Abstract
If all states want to survive, why do some of them enter unpropitious alliances? International Relations (IR) theory’s conventional answer is that imperfect information and systemic complexity result in miscalculation. This explanation begs the question: any alliance that fails is a miscalculated one, so the puzzle is not whether but why such mistakes are made. This article imports from recent scholarship on network theory and interpersonal trust to offer an alternative explanation. Alliances are not entities ethereally formed out of strategic imperatives, but products of interactions within transnational social networks of political, military, and business elites in the prospective allies. Such interactions enable alliances because people who are connected to each other through mutual association or previous exchanges develop mutual trust and gain subjective certainty about each other’s intentions and capabilities, which points at a previously ignored mechanism in alliance behavior: brokerage. In a case study that combines theory-based archival research and social network analysis, this article uses historical evidence on the Turco-German alliance to empirically demonstrate the brokerage role Colmar von der Goltz, the head of the German military mission to the Ottoman Empire, played in the two countries’ relations at the turn of the century and their eventual alliance in the First World War. The analysis points at a potential means of bridging IR, history, and sociology while expanding our understanding of alliance behavior and providing policy-relevant insights on geo-economic competition and the weaponization of interdependence at a time of growing strategic rivalry on the world stage.
Building bridges between International Relations (IR) and history is an endeavor that attracted much attention from both sides of the gap (Elman and Elman, 2001). One challenge associated with this task is to redress the Eurocentric biases that excluded from the discipline’s scope many cases and issues from outside Europe (Hobson, 2012; Kayaoglu, 2010). Even though the First World War is IR’s “most analyzed and contested case” (Copeland, 2001: 56) and an integral part of its history as an academic discipline (Porter, 1972), the Ottomans, whose decline was a major cause of the war (Anievas, 2013: 734–735), were virtually invisible to IR scholars until quite recently (Bulutgil, 2017; Kadercan, 2014; Nisancioglu, 2014; Savage, 2011; Zarakol, 2010). In International Organization, they appear only in two articles, with no direct relevance (Narang and Nelson, 2009; Tetreault, 1991). International Security, another prominent journal, has only two articles that directly deals with the Ottoman Empire in its publication history (Bulutgil, 2017; Kadercan, 2014).
Such neglected cases as the Ottoman Empire abound with empirical puzzles. This article concerns itself with one such example. Both realists and constructivists agree that all states want to survive (Mearsheimer, 1994: 10; Wendt, 1995: 72). Then, how do unpropitious alliances like the one between the Ottoman Empire and Wilhelmine Germany during the First World War come to be? The dominant strand in IR’s alliance theories is neorealism, which started to find purchase among Ottoman historians as well (McMeekin, 2011; Reynolds, 2011). Neorealists argue that “states form alliances to protect themselves [from] the threats they perceive” (Walt, 1987: x). These threat perceptions are shaped by geographic proximity, offensive capabilities, aggressive intentions, and the distribution of power (Walt, 1987: 22). The last factor is viewed as especially important: systems with more than two powers—multipolarity—are expected to be unstable because “uncertainties about the comparative capabilities of states multiply as numbers grow and estimates of the cohesiveness and strength of coalitions are hard to make” (Waltz, 2000: 6). Thus, interstate conflict has a higher likelihood because states either rush into war—chain-ganging—or avoid it until it is too late—buck-passing— as happened in the two world wars (Waltz, 1979: 165–169).
Jervis’s (1978) concept of security dilemma, Waltz’s (1979) chain-ganging hypothesis, and the subsequent scholarship building on them (Christensen and Snyder, 1990, 2011; Lieber, 2007; Snyder, 1984; Tierney, 2011; Van Evera, 1984) offer a realist explanation of the origins of the First World War, but this corpus does not suffice to explain specific decisions such as why the Ottomans sided with Germany as opposed to another power like France, which was their main creditor (Geyikdagi, 2011: 51), largest investor (Geyikdagi, 2011: 57), and a major arms supplier (Grant, 2002: 32–33). Indeed, archival documents reveal a dissonance between IR’s theories and history’s accounts of alliance behavior. The Turco-German alliance was the product of frantic negotiations, not a predetermined outcome of the strategic environment. Neither were the Ottomans exclusively interested in an alliance with Germany nor was Germany keen to have the Ottomans on its side (Aksakal, 2008: 93–102). Istanbul was rebuffed the first time it sought an alliance with Berlin in July 1914 (Weber, 1970: 62–63) and it actively courted Britain, France, and Russia until Germany ultimately agreed to an alliance on 2 August (Ahmed, 1984: 14–15). Even thereafter, neither side had much appetite to fight alongside the other. “For three long months after signing the German alliance,” writes Aksakal (2011), “[the Ottomans] did everything they could to stay out of the fighting” (198). German officials similarly believed that the Ottoman Empire was “militarily too weak to be of any value” (Aksakal, 2008: 123) and viewed its neutrality as of greater value than its involvement in the war (Aksakal, 2008: 169).
So, how did their unpropitious alliance come to be? The conventional answer is that one or both of the allies made a miscalculation. Decision-makers operate in a world of imperfect information. They cannot have certainty of such factors as the capabilities and intentions of their adversaries (Yarhi-Milo, 2014) or how much time they have before the window closes for a certain course of action (Edelstein, 2017). Moreover, some features of the strategic environment are inherently unknowable: international politics is a complex system with interconnected elements and emergent properties; its processes are nonlinear, its outcomes cannot be understood by adding together the units or their relations, and many of the results of actions are unintended (Jervis, 1997: 6). Therefore, decision-makers require both the ability to accurately estimate and mitigate risks and the agility to improvise and innovate in the face of uncertainty, and their success varies to the extent these capacities differ (Katzenstein and Seybert, 2018). While all of this is true, however, it is not a sufficient answer to the puzzle at hand. Granted that all states want to survive, and neither Wilhelmine Germany nor the Ottoman Empire did, their alliance was obviously miscalculated. As Greek premier Eleftherios Venizelos reportedly said about Greece’s failed invasion of Asia Minor in 1919, “every enterprise that does not succeed is a mistake” (Carr, 1939: 67).
The question is not whether but why the mistake was made, and answering it requires a bridge over the chasm between the realist literature on structural power and the constructivist literature on perception and agency (Zala, 2017). Miscalculation happens because the world decision-makers see and the world in which their decisions will be carried out are different milieus, and this difference is mediated by the decision-makers’ goals, calculations, and perceptions (Jervis, 1976: 13). Hence, the external setting alone cannot explain or predict behavior. Constructivists capture this point by arguing that the environment in which states act is social as well as material and that this setting affects how states conceive their interests (Checkel, 1998: 325), but alliances have not drawn much attention from them (Masala, 2016: 385) except for a few canonical works (e.g. Barnett, 1996; Risse-Kappen, 1996).
