ssrn-5334999

1 In Search of Coherence: Assessing the Structure of Geopolitical Explanations Mateusz Ambrożek, PhD Faculty of Business and International Relations Vistula University ul. Stokłosy 3, 02-787 Warszawa, Poland e-mail: m.ambrozek@vistula.edu.pl Abstract: Classical and neoclassical geopolitics are widely criticized within IR theory. Stefano Guzzini’s dilemma captures this critique: classical geopolitics is deterministic and flawed, while the neoclassical variant is accurate but trivial. The central problem with this critique – aimed at rejecting geopolitics as a research approach – lies in its reliance on heterogeneous types of objections: empirical in the classical case, and theoretical in the neoclassical version. This inconsistency hinders a unified evaluation. This article proposes a framework based on explanatory coherence, understood as a key feature of scientific explanation. Coherence is conceptualized along two dimensions: quantitative (the number of variables involved in the explanation), and qualitative (the internal consistency of the mechanisms). Three types of explanations are examined: classical geopolitics, positive feedback between geography and political phenomena, and neoclassical geopolitics. Only classical geopolitics fulfills both dimensions of coherence; the positive feedback satisfies one, and the neoclassical satisfies neither. Consequently, only classical geopolitics may be regarded as coherent among the examined geopolitical explanations. However, the IR consensus rejecting geographical determinism implies the necessity of rejecting geopolitics as such. Therefore, this article provides a decisive argument against geopolitics – an argument that Guzzini could not provide. Keywords: Geopolitics, classical geopolitics, neoclassical geopolitics, geographical determinism, coherence -- 1 of 31 -- 2 Introduction Geopolitics, in both its classical and neoclassical variants, remains a contested approach within international relations (IR) theory. This contested status stems from two fundamental issues. On the one hand, classical geopolitics has been criticized for its geographical determinism, which reduces complex political phenomena to spatial conditions. On the other hand, its neoclassical counterpart – an attempt to overcome determinism by incorporating non- geographical variables – has been accused of insufficiently emphasizing the geographical factor and lacking theoretical clarity. As a result, geopolitics faces a dilemma aptly formulated by Stefano Guzzini, who argues that it ‘[i]s either distinct and determinate but wrong or indistinct, indeterminate and correct but trivial’ (Guzzini, 2017: 428). This tension constitutes one of the most significant challenges confronting the approach. The absence of a clearly defined ontological foundation undermines the epistemological status of geopolitics and calls into question its place within the broader theoretical landscape of IR. However, the problematic nature of Guzzini’s argument lies in the asymmetry of his evaluative criteria: classical geopolitics is assessed in empirical terms, while its neoclassical variant is judged on theoretical grounds. Yet, a proper assessment of geopolitics requires a consistent theoretical analysis – one that asks whether geopolitical explanations, despite their evolution, meet the fundamental criterion of theoretical adequacy: coherence. This article seeks to conduct precisely such a theoretical analysis. It seeks to accomplish what Guzzini did not – to examine how the reconceptualization of the role and significance of geography, along with the inclusion of additional, non-geographical variables, has affected the coherence of geopolitical explanations. Both classical and neoclassical variants are evaluated across two dimensions of theoretical coherence: quantitative (the number of variables included in the explanation) and qualitative (the internal consistency of the explanatory mechanisms). The theoretical analysis proceeds in two stages. In the first, three types of geopolitical explanation are identified and classified based on the relationship between geographical factors and political phenomena. The first type (classical geopolitics) assumes a unidirectional, deterministic influence of geography on politics. The second type, rejecting determinism, posits that states are capable of modifying the effects of geographical space through technology, resulting in positive feedback. The third type (neoclassical geopolitics) is characterized by a multitude of vague explanatory mechanisms and the arbitrary selection of variables, with geography being acknowledged rhetorically but playing a marginal role in practice. -- 2 of 31 -- 3 In the second stage, drawing on the two aforementioned dimensions of coherence, the analysis demonstrates that the further geopolitics diverges from the classical model, the more it loses its internal coherence. Only the first type of geopolitical explanation satisfies the criteria for both dimensions of coherence. The second type meets only the qualitative dimension, while the third type fails to satisfy neither. These