: how to connect the game or simulation with theory. Notes 1. In the country where the study was conducted, this type of scholarship does not require IRB approval. The study followed the national and institutional guidelines on research integrity. 2. Wendt himself argues that, while the Hobbesian culture of anarchy has been predominant through long periods of human history, since the Westphalian peace in 1648, world politics has moved towards a Lockean culture of live-and-let-live. Wendt also envisions the possibility that we may move towards a more Kantian culture in the future (Wendt 1999(Wendt [2010). 3. In theory, a security community does not require collective security arrangements. However, in practice they tend to coincide and reinforce each other (Wendt 1999(Wendt [2010: 298-302).
Removing this condition would in the game context invite more aggression and put the "community" in question.