This article is a variation on the constructivist theme that “states are embedded in dense networks of transnational and international social relations that shape their perceptions of the world and their role in that world” (Finnemore, 1996: 2). Its guiding intuition was perfectly summarized by Marschall von Bieberstein, the German ambassador to the Ottoman Empire from 1897 to 1912, in a letter to the Chancellor, Prince von Bülow: “economic interests are so much in the forefront of international life that any attempt to separate them from the political questions would be necessarily vain” (Bieberstein to Bülow, 29.12.1907). In the academic literature, Hirschman (1945 [1980]) had captured this logic by showing how international economic relations affect domestic politics, which in turn shapes national interests, and asserting that “commercial and financial relations [are] the root cause of the influence or power position which a country acquires in other countries” (16). While Hirschman studied how firms and regions exert pressure on central government in favor of the state to which they owe their interest, this article imports from the growing literatures on network dynamics (Cranmer et al., 2012a, b; Kinne, 2013; Maoz et al., 2007; Warren, 2010) and interpersonal trust (Holmes, 2018; Trager, 2017; Wheeler, 2018) to explore a similar dynamic in transnational elite networks. Alliances are not entities ethereally formed out of strategic imperatives, but products of interactions within transnational social networks of political, military, and business elites in the prospective allies. These interactions enable alliances because they build mutual trust, which gives the prospective allies subjective certainty about each other’s intentions and capabilities (Wheeler, 2018: 12). A key mechanism in this process is brokerage (Stovel and Shaw, 2012). Brokers are actors whose position within the network gives them significant information and control advantages due to their role in the flow of goods, information, opportunities, or knowledge and allows them to bridge structural holes by connecting nodes that would otherwise remain unconnected (Burt, 1992). Brokerage also reflects the co-constitutivity of agent and structure through its “switching effects”: brokers make and break ties among actors, which alters the network structure, and, in turn, changes the opportunities for and constraints on individual action (Goddard, 2009: 250).
This article is an example of theory-based archival research (Larson, 2001). It pieces together a social network of officers, diplomats, and capitalists using Ottoman archival documents and applies social network analysis (SNA) methods to demonstrate the brokerage role of Colmar von der Goltz, the head of the German military mission to the Ottoman Empire, who enabled the Turco-German alliance via the relations of similarity, connectedness, and exchange he formed among Turkish and German elites. First, Goltz was instrumental in advancing Germany’s interests in the Ottoman Empire, especially as a de facto lobbyist for German armament firms like Krupp and Mauser (Yorulmaz, 2014). Second, thanks to his service as an officer of the Imperial Household (Mabeyn-i Humayun) and personal connections to senior German officials like Alfred von Kiderlen-Wächter of the German Foreign Office (Auswärtiges Amt), Goltz was a key conduit in the flow of information between Berlin and Istanbul. Third, Goltz played a crucial role in creating ideational homophily among German and Turkish military elites, which formed ties that persisted over time. As the architect of modern military education in the Ottoman Empire, Goltz was a formative influence on generations of officers who studied under his curriculum, trained in his methods, and looked up to him (Nezir-Akmese, 2005: 24–33). These officers, known as the “Goltz generation,” later became enablers of German influence as they rose up to the senior ranks, tilting crucial decisions in Berlin’s favor out of ideological sympathy or personal profit.
The remainder of the article is organized as follows. First, I explore how network theory enriches the study of alliances by enabling the study of social capital and elite brokerage as factors that affect alliance behavior. Second, I sketch out a transnational social network between Turkish and German elites and present quantitative empirics that demonstrate the brokerage role played by Goltz (and his star cadet, Mahmut Sevket Pasha) in the Turco-German alliance. Third, I provide a detailed qualitative study of this network’s actors, processes, and interactions in the period from the Congress of Berlin in 1881 to the start of the First World War in 1914, demonstrating its resilience and influence. In the penultimate section, I discuss this article’s contributions to the study of alliances, its relevance to current policy questions, and possible directions for future scholarship.
Constructing alliances: social capital and elite brokerage
Analyzing alliance behavior from a social network perspective that gives primacy to actors and their relations as opposed to states and their actions requires two theoretical maneuvers away from IR theory’s traditional assumptions. The first maneuver is to replace entities with interactions as the primary object of analysis. All major theories in IR share a substantivist ontology that assumes that entities precede interactions: “units come first, then, like billiard balls on a table, they are put into motion and their interactions are the patterns we observe in political life” (Jackson and Nexon, 1999: 293). In these theories, the entities themselves (i.e. their constitutive properties) are unchanged; it is their attributes (i.e. capabilities and preferences) that vary. In contrast, this article adopts a relational ontology (i.e. network theory) that conceives of structure as a network of relations in which agency is not a singular, essential entity, but rather a layer of processes that the structure constrains into a social space without predetermining in one way or the other (Jackson and Nexon, 1999: 317–318). Relational ontologies are already adding new impetus to theory building, which has long taken a backseat to hypothesis testing as even some of the discipline’s leading scholars have lamented (Mearsheimer and Walt, 2013). The quantum-holographic approach Pan (2020) develops in this issue perfectly illustrates the power and potential for innovation relational ontologies can bring into IR theory.
The second maneuver is to reconceptualize power as a relational attribute deriving from one’s position within a network, which determines how easily one can access resources and information from others and shape the flows and interactions among them (Hafner-Burton et al., 2009: 570). IR scholars are already adopting network approaches to develop new theoretical perspectives on alliances (Maoz, 2010). Recent scholarship has shed light on previously unexplored dimensions of their behavior like how network dynamics operate differently in military alliances and trade relations (Maoz, 2006, 2012; Maoz et al., 2006). Maoz et al. (2007), Warren (2010), Cranmer et al. (2012b), and Kinne (2013) demonstrated that the formation and operation of alliances are strongly influenced by triadic closure (Burt, 1992, 2005; Granovetter, 1973). Warren (2010) and Li, Bradshaw and Clary, (2017) found empirical evidence for the power of indirect ties, showing that triadic patterns of amity and enmity exercise a powerful influence over the selection of alliance partners. Cranmer et al. (2012a) affirmed that exponential random graph models (ERGMs), a commonly used network analysis technique, gives different results than dyadic treatments (e.g. logistic regression) on how joint democracy, major power status, and history of conflict affect alliance behavior. Network-centric approaches also empirically observed the coevolution of military cooperation and economic exchanges (Kinne, 2018; Kinne and Bunte, 2018).
This article argues that alliances are constructed through interactions within transnational social networks of political, military, and business elites in the prospective allies. In every social network, “certain people are connected to certain others, trusting certain others, obligated to support certain others, dependent on exchange with certain others” (Burt, 2005: 4). The extent, nature, and power of these connections varies as a function of prior contact, exchange, and attendant emotions (Burt, 2005: 11). Not everyone is connected to everyone else, not everyone holds the same type or level of information and resources, and, therefore, each actor has a different set of actions and opportunities available to them at any given moment.
Sociologists describe this asymmetry as “social capital”: the “sum of the resources, actual or virtual, that accrue to an individual or a group by virtue of possessing a durable network of more or less institutionalized relationships of mutual acquaintance and recognition” (Bourdieu and Wacquant, 1992: 112). Social capital makes possible the achievement of certain ends that would not be attainable in its absence: people who are connected to each other through mutual association or previous exchanges develop rational expectations about each other’s behavior, which reduces uncertainty and enables cooperation. Social capital is relationally defined (Burt, 2004: 351), and certain actors possess more of it because they are uniquely positioned to bridge the network’s structural holes by forming connections between others who would otherwise remain unconnected (Burt, 1992). This bridging ability is known as brokerage (Stovel and Shaw, 2012), and an actor’s social capital is a derivative of the brokerage opportunities available to it (Burt, 1997: 355).
This article argues that brokerage is an important but ignored mechanism in alliance behavior. Information asymmetries and commitment problems are not exclusive to relations with adversaries (Fearon, 1995). Alliance commitments are fulfilled only 75% of the time (Leeds et al., 2000). So, how could a state be certain of its ally’s capabilities or its commitment to mutual defense? This is a problem of trust, and social capital plays a crucial role in solving it: relations of material exchange, mutual association, and ideational affinity build trust among the decision-making elites of prospective allies and give them subjective certainty about each other (Wheeler, 2018: 12).