findings imply that only classical geopolitics can be considered coherent, which would imply its theoretical validity within IR theory. However, the widespread critique of geographical determinism across the discipline necessitates the rejection of geopolitics as such. In this sense, the article offers a decisive theoretical argument against geopolitics – an argument that, it seems, Guzzini himself was unable to fully articulate. The structure of the article comprises five sections. The first outlines the fundamental features of the geopolitical approach, the two main strands of its criticism, and a reformulation of Guzzini’s dilemma. The second section discusses the two dimensions of theoretical coherence. The third presents an original typology of geopolitical explanations based on the identification of three types of geopolitical explanations. The fourth section provides an assessment of the coherence of these models in both quantitative and qualitative terms. The final section offers conclusions drawn from the analysis and reflects on their implications for contemporary IR theory. Classical and Neoclassical Geopolitics as Research Approaches in IR Theory This literature review focuses on two key issues. First, it outlines the fundamental definitions of geopolitics and discusses two major strands of criticism directed at this research approach. The first critique, aimed at classical geopolitics, concerns the charge of geographical determinism. The second critique arises from neoclassical attempts to reinterpret geopolitics by incorporating non-geographical variables, which raises the question of whether such explanations can still be legitimately described as ‘geopolitical’. The second part of the review formulates the core research problem, which is presented as a reinterpretation of the dilemma proposed by Stefano Guzzini. Fundamental Characteristics of the Geopolitical Approach Geopolitics constitutes a research approach within the field of IR theory, focusing on the analysis of how spatial conditions influence the behavior of political agents. For instance, Saul Cohen defines geopolitics as the study ‘[o]f the interrelationship of the body politic and nature’ -- 3 of 31 -- 4 (Cohen, 1982: 228), while Stefano Guzzini conceptualizes it as the ‘[e]mpirical relationship between geographic factors and state policies’ (Guzzini, 2012: 13). The primary objective of geopolitics is to explain how geographical factors – such as location, climate, topography, borders, access to the sea, natural resources, or demographic structure – influence states’ interests, security strategies, and capacities (Gottman, 1951; Kelly et al., 2020; Owens, 2015). A defining feature of geopolitics is its treatment of geographical space as an objective, material factor that not only conditions but also constrains political behavior. Geography is not merely a backdrop to political events; it constitutes an integral element of their very structure. Scholars such as Mark Bassin (Murphy et al., 2004) and Colin S. Gray (2012) emphasize that all political behavior is spatially embedded, implying the necessity of accounting for geographical constraints in the analysis of international processes. Geopolitics does not simply acknowledge the importance of space – it identifies where and how this importance materializes, thereby enabling the delineation of strategic options and the formulation of effective political strategies. Harvey Starr adds to this perspective by noting that the spatial environment establishes the parameters of political choice, shaping which decisions are more likely, feasible, or rational. In this way, geographical space informs the calculations of decision-makers by narrowing the range of plausible spectrum of behavior in international politics (Starr, 1991, 2013). Two Critiques of Geopolitics in IR Theory Classical geopolitics, understood as an approach centered on the deterministic influence of geographical space on political behavior, has been subject to serious theoretical objections. One of the most frequently cited critiques comes from Benno Teschke, who asserted that ‘[g]eographical factors do not directly impact on politics’, since natural conditions only exert political influence within a specific international order (Teschke, 2006: 333, see also Kristof, 1960). This concise claim was further developed by Daniel Deudney, who argued that ‘[h]umans are not mechanical or natural automatons, but rather engage in consciously purposive practical activities’ (Deudney, 2000). As Jason Dittmer observed, geopolitics has thus long grappled with the fundamental accusation of geographical determinism – the assumption that the ‘[d]ynamics of international relations as following from the permanence of landscape and climate’ (Dittmer, 2014: 385). In his view, ‘[u]nderstanding the strategic importance or limitations of a given territory was understood as advantaging one state in their competition with others’ (Dittmer, 2014: 385). -- 4 of 31 -- 5 It appears that geopolitical scholars have not developed a consistent position on determinism. Rather than offering a