A network-centric approach to alliance behavior has several advantages. It sidesteps IR’s levels-of-analysis problem because SNA can cut across the individual, domestic, and systemic levels. It helps to bridge the gap between realist materialism and constructivist idealism because the same methodology can be applied to study both material and nonmaterial flows. It is adept in tracing path dependence because it takes into account the social history of interpersonal interactions. And, finally, it captures the co-constitutivity of agent and structure because it takes nodes and their behaviors as mutually dependent, not autonomous, and views the persistent patterns of association among nodes as creating structures that can define, enable, or restrict their behavior (Hafner-Burton et al., 2009: 562).
It also demonstrates the possibilities theory-based archival research opens up for theoretical innovation and interdisciplinary cross-fertilization. Historical substance adds nuance, diversity, and complexity to IR scholarship while IR’s theoretical and methodological repertoire offers an effective toolkit to extract and process information from historical sources. Archives give historians access to information that would be prohibitively difficult to access in a contemporary setting. IR scholars searching for evidence of a general committing acts of industrial espionage in another country to benefit one company over another (Goltz to Mauser, 25.11.1887), a head of state lobbying to secure business for an ally (Kaiser Wilhelm II to Chancellor Caprivi, 19.11.1891), or an arms dealer currying favor with another country’s senior officials (Huber to Mauser, 08.12.1892) would face legal liabilities, ethical concerns, and risks to their personal safety. Empirically and systematically examining the patterns, practices, and consequences displayed in the archival material uncovered by historians, however, requires rigorous analytical methods like SNA.
Quantitative empirics
As described earlier, brokers gain their influence from their centrality in a network. There are many centrality measures, but the two that are most relevant for this article are betweenness and eigenvector. Betweenness centrality is the extent to which a node lies on the shortest path between other nodes (Newman, 2010: 185). A drawback of this measure, however, is that it cannot account for how a node’s importance in a network increases by having connection to other nodes that are themselves important. Eigenvector centrality resolves this problem by giving each node a score proportional to the sum of the scores of its neighbors (Newman, 2010: 169). To illustrate this distinction, take the example of the movie actor collaboration network (Newman et al., 2002; Watts and Strogatz, 1998): Robert de Niro has the highest eigenvector centrality because many of his collaborators are Hollywood’s leading stars themselves, but AR Rahman and Irrfan Khan’s betweenness centrality scores are higher than de Niro’s because they connect Hollywood and Bollywood.
There are also structural properties that affect brokerage. One is density, which is measured as the proportion of direct ties in a network relative to the total number possible. The other is the clustering coefficient, which indicates how tightly knit the network is. In dense and tightly knit networks, brokerage is much less valuable because more nodes have connections with each other, decreasing the number of structural holes in the network and eliminating opportunities for brokers that bridge them.
Combining archival research with SNA is an ambitious enterprise. Archival research is time-consuming and can require significant resources. Many methods of SNA use statistically advanced techniques that require extensive training. These challenges, however, are not as a high barrier to entering such research as is commonly assumed. To demonstrate it, this article deliberately adopted an approach that requires the least level of statistical sophistication and material resources. Brokerage is easily measured using descriptive empirics like centrality and density. The network was pieced together using primary sources already identified in secondary literature like Stern (1977), Wallach (1976), and Yorulmaz (2014). It must be cautioned that using secondary literature alone would be problematic as it requires uncritical acceptance of the accuracy of another scholar’s interpretation of a primary source. Hence, the ideal scenario is for the researcher to work with the original documents, possess the language proficiency required to read and understand them, and have the material resources to follow the archival trail from them, as was done in this article.
The social network studied in this article was compiled from over 50 primary sources, including material from various archives in different countries. The network has 37 nodes and 79 edges, as visualized in Appendices A and B. Its nodes are individuals—political, military, and business elites—and the edges indicate a material (e.g. money) or immaterial (e.g. information) exchange between them. The edges are not weighted because the content of the exchange holds more significance than their frequency. Statistical analysis using network-analytic software provides empirical support for Goltz’s brokerage role in the transnational social network of German and Ottoman elites. Goltz has an eigenvector centrality of 1.00, the highest score possible, and a betweenness score of 197, which is the highest in the network and about four times the network average. A second, but equally important, finding is that Goltz’s influence persists even in his absence, largely thanks to his star cadet, Mahmut Sevket Pasha, stepping in his place. Mahmut Sevket Pasha has the second-highest eigenvector centrality (0.650) despite having fewer direct connections (i.e. degree centrality) in the network, demonstrating his position of influence. The network’s structural properties also add to the importance of brokerage: its density is fairly low (0.119/1.00), as its average clustering coefficient (0.25/1.00). Only a quarter of possible connections between three nodes with two or more nodes connected (i.e. triplets) are complete. It must be cautioned even though the Sultan and the Kaiser also have high centrality scores, this is misleading since they are the endpoints, not bridges, in the flow of money, information, and influence.
The following section provides a comprehensive narrative of the role Goltz played in brokering the Turco-German alliance through three interrelated processes that created similarity, connectedness, and exchange. The narrative also demonstrates how elite brokerage added to the resilience of the relations between Germany and Turkey. In the early years, Bismarck and his banker, Gerson von Bleichröder, were the dominant influences in Turco-German diplomacy. With Kaiser Wilhelm II’s accession to the throne in 1890, Bismarck fell from power, but, thanks to Goltz’s actions, Berlin’s business in the Ottoman empire went on as usual. Goltz’s retirement in 1895 brought a decline in German influence, particularly in the arms trade, but the tide reversed with Marschall von Bieberstein’s appointment as ambassador in 1897 and Kaiser Wilhelm’s second state visit to the Ottoman Empire (Orientreise) the following year. After the Young Turk Revolution in 1909 deposed the pro-German Abdülhamid II, Berlin initially feared that the new regime would reverse its influence, but these fears did not come to pass. German influence persisted among Ottoman elites, especially in the military, and Mahmut Sevket Pasha, one of Goltz’s longtime protégés, rose to prominence after his role in suppressing the Counter-Revolution made him a revolutionary hero. By 1914, Berlin’s influence in Istanbul had grown roots deep enough to pull the Ottomans into war on Germany’s side in spite of opposing forces, and it endured for decades thereafter.
Colmar von der Goltz and the brokerage of the Turco-German alliance
Origins: 1876–1885
Germany was a latecomer to the scramble over the Ottoman Empire. In his December 1876 speech to the German parliament, Bismarck had famously remarked that the “Eastern Question” is “not worth the healthy bones of a single Pomeranian grenadier” (Schöllgen, 1984: 16). This outlook changed with the reign of Sultan Abdülhamid II. Abdülhamid II was installed to the throne in 1876 to establish constitutional monarchy, but instead he abolished the parliament and went on to rule the country as “the most despotic of sultans” for three decades (Deringil, 1991: 345). The Sultan “hated Russia, feared Britain, and looked down on France,” but he had strong pro-German sympathies (Karal, 1962: 173). He saw in Germany not only an ally against Russia, with whom the Ottomans had been in a constant state of war for over three decades, but also a benefactor for his autocratic ambitions.
Less than 6 months before Abdülhamid II’s coronation, Istanbul had declared default due to a trifecta of growing trade deficits, increasing military expenditures, and decreasing public revenues that sunk the country into a debt crisis (Tuncer, 2015: 58–62), and, only a few months into his reign, the Russo-Turkish War of 1877–1878 broke out, which ended disastrously for the Ottomans and further worsened their problems. When the Congress of Berlin convened in June 1878, Ottoman debt was standing at over 200 million pounds and more than half of the budget was spent on servicing it (Pamuk, 1987). Resolving the debt issue took over a half-decade of diplomacy and culminated in the Decree of Muharram in December 1881, which relinquished the management of the Ottoman Empire’s public finances to a European condominium, the Ottoman Public Debt Administration (Düyun-u Umumiye-i Osmaniye Varidat-ı Muhassasa İdaresi, OPDA).
The OPDA was Germany’s first foray into the Ottoman Empire. The force behind it was Bismarck’s banker, Gerson von Bleichröder, whose business associates like Justizrat Primker and Rudolf Lindau served as its German delegate (Stern, 1977: 419), and profited handsomely from it as his Régie Company acquired in 1883 the lucrative concession for the Ottoman tobacco monopoly (Birdal, 2010: 135). Immediately after the Decree of Muharram, the Sultan dispatched two envoys—Ali Nizami Pasha and Resit Bey—to Berlin, where they met with Bismarck, whose interest in and knowledge of the Ottoman Empire’s affairs surprised his guests (BOA, Y.EE. 7/6, 24.12.1881). A few months later, at the Sultan’s request, Germany dispatched a four-person military mission to the Ottoman Empire: Major-General Kähler, the mission chief, and Colonels Kamphövener, Ristow, and von Hobe, who were seconded to the infantry, artillery, and cavalry corps. They were soon to be joined by a fifth member: Colmar von der Goltz, who was put in charge of military education.
The Ottoman elites, including the high command, were notoriously corrupt (Kähler to Headquarters, 27.06.1882) and they viewed the German reformers with distrust (Radowitz to Bismarck, 01.03.1884). Furthermore, the Sultan feared another coup, so he kept the cadets confined to the barracks, banned trainings and maneuvers, and filled the military academy with his informers (Nezir-Akmese, 2005: 23). This made Goltz’s job especially difficult, but he ultimately succeeded in getting the Sultan to pull his spies off, which also earned Goltz the respect of his Ottoman counterparts (Radowitz to Bismarck, 07.03.1884). Hence, when Kähler died in 1885, he was the person that took his place.
Goltz’s rise to Bismarck’s fall: 1885–1890
One of Goltz’s first actions was to convince the Sultan into an ambitious plan for coastal fortifications to defend the Bosporus and the Dardanelles against a possible Russian attack, which served as a pretext for the purchase of 891 guns from Krupp (Yorulmaz, 2014: 104), followed by 17 German-made torpedo boats (BOA. Y.PRK.ASK. 35/60, 20.10.1886). This was a significant victory for Krupp, which had been courting the Ottomans for years. Krupp’s agent in Istanbul, Otto Dingler, had even caught a break in 1883 after convincing the Navy Minister to replace the fleet’s British-made muzzle-loaded guns with 190 Krupp-made breech-loading guns (BOA. Y.PRK.KOM 4/32, 28.11.1883). Ahmet Muhtar Pasha, who oversaw military purchases, visited Essen as a guest of Krupp (Beziehungen zur Türkei, 25.05.1916), but no deal was ultimately agreed.
Goltz had not intended for a long stay in Turkey, so, a few months later, he relayed his plans to return to Germany, but his work was just beginning, and neither the Sultan nor the Kaiser had any intentions of letting him leave. Privy secretary (mabeyn başkatibi) Ragip Bey was meeting with Goltz “almost daily” to change his mind (Memorandum of Ritter von Manega, 06.06.1886), and, on 19 July, under the Kaiser’s orders, Goltz ended up renewing his contract for another term (BOA, Y.PRK.ASK.33/62, 22.07.1886).
The Krupp purchase became Goltz’s modus operandi: he sold the Sultan and his officers an idea, and German companies sold them the means to turn it into reality. Goltz’s next big idea was to reform the conscription law (Hartmann, 2016) and massively expand the army along the lines of the “military-nation” concept he advocated in his 1883 book, Das Volk in Waffen, which was approved in October 1886. More soldiers meant more weapons, and more business for the German companies that sold them. Arms industrialist Isidor Loewe already had advance knowledge about Goltz’s plans via the Huber Brothers, his agents in Istanbul, and he forged a partnership with rifle manufacturer Paul Mauser as early as May (Abschrift von der Versammlung, 02.05.1886). When Mauser arrived in Istanbul on 22 November, less than a month after the law’s passing, the commission tasked with choosing the new rifles had already started working, but Goltz and Kamphövener were both on the commission, and, through their intervention, Mauser belatedly joined the tests (Seel, 1993: 45).
Around the same time, there arose in Istanbul the rumors that Mauser was offering the Ottomans an inferior model. To dispel them, Goltz requested Bismarck’s War Minister, Paul Bronsart von Schellendorf (Goltz to Schellendorf, 13.12.1886) to write to the Sultan and meet with the Ottoman ambassador, Tevfik Pasha (BOA, Y.PRK.MYD 29/19, 31.10.1887). Bismarck also sent a letter, presumably at Goltz’s urging, and, according to British diplomats, it was what tipped the scale in Mauser’s favor (Unsigned to Salisbury, 27.01.1887). When Mauser received the official order (BOA, Y.A.HUS.486/9, 06.02.1887), Loewe excitedly credited his success to the “very energetic and tireless support of the Chancellor [Bismarck] and the self-sacrificing support of Goltz Pasha” (Seel, 1981: 802).
Two of Goltz’s local contacts, Ragip Bey and Mahmut Sevket, also played key roles. Ragib Bey, the privy secretary, was one of Goltz’s most reliable sources inside the court and a personal friend of Paul Mauser, with whom he regularly exchanged letters (Ragib to Mauser, 17.10.1887). Krupp’s agent in Istanbul, Menshausen, described Ragib Bey as Germany’s “only friend” and his “only influential contact” (Menshausen to Geheimrat, 23.12.1891). Ragip Bey was also renowned for his avarice, and the British ambassador, Sir William White, believed that he was on Mauser’s payroll (White to Salisbury, 28.06.1887). Mahmut Sevket Pasha, then a young colonel, had been a Goltz protégé since his days as a cadet. Goltz introduced him to Mauser as the “most suitable person” to conduct business (Seel, 1981: 800), a judgment that was vindicated via his help in sidelining Mauser’s chief detractor, war minister (Serasker) Ali Saib Pasha, during the commission in 1886 (von der Goltz, 1913: 35).
The Ottomans also had to pick cartridges for their rifles, and this time Mauser had a German rival, the Lorenz Company, which had both the experience and the technology, as well as backing from Ali Saib Pasha himself (BOA, Y.PRK.MYD.6/21, 14.03.1887). Berlin, however, was on Mauser’s side because his cartridges were to be produced at the Pulverfabrik-Rottweil, whose owners, the Duttenhoffer Brothers, were close friends with Bismarck (BOA, Y.PRK.MYD.6/34, 08.05.1887). Bismarck wrote a letter to the Sultan on Rottweil’s behalf and praised it as the choice “most appropriate for the Mauser rifles currently in the hands of the Ottomans” (BOA, Y.PRK.EŞA.6/61, 19.04.1887). The German ambassador, Radowitz, also extolled that it is “the best powder for the ammunition used in the Mauser rifles” (BOA, Y.PRK.ASK.39/67, 21.04.1887). The reality, however, was starkly different. The Ottomans knew neither that “the performance of the [Rottweil] powder is not yet clear” (Mauser to Goltz, 30.11.1887), nor that Bismarck was personally invested in Pulverfabrik-Rottweil (Stern, 1977: 209; Vagts, 1968: 216–217). Once again, Goltz saved the day. He surreptitiously shared with Mauser the test results from the powder samples of its French competitor (Goltz to Mauser, 25.11.1887) while erroneously reporting to the Sultan that “the French powder, the quality of which was exaggerated, was not as durable as desired” (BOA, Y.PRK.ASK.52/77, 02.01.1889). In September 1887, Pulverfabrik-Rottweil won the contract for 100 million cartridges.