clear defense of it, they have sought to reframe the issue, emphasizing that geography constitutes only one among many factors shaping political reality. An illustrative example is Colin S. Gray’s position, who held that ‘[p]hysical geography imposes a physical discipline upon the manifestations of the human will in behavior’, while simultaneously affirming that ‘[i]s the will that is in the driving seat’ (Gray, 2012: 297). Gray thus attempted to articulate a middle-ground position, which he described as ‘[t]he need to steer a prudent course between the hazard of determinism (“geography is destiny”), and the scarcely less misleading siren call of constructivism (“destiny is what we choose to make it”)’ (Gray, 2012: 299). In his view, geography both enables and constrains political behavior, depending on how effectively it is harnessed by human agency (Gray, 1996). A similar ambivalence can be found in the writings of Jakub Grygiel, who considered geopolitics to be ‘[a]n objective reality, independent of state wishes and interests’ (Grygiel, 2006: 24). In his view, ‘[a] foreign policy that does not reflect the underlying geopolitics cannot increase or maintain the power of a state’ (Grygiel, 2006: 36), as leaders may ‘[f]ail to read geopolitics and geopolitical change correctly and thus do not formulate and implement an appropriate response to changes’ (Grygiel, 2006: 24). Grygiel thus acknowledges determinist implications, suggesting that the failure to align foreign policy with geopolitical constraints results in inevitable strategic decline. At the same time, however, he notes that ‘[g]eopolitics does not determine foreign policy. Geopolitics simply limits the spectrum of strategic options available to a state’ (Grygiel, 2006: 36), and that ‘[w]e cannot predict the course of a state’s foreign policy by simply looking at the geopolitical variables’ (Grygiel, 2006: 36). This tension leads him to conclude that ‘[i]f geography must have explanatory power, it has to allow a degree of variation’ (Grygiel, 2006: 21), since ‘[c]onstant reality … cannot explain the often dramatic changes in foreign policy’ (Grygiel, 2006: 21). This internal ambiguity was already noted in the 1960s by Ladis Kristof. In his interpretation, determinism does not imply that the environment overrides human will, but rather that it ‘[e]liminates, under certain circumstances at least, all possible alternative courses of action’ (Kristof, 1960: 26). Kristof identifies a fundamental tension between acknowledging geography as a persistent factor and maintaining the belief in human capacity to transform the environment. In his view, resolving this dilemma requires accepting the inevitable embeddedness of human agency in geographical space, while also recognizing the capacity to modify that space within the bounds of technological capabilities. -- 5 of 31 -- 6 As a result, the classical geopolitics underwent an evolution toward more nuanced formulations, now commonly referred to as neoclassical geopolitics. This approach increasingly emphasized the interaction between geographical factors and elements specific to the social world. Technological variables, in particular, came to occupy a prominent position, being treated as intervening factors capable of modifying the influence of physical space in IR (e.g. Flint, 2022; Scholvin, 2016a, 2016b). It is precisely the ambiguity surrounding the role of geography that became the subject of a second critique – directed specifically at the neoclassical variant. As Mark Bassin aptly noted, ‘[i]t is not enough merely to insist that geography and space 'matter' to the political process. The real challenge … is to explain in meaningful detail how it matters’ (Murphy et al., 2004: 626). This difficulty becomes particularly pronounced when explanatory mechanisms refer to variables that are ontologically distinct from geographical ones. This issue was also recognized by Stefano Guzzini. In his view, rejecting the primacy of geography leads to a theoretical dilemma: either geopolitics is a specific approach requiring a clearly defined explanatory role – an expectation that becomes impossible if its determinist character is simultaneously disavowed – or it is not a specific approach at all, but rather a general form of materialism, in which case its use becomes redundant. In the latter case, one would need to justify why, despite its lack of uniqueness, geopolitics continues to be employed – something that is rarely done. When neoclassical geopolitics ‘[a]llows for a multiplicity of equal explanatory factors, or inflates the definition of geography sufficiently to include everything from historical lessons to state forms, then it is redundant, since it loses both a specific explanatory added value as compared to already existing approaches and, indeed, its geographical identity’ (Guzzini, 2012: 40). As Guzzini emphasizes, the introduction of new variables ultimately leaves ‘[t]he ambivalence about the actual role of geographical factors untouched’ (Guzzini, 2012: 38). In other words, for neoclassical geopolitics to avoid cognitive redundancy, it would