Bismarck’s fall to Goltz’s retirement: 1890–1895
In the meanwhile, German investors were setting their eyes on another prize: railroads. The Mauser company’s majority stake belonged to Württembergische Vereinsbank, and, through its Turkish business, the bank’s director Alfred von Kaulla came into contact with Wilhelm von Pressel in early 1888. Pressel was the longtime advisor to the Ottoman Railways and he had drawn up plans for a railroad from Istanbul to Basra but his fundraising efforts had repeatedly failed (Özyüksel, 2016: 24). Kaulla partnered with a fellow banker, Deutsche Bank’s Georg von Siemens, and they set out to turn Pressel’s vision into reality (Özyüksel, 2016: 27). In a fortuitous coincidence, Baron Hirsch was looking to sell his ill-fated Oriental Railways, an ambitious scheme to build 1332-kilometer-long railway linking Vienna to Istanbul (Özyüksel, 2016). This was an opportunity for Kaulla and Siemens to extend their railroad from Basra to Berlin, but such an investment was impossible without official backing, which they secured with the help of Bleichröder, who also had a stake in the Oriental Railways. At his pleading, Bismarck halfheartedly gave his blessing, but he also warned Kaulla and Siemens that they “must not count upon the protection of the German Empire against eventualities connected with precarious enterprises in foreign countries” (Stern, 1977: 421).
Later that year, Kaiser Wilhelm II acceded to the throne, and Bismarck’s power started to abate. Two years later, he was forced out of office, but Germany’s business went on as usual thanks to Goltz, who was now a member of the Imperial Household (Mabeyn-i Humayun) and enjoying direct access to the Sultan and his courtiers. A few years later, a new crisis erupted after Mauser’s Belgian licensee, Fabrique Nationale, unknowingly revealed that Mauser had a newer model that it kept secret the Ottomans. Even though Mauser quickly upped his offer, Istanbul was furious (Huber to Mauser, 08.12.1892), and its placation required a “generous distribution of gratuities (baksheesh)” (Col. Chermside’s Despatches No. 19–21, 01.07.1890; Berlin (Direction) to Mauser, 04.09.1890).
Soon enough, Mauser was back in favor. In late 1893, the company secured another contract for 200,000 additional rifles, for which Mauser personally thanked Goltz, who once again did “what needed to be done” (Mauser to Goltz, 26.11.1893; Huber to Mauser, 09.12.1893). Mahmud Sevket Pasha, who was quickly rising up the ranks, also played a key role in this success. “[Mahmud Sevket Bey] sacrificed so much for us,” wrote Huber to Mauser, “that we are inclined to assumed that he would be prepared and expecting to earn something decent” (Huber to Mauser, 08.12.1892).
Kaiser Wilhelm II’s reign was also profitable for the Krupps, with whom the Kaiser was so close that he “regarded himself and was regarded by them as family” (Yorulmaz, 2014: 166). In late October, Goltz received from Menshausen the news that the Ottomans were in talks with Krupp’s French rival, the Canet Company, which he immediately reported to both the Kaiser (Goltz to Kaiser Wilhelm II, 28.10.1891) and Kiderlen (Goltz to Kiderlen, 10.11.1891). Alfred Krupp had also asked the Kaiser for help, who ordered the ambassador in Istanbul, Radowitz, to “give his support to Mr. Krupp’s representative [Menshausen] and point out to the Sultan that such a concession to France, which would ipso facto damage the German industry, could not be without political consequences” (Kaiser Wilhelm II to Chancellor Caprivi, 19.11.1891).
Two weeks later, the Ottoman ambassador in Berlin, Ahmet Tevfik Pasha, was called into a meeting with the Kaiser for an ultimatum. Bemoaning the “horrible and upsetting impression [left by] rumors running around in the circles of the German industrialists that Abdulhamid intends to turn away almost absentmindedly from the German industry and contemplates entrusting the French with orders for ships and cannons,” the Kaiser warned that “if the Sultan, despite his satisfaction over many years with proven deliveries, and the progress of railroad construction undertaken by German industry, puts the German interest behind the French interest, German capital would have no reason to take risks to accommodate the Sultan’s private wishes or to continue their efforts supporting his country’s progress” (Kaiser Wilhelm II to Chancellor Caprivi, 09.12.1891).
Soon after the meeting, Kiderlen wrote to Goltz, asking him to “control and correct Tevfik Pasha’s report” and informing that Berlin would be “very interested in knowing how it is received” (Kiderlen to Goltz, 10.12.1891). In the meanwhile, Menshausen was also plotting with the Foreign Office to block the Ottoman Imperial Bank from funding the French purchase (Menshausen to Geheimrat, 23.12.1891). A week after the Kaiser’s ultimatum, on 16 December, the Sultan granted an audience to Goltz Pasha and assured him that he does not intend to change course to a policy that favors France (Goltz to Geheimrat, 18.12.1891). Menshausen secretly met with Ragip Bey the same night and was given similar assurances (Menshausen to Krupp, 20.12.1891). A few days later, the Canet deal was cancelled.
Goltz’s retirement to the Young Turk Revolution: 1895–1909
With his second term nearing completion, Goltz was looking forward to his return to Germany, but he was once again implored to stay. Ambassador Radolin strongly advised against Goltz’s departure, describing him as “invaluably important” for Germany’s influence over the military and the success of its arms trade (Radolin to Caprivi, 20.5.1893). But, Goltz’s intimate knowledge of Istanbul’s inner workings, its corruption and intrigue, had also earned him powerful enemies, in and out of uniform (Goltz to Radolin, 09.08.1893). Goltz eventually agreed to stay, but the courtiers and the generals gave him so much trouble during his command inspection (Generalstabsreise) on the Greek border a few months later that he retired soon after it ended (Krethlow, 2012: 123).
Goltz’s departure left a void because he was uniquely powerful. No one else enjoyed the kind of access and influence he had on both the Sultan’s court and the High Command (Saurma to Hohenlohe, 7.8.1895). For several years after his departure, there were only three German officers on official duty in Istanbul, including Major-General Grumbkow, who was left paralyzed after a stroke, and Kamphövener Pasha, whose quarter-century of service barely left an imprint (Kramer and Reinkowski, 2008: 84). With Bieberstein’s appointment as ambassador in July 1897 (Özyüksel, 2016: 47), the embassy became more active, and the military attachés came to serve a role similar to Goltz’s, albeit with much less influence.
The most important factor in Germany’s continued influence was Kaiser Wilhelm’s personal diplomacy, especially during his second official visit (Orientreise) in October 1898, which was, as affirmed at the time by both the Kaiser and foreign observers, a total success for Berlin (Angell to May, 25.07.1898; Wilhelm II, 1922: 90). The Kaiser eagerly flaunted his “warm friendship” with the Sultan (Bieberstein to Hohenlohe, 24.05.1898) in his almost-daily telegraphs to him (Bülow, 1931: 542), as well as public occasions like his Damascus speech where he assured “his Majesty the Sultan and the 300 millions of Muslims . . . [who] revere in him their Khalif that the German Emperor will ever be their friend” (Bülow, 1931: 254), an episode that so deeply affected the Sultan that he fondly recounted it to the German ambassador almost a decade later (Bieberstein to Auswärtiges Amt, 22.03.1907).
The Orientreise, however, was more than a friendly visit. As British diplomats observed, it secured for Berlin “the concession of the Baghdad Railway, a monopoly on all order for military munitions for the Turkish Army, and a privileged position for all industrial and commercial concessions which it was in the power of the Sultan to bestow upon his friend and patron” (Annual Report for Turkey, 1907).
The Baghdad Railway was the most prized victory, especially considering that Deutsche Bank almost sold it to the Russians 2 years ago as it saw no hope for it (Barth and Whitehouse, 1998: 122). Along with it, Germany received two other assets: a telegraph cable from Berlin to Istanbul (BOA, Y.A.HUS.390/88, 01.11.1898) and a new port in Istanbul, the Haydarpasa quays, which Britain and France had strenuously opposed (O’Conor to Salisbury, 08.02.1899). Equally important was the Kaiser’s dedication of a new Protestant church in Jerusalem, the Church of the Redeemer (Erlöserkirche), and his acquisition of a site of historic significance, the Abbey of the Dormition, which was viewed by Berlin as blows to French influence in the Levant (Bülow to Kaiser Wilhelm II, 04.06.1898). The Orientreise also won Krupp a new order for 162 guns and 30,000 shrapnel shells (Kossler, 1981: 254).
Young Turk Revolution to the First World War: 1909–1914
The Young Turk Revolution in April 1919 toppled Sultan Abdülhamid II’s government, whose pro-German sympathies and penchant for corruption had given Berlin such influence that, less than a year before his ouster, the German ambassador in Istanbul was boasting that the “War Minister, the Undersecretary of the Ministry of War, Chief of General Staff, the commanders of the Guards Corps and other important commands are all in the hands of the officers who have served in Germany” (Bieberstein to Bülow, 03.09.1908). In contrast, the Young Turks had a “deep hatred” for the Hamidian regime (Hanioglu, 1995: 26) and were aspiring for a constitutional monarchy like in Britain and France, where their success was received with “great sympathy and pleasure” (Ahmed, 1966: 302).
Yet, Berlin’s initial fears of an anti-German turn did not come to pass, largely because many of the new regime’s military leaders were of the “Goltz generation” (Grüßhaber, 2018: 26–103). Particularly prominent was Mahmut Sevket Pasha, who the revolutionaries regarded as a “savior and intellectual godfather” (Provence, 2017: 12) and later appointed as their war minister. Many of Mahmut Sevket’s aides-de-camp were also Goltz’s cadets. One of them even authored a biography of Goltz during his retirement (Demirhan, 1960). The Ottoman ambassador in Berlin, Osman Nizami Pasha, was another Germanophile, who Bieberstein described as “a loyal friend and adviser, [and a] reliable informant [who] used his whole influence in order to make Germany the unique supplier for all Turkish arms orders, [and is] in his mind, more German than Turkish” (Kiderlen to Bülow, 20.08.1908). Consequently, Germany’s influence continued unabated. Within a year of the revolution, Kiderlen was reporting with delight that “the belief that the Young Turk movement is pro-English and anti-German [is] gravely mistaken” (Kiderlen to Bülow, 20.08.1908).
The return to normalcy took slightly longer for Krupp and Mauser because of their close ties to the Hamidian regime. When the Young Turks came to power, one of their first actions was to ask Huber to produce a list of the names of all those who under the old regime had taken bribes in return for orders. There is no evidence that such a list was ever produced, but, had one existed, it would have incriminated many powerful figures, most likely including Mahmut Sevket Pasha himself. Krupp lost an artillery contract in 1908, and Mauser was passed over for its American rival Browning the following year, but, eventually, they regained some of their footing. In 1910, Krupp won a contract for 106 guns, followed by an order for 90 more the following year. The monopoly they enjoyed under Abdülhamid II, however, no longer existed, and they faced increasing competition from France on artillery sales and the British on naval purchases (Grant, 2002: 32–33).
In 1913, a coup d’etat changed the scene once again. At first, it seemed like another victory for Berlin. Two of the “Three Pashas” behind the “Raid on the Sublime Porte” (Babıali Baskını), Enver and Talat, were sympathetic to Germany, and Mahmut Sevket Pasha was installed as the new regime’s grand vizier, though his tenure proved short-lived as he was assassinated 6 months later. A full exposition of Enver Pasha’s relationship with Germany would require a separate article, but his friendship with Hans Humann deserves particular mention (Gottschlich, 2015: 114). The Turkish-born son of a German archeologist, Humann was one of Germany’s “most valuable contacts” during his posting to Istanbul as the naval attaché (Trumpener, 1984: 115). The American ambassador in Istanbul, Hans Morgenthau Sr., describes Humann in his memoirs as “a man of great influence [and] a personal emissary of the Kaiser” (Morgenthau, 1918: 251–222). According to Franz von Papen (1953), Humann’s classmate, who later became Nazi Germany’s ambassador to Turkey, Enver and Hans’s friendship dated back to their childhood (89). Although there is no evidence to back this claim, there is no doubt of their camaraderie: after the defeat in the First World War, Enver took up refuge at the Potsdam villa of Humann’s sister, Maria, and lived there until he left for Central Asia, where he organized the pan-Turkist revolt that cost him his life in 1922.
Despite Enver’s influence, it would be misleading to say that Germany remained uncontested. Some of the new regime’s most influential figures had loyalties elsewhere: Said Halim Pasha, the Grand Vizier, and Cemal Pasha, the Navy Minister, were Anglophiles while Cavid Bey, the Finance Minister, was close with the French. These divisions allowed the Ottomans to conduct a balancing game in the lead-up to the First World War. Istanbul bought massive amounts of arms and ammunition from both France and Germany (Grant, 2002; 32–33) and conducted alliance negotiations with both the Entente and the Central Powers until the very last minute (Aksakal, 2008).
Istanbul’s balancing game came an end after an arms deal gone wrong and the diplomatic crisis that ensued. The Greco-Turkish War of 1897 had entered the two countries into a naval arms race in the Aegean. In March 1910, Greece acquired from Italy the armored cruiser Georgios Averof, which shifted the balance of power and caused alarm in Istanbul (Halpern, 1971: 329). The Ottoman Navy was a pocket of British influence, because of the British naval mission that arrived after the Young Turk Revolution (Rooney, 1998). Istanbul’s entreaties for a naval purchase, however, was turned down by London. As a result, the pro-British naval minister Halil Pasha was removed from office. His replacement was a Germanophile, Mahmud Muhtar Pasha, whose father was hosted as Krupp’s guest during its failed sale in 1883, and he quickly arranged for the purchase of two Brandenburg-class pre-dreadnoughts from Germany (Guleryuz and Langensiepen, 1995: 16–18). These purchases earned Germany a great deal of sympathy in Istanbul. Sultan Resad himself remarked that he “would never forget the service provided by Germany in such critical times” (Miquel to Hollweg, 01.09.1910). Mahmut Sevket Pasha welcomed the commander of the two ships with similar words (Koch, 21.09.1910).
Increasing cooperation between Turkey and Germany raised worries in Britain. To placate London, Istanbul also ordered from Vickers & Sons two state-of-the-art dreadnoughts, Reşadiye and Sultan Osman-ı Evvel, but, as they neared completion, Archduke Franz Ferdinand was assassinated on 28 June, 1914, and the war got into gear. Worried that the Ottomans would enter the war on the side of Germany, and not wanting to put two brand-new warships in the hands of a prospective adversary, Britain first postponed delivery of the ships, and, then requisitioned them for the Royal Navy on 29 July, while an Ottoman delegation was already in the country for the ships’ handover.
The news of the seizure, which coincided with Greece’s purchase of two American battleships, USS Idaho and USS Mississippi, drove the Ottoman public into an anti-British frenzy (Guvenc, 2005). Because of the government’s financial troubles, the battleships were publicly funded via donations to the Ottoman Navy Association, so, in the eyes of the Turkish public, the seizures amounted to theft. Less than 2 weeks later, they also became the pretext for Ottomans to enter the First World War. On 10 August, two battleships of Germany’s Mediterranean Fleet, SMS Goeben and SMS Breslau, escaped an Anglo-French fleet and entered the Dardanelles. The Ottoman Empire was still neutral, so Istanbul had to deny the ships free passage. Instead, in a tit-for-tat move, it declared that the two ships, renamed Yavuz and Midilli, now belonged to the Ottoman Navy, and their commander, Adm. Wilhelm Souchon, was entered into Ottoman service (Aksakal, 2008: 110–118). Two months later, Enver Pasha ordered Souchon and his two battleships to raid Russia’s Black Sea ports, allegedly in conspiracy with Humann (Einstein, 1918: 12), and entered the Ottoman Empire into the war that became its undoing (Aksakal, 2008: 178–182).
Conclusion
This article underlines the untapped potential of research that combines network-analytic methods with archival material, adds to a growing body of scholarship importing network thinking into alliance theory, and provides a convincing case for reconceptualizing alliances as products of interactions within transnational social networks of political, military, and business elites in the prospective allies. These interactions do not take place ex nihilo. Their extent, nature, and power vary as a function of prior contact, exchange, and attendant emotions (Burt, 2005: 11). Not everyone is connected to everyone else, not everyone holds the same type or level of information and resources, and, therefore, each actor has a different set of actions and opportunities available to them at any given moment. These differences also shape the preferences formed, resources possessed, outcomes sought, and strategies employed by the decision-making elite. Brokers hold an outsized influence in this process, because, as they make and break network connections, they also change the opportunities for and constraints on individual action. Thus, brokerage helps to bridge the agency-structure divide in explaining alliance behavior. It also illustrates how prospective allies overcome the problem of uncertainty: people who are connected to each other through mutual association or previous exchanges develop mutual trust and gain subjective certainty about each other. Finally, a network-centric approach that explores elite brokerage offers valuable information on the informal politics of alliance behavior like personal favors, palace intrigues, backroom deals, and quid pro quos, which offers an interesting convergence with the growing scholarly literature on the globalization of corruption (Cooley and Sharman, 2017; Cooley et al., 2018)
Looking forward, this paper presents at least two different directions for future research. One is to dive deeper into the history, studying the networks other European powers like Britain and France had in the Ottoman Empire and why they failed to counter German influence, as well as comparable historical cases of political control through economic penetration like the Principality of Bulgaria (1878–1908), Khedivial Egypt (1867–1922), Kingdom of Greece (1832–1935), and the Kingdom of Serbia (1882–1918) to piece together a global picture of the pre-First World War great-power competition (Tuncer 2015). Another is to put it in the modern context and explore similar mechanisms in the United States and China’s growing strategic rivalry (Allison, 2017; Beckley, 2018; Brands and Cooper, 2019; Brooks and Wohlforth, 2016; Christensen, 2006; Glaser, 2015; Kirshner, 2012; Mattis, 2018; Schweller and Pu, 2011; Shambaugh, 2018). As Winecoff (2020) demonstrates in this issue, network approaches are well-suited to illustrating less apparent dimensions of this great-power competition. Geo-economic competition and the weaponization of interdependence are already attracting growing interest from the academic and policy scholars (Blackwill and Harris, 2016; Farrell and Newman, 2019; Wright, 2017). It is difficult to ignore the similarities between Germany’s economic penetration of the Ottoman Empire through sovereign debt, arms deals, and infrastructure projects such as the Berlin-Baghdad Railroad; its cooptation of Ottoman civilian and military elites; and its subsequent control over Ottoman policy and the current policies and actions of China (Dreher et al., 2018, 2019; Flores-Macias and Kreps, 2013; Kastner, 2016; Norris, 2016; Strüver, 2016; Tuman and Shirali, 2017), as well as Russia (e.g. Sherr, 2013).
Third, this article demonstrates the importance of the ideational dimensions of great-power politics (Haas, 2005; Owen, 2010). The ties between Germany and Turkey were not only material. As described earlier, one source of Goltz Pasha’s power was his influence over military education. Generations of Ottoman officers viewed him as a role model, studied in the curriculum he designed, and were ensconced in the mindset this education instilled. It also created many opportunities for Turkish and German officers to commingle together and build relationships, which allowed German influence to continue unabated even after Bismarck’s downfall, Goltz Pasha’s retirement, Abdülhamid II’s overthrow, and Mahmut Sevket Pasha’s assassination. Ideational homophily matters because it endures. Many of the nationalist leaders that reshaped the ex-Ottoman imperium in the interwar period were all trained by, sympathetic to, and steeped in the worldview of Germany (Provence, 2017).
Take the example of Kemal Ataturk, the founder of modern Turkey. In military school, Ataturk had studied German and read the classics of German military thinking; an 1891 edition of Goltz’s Das Volk in Waffen, dating back to his school days, remained in his library until the end of his life (Gawrych, 2013: 9). He was a lieutenant in Mahmut Sevket Pasha’s armies during the suppression of the 1908 Counter-Revolution. He had personally met Goltz, who he described as a “great savant” (Gawrych, 2013: 10; Hanioglu, 2011: 36), and authored a book (Ataturk, 1918[2010]) that was influenced by his thinking. During the First World War, Ataturk served under German generals and earned an Iron Cross from Kaiser Wilhelm II for his service (Gawrych, 2013: 48). He was a staunch Anglophobe, if not exactly a Germanophile (Provence, 2017: 135). The ideational homophily between two countries influenced bilateral relations for decades. After the re-establishment of diplomatic relations in 1924, many officers forced out by the Versailles Treaty were recruited to teach at the Turkish War Academies (Özgüldür, 1993: 64–65). The Nazis similarly viewed Turkey as a role model for themselves (Ihrig, 2014: 145) and clamored to restore German influence over Turkey through propaganda tours for journalists and training programs for the cadets, invested heavily in the country’s fledgling defense industry, and dispatched civilian and military advisers whose number exceeded 2000 by 1940 (Özgüldür, 1993: 65).
Finally, whether the mechanisms this article illustrates in the late 18th century are part of a particular repertoire that states use in the emergence or existence of multipolarity (e.g. Goddard and Nexon, 2016: 11–13) is a question that closely pertains to global power politics at the present and would reveal valuable lessons for contemporary scholars and practitioners of policy and strategy.
Footnotes
Appendix
Acknowledgements
The author thanks Marlies Glasius and Nina Tannenwald for their unceasing support and Lauren Sukin and Omer Faruk Yalcin for their excellent comments on several previous drafts of this article. He also thanks Nick Beauchamp, Christine Bricker, Bruce Desmarais, Chelsea Estancona, Justin Gross, Michael Heaney, Burak Kadercan, David Kang, Goksu Kolayli, Gunes Kolsuz, Seckin Kostem, Caglar Kurc, Paul MacDonald, Elizabeth Menninga, Alexander Montgomery, Peter Mucha, Kerem Nisancioglu, Chengxin Pan, Zhenggi Pan, Or Rabinowitz, Yotam Shmargad, Alp Eren Topal, Nicholas Wheeler, Einar Wigen, W.K. Winecoff, Kevin Young, Mara Dolan, Amalia Perez and the anonymous reviewers for their feedback, as well as the participants at the EJIR 20th Anniversary Workshop, 2019 ISSS-IS Joint Annual Conference, 2019 Political Networks Conference, and the Brown University Comparative Politics and International Relations Workshop. An earlier version of this article won the 2020 Carl Beck Award of the International Studies Association and a poster award at the 2019 Political Networks Conference.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Primary sources
BOA.Y: Başbakanlık Osmanlı Arşivi, Yıldız [Ottoman State Archive]
A.HUS: Sadaret Hususi Maruzat Evrakı [Office of the Grand Vizier-Private]
A.RES: Sadaret Resmi Maruzat Evraki [Office of the Grand Vizier-Official]
EE: Esas Evrakı [High Offices]
PRK.ASK: Askeri Maruzat Evrakı [Military Affairs]
PRK.EŞA: Elçilik ve Şehbenderlik Maruzatı [Ambassadorial and Consular Affairs]
PRK.MYD: Yaveran ve Maiyet-i Seniye Erkan-ı Harbiye Dairesi [High Command-General Staff]
PRK.KOM: Yıldız Perakende Evrakı Komisyonlar Maruzatı [Commissions]
BArch-MA: Bundesarchiv Militärarchiv [German Military Archives]
HA, Krupp: WA Historisches Archiv-Krupp – Werk-Archiv Essen [Krupp Company Archıves]
NA: National Archives, London
NARA: National Archive and Records Administration; College Park, MD
PA.AA: Politisches Archiv des Auswärtigen Amtes [German Foreign Ministry]
SA: Stadt-und Zeitungsarchiv Oberndorf am Neckar [Oberndorf City Archive]
Gooch and Temperley:
British Documents on the Origins of the War, 1914–1918. London: Her Majesty’s Stationary Office.
Lepsius et al.: Lepsius J, Mendelssohn-Bartholdy M and Thimme FWK (eds) (
) Die große Politik der europäischen Kabinette, 1871–1914: Sammlung der diplomatischen Akten des Auswärtigen Amtes. Berlin: Deutsche Verlagsgesellschaft für Politik und Geschichte
BOA, Y.A.HUS.486/9, 06.02.1887, in: Mübayaat-ı Mühimme Defteri (HH 26363:7)
BOA, Y.A.HUS.390/88, 01.11.1898
BOA, Y.EE. 7/6, 24.12.1881
BOA, Y.PRK.ASK.33/62, 22.07.1886
BOA. Y.PRK.ASK. 35/60, 20.10.1886
BOA, Y.PRK.ASK.39/67, 21.04.1887
BOA, Y.PRK.ASK.52/77, 02.01.1889
BOA, Y.PRK.EŞA.6/61, 19.04.1887
BOA. Y.PRK.KOM 4/32, 28.11.1883
BOA, Y.PRK.MYD.6/21, 14.03.1887
BOA, Y.PRK.MYD.6/34, 08.05.1887
BOA, Y.PRK.MYD.29/19, 31.10.1887
Abschrift von der Versammlung, 02.05.1886, in: SA, Oberndorf: M-A3
Angell to Day, 25.07.1898, in: NARA: M46/65
Annual Report for Turkey for 1907, in: Gooch and Temperley, Vol. 5, p. 43.
Beziehungen zur Türkei, 25.05.1916, in: HA, Krupp: WA 7f -886
Col. Chermside’s Despatches No. 19-21, 01.07.1890, in: NA: FO 78/4276
Berlin (Direction) to Mauser, 04.09.1890, in: SA, MA-4
Bieberstein to Auswärtiges Amt, 22.03.1907, in: PA.AA. R13775
Bieberstein to Bülow, 26.12.1907, in: NA: GFM 10/11
Bieberstein to Bülow, 03.09.1908, in: Lepsius et al., Vol. 25: No. 8911
Bieberstein to Hohenlohe, 24.05.1898, in Lepsius et al., Vol. 12.2: No. 3340
Bülow to Kaiser Wilhelm II, 04.06.1898, in: Lepsius et al., Vol. 12.2: No. 3368
Goltz to Geheimrat, 18.12.1891, in: PA.AA. R13763
Goltz to Kaiser Wilhelm II, 28.10.1891, in: PA.AA. R13763.
Goltz to Kiderlen, 10.11.1891, in: PA.AA. R13763.
Goltz to Mauser, 25.11.1887, in: SA: M-A8
Goltz to Schellendorf, 13.12.1886, in: SA: M-A8
Huber to Mauser, 08.12.1892, in: SA: M-A8
Huber to Mauser, 09.12.1893, in: SA: M-A8
Kähler to Headquarters, 27.06.1882, in: BArch-MA: N65/11
Kaiser Wilhelm II to Chancellor Caprivi, 19.11.1891, in: PA.AA. R13285
Kaiser Wilhelm II to Chancellor Caprivi, 09.12.1891, in: PA.AA. R13763
Kiderlen to Bülow, 20.08.1908, in: PA.AA. R13745
Kiderlen to Goltz, 10.12.1891, in: PA.AA. R13763
Mauser to Goltz, 26.11.1893, in: SA: M-A8
Mauser to Goltz, 30.11.1887, in: SA: M-A8
Memorandum of Ritter von Manega, 06.06.1886 in: PA.AA. R13237
Menshausen to Geheimrat, 23.12.1891, in: PA.AA. R13763
Menshausen to Krupp, 20.12.1891, in: PA.AA. R13763
Miquel to Hollweg, 01.09.1910, in: PA.AA. R13312
O’Conor to Salisbury, 08.02.1899, in: NA: FO78/5000
Radolin to Caprivi, 20.05.1893, in: PA.AA. R14311
Radowitz to Bismarck, 01.03.1884, in: PA.AA. A1532
Radowitz to Bismarck, 07.03.1884, in: PA.AA. A1605
Ragib to Mauser, 17.10.1887, in: SA: M-A3
Reinhard Koch, ‘Militärpolitischer Bericht über die Überführung Euerer Majestät Schiffe “Kurfürst Friedrich Wilhelm” und “Weißenburg” und Übergabe an die Kaiserlich Ottomanische Regierung’, 21.09.1910, in: PA.AA. R13313
Saurma to Hohenlohe, 07.08.1895, in: PA.AA. A-9091
Unsigned to Salisbury, 27.01.1887, in: NA: FO 78/4002
White to Salisbury, 28.06.1887, in: NA: FO 78/3998
Bülow, B. (1931) The Memoirs of Prince von Bülow. New York: Little, Brown & Co.
Einstein L (1918) Inside Constantinople: A Diplomatist’s Diary During the Dardanelles Expedition, April-September 1915. New York: E.P. Dutton.
Morgenthau H (1918) Ambassador Morgenthau’s Story. Garden City, NY: Doubleday.
von der Goltz C (1899) Das Volk in Waffen: Ein Buch über Heerwesen under Kriegführung unsered Zeit. Berlin: V. Decker.
von der Goltz C (1913) Erinnerungen an Mahmud Shewket Pascha. Deutsche Rundschau 157: 32–46.
von Papen, F (1953) Memoirs. New York: E.P. Dutton & Co.
Wilhelm II (1922) The Kaiser’s Memoirs. New York: Harper & Brothers.