need to demonstrate the primacy of a particular form of geographical knowledge. However, as Guzzini points out, ‘[g]eography or space generally matters is really not the point for which one would need ‘geopolitics’; it is consistent with almost all IR theories’ (Guzzini, 2012: 19). For geopolitics to retain its relevance, it would thus need to gravitate toward recognizing geography as a determining factor. Yet, as Guzzini rightly observes, neoclassical geopolitics offers no theoretical justification for such a position, as it consistently rejects geographical determinism. -- 6 of 31 -- 7 Guzzini therefore formulates the dilemma facing geopolitics as a research approach: [i]f geopolitics comes with a defence of (material) determinism, it is not credible; if it renounces such a defence, this undermines the very basis of its approach and identity’ (Guzzini, 2012: 35- 36). In his formulation, geopolitics appears to be ‘[e]ither distinct and determinate but wrong or indistinct, indeterminate and correct but trivial’ (Guzzini, 2017: 428) This dilemma should be regarded as one of the most serious theoretical challenges confronting geopolitics. When coupled with Christopher Fettweis’s critique of geopolitics as lacking scientific rigor (Fettweis, 2015), the inability to define a clear ontological foundation calls into question geopolitics’ place within the system of knowledge and the justification for its continued existence in IR theory. The Reformulation of Guzzini’s Problem The issue with Stefano Guzzini’s argument lies not in the substance of his critique, but in the form in which it was presented. While both variants of geopolitics are deemed inadequate within contemporary IR theory, Guzzini rejects each on different grounds. In the case of classical geopolitics, he relies on empirical reasoning, since it is not possible to establish the truth or falsity of a theoretical explanation on purely conceptual grounds. Challenging classical geopolitics thus requires the presumption of empirical evidence that leads to its falsification. By contrast, his critique of neoclassical geopolitics is entirely theoretical. Guzzini deems its explanations banal and, therefore, less original than competing approaches such as realism, which address similar research problems (Fettweis, 2015; Guzzini, 2017). However, a consistent and comparable evaluation of both variants necessitates the application of a uniform set of analytical criteria. The aim of this article is to conduct such an assessment within a theoretical framework, with particular attention to the criterion of theoretical coherence. Specifically, it asks whether – despite the evolution from classical to neoclassical formulations – geopolitical explanations retain coherence, understood as both logical consistency and the completeness of mechanisms, based on the presence of inferential connections among all components of the explanation (BonJour, 1985: 93-101). It is important to clarify that theoretical coherence is not the sole criterion for evaluating the value of a scientific explanation. It neither determines the truth of an explanation nor replaces other criteria, such as empirical adequacy (e.g. van Fraassen, 1980), heuristic value (e.g. Lakatos, 1978), or the capacity to generate testable hypotheses (Popper, 2002). Nevertheless, coherence constitutes a necessary precondition for an explanation to qualify as academic at all. -- 7 of 31 -- 8 This is because the aforementioned criteria can only be applied once an explanation has been shown to be internally coherent. Put differently, a lack of coherence precludes not only the generation of hypotheses but also their empirical testing and the attribution of any heuristic value to the approach (Bartelborth, 1999; Fetzer, 2000; Kuhn, 1970; Popper, 2002). Accordingly, theoretical coherence emerges as a necessary condition for assessing whether geopolitics contributes any academic value to IR theory. Should the analysis demonstrate that coherence has systematically declined as geopolitics has evolved toward its neoclassical form, Guzzini’s argument for abandoning geopolitics as a distinct research approach would appear justified. Conversely, if coherence has been retained, further empirical investigation would be required to assess the continued utility of geopolitical reasoning in IR. To carry out such an analysis, one must undertake the task that Guzzini himself did not: to examine how the reconceptualization of geography’s role, along with the inclusion of non- geographical variables, has affected the coherence of geopolitical explanations. The framework through which this analysis will be pursued is presented in the next section. Toward a Framework for Evaluating Coherence in Geopolitical Explanations This article assumes that the coherence of geopolitical explanations can be evaluated along two complementary dimensions: a quantitative one (associated with the principle of Ockham’s razor) and a qualitative one. Taken together, these dimensions form an analytical tool for estimating the degree of coherence of a given geopolitical explanation. This subsection is divided into three parts. The first discusses the quantitative dimension of coherence – its advantages as well as its limitations, which point to the need for supplementing it with a qualitative component. The second part focuses specifically on the qualitative aspect of coherence. The third part outlines how both dimensions will be applied in the subsequent stages of the analysis. Quantitative Dimension of Coherence According to the quantitative understanding of coherence, explanations should ideally operate with the minimal number of concepts, assumptions, and mechanisms. This principle, known in the philosophy of science as Ockham’s razor, serves as a heuristic rule for selecting among competing explanations. When confronted with multiple explanations accounting for the same phenomenon, the preferred explanation is the one that introduces the fewest assumptions and/or -- 8 of 31 -- 9 concepts, while still offering a comprehensive account of the phenomenon under investigation (e.g. Sober, 1975, 2015; Walsh, 1979). Applying this principle to the evaluation of scientific coherence carries important implications. The more complex an explanation becomes, the more likely it is to contain internal inconsistencies. As additional elements are introduced, the number of explanatory mechanisms increases, which in turn raises the likelihood of incoherence within the explanation. However, it must be emphasized that the relationship between simplicity and coherence is not direct, which constitutes the first limitation of this approach. The risk of incoherence arises from the number of explanatory mechanisms, not from the sheer number of variables involved. In other words, an increase in the number of variables does not necessarily weaken the coherence of an explanation, even if it increases complexity, provided the mechanisms are structured in a way that preserves internal consistency. Accordingly, Ockham’s razor should not be used as an arbitrary standard for eliminating complex explanations, but rather as a preliminary filtering tool. In this way, the quantitative dimension of coherence helps identify those explanations whose complexity necessitates deeper analysis through the qualitative dimension discussed below. A second limitation pertains to the specific nature of IR scholarship, which is inherently complex and characterized by multicausality. Explanations in this field typically require the inclusion of multiple interrelated factors that, taken together, form an indivisible explanatory whole. As Milja Kurki notes, research in IR must remain sensitive to the fact that causes are often complex and indirect, and that scholars should therefore embrace plural forms of causal explanation (Kurki, 2008). Colin Wight similarly argues that causation in IR must be understood in terms of the properties and interactions of multiple entities (Wight, 2006). These observations are particularly relevant in the critiques of geopolitics, especially in its neoclassical variant. They suggest that the mere inclusion of multiple variables in an explanatory model should not automatically lead to dismissing such a model as vague, trivial, or incoherent, as Guzzini has implied. What matters is not the number of variables, but how they are integrated into the explanatory structure. From this discussion, it follows that evaluating the coherence of an explanation based solely on simplicity – that is, by applying only the quantitative dimension – would be analytically insufficient. In the case of multicausal phenomena, such an approach risks discarding valid but -- 9 of 31 -- 10 complex explanations in favor of simplified but incomplete models. The analysis of coherence in geopolitical explanations should therefore go beyond the mere count of variables and include an assessment of the internal consistency of explanatory mechanisms – a task addressed in the following section. Qualitative Dimension of Coherence Building on the arguments outlined earlier, coherence requires consideration in both its quantitative and qualitative dimensions. This means that an assessment of coherence should take into account not only the number of elements that constitute a given explanation, but also the way in which those elements are interrelated. The structure of an explanation should be both logically consistent and functionally integrated. As Paul Thagard observes, an explanation should be internally consistent: its components should not duplicate one another’s functions, and together they should form a coherent structure. From this perspective, an explanation whose mechanisms are overlapping, redundant, or contradictory can be deemed incoherent (Thagard, 1989). A similar view is expressed by Peter Godfrey-Smith, who argues that coherence may be evaluated by analyzing the ‘architecture’ of an explanation – namely whether it is conceptually transparent, logically structured, and capable of conveying causal relationships without requiring resolution of internal contradictions (Godfrey-Smith, 2003: 190-200). Michael Strevens likewise treats coherence as an epistemic virtue conducive to effective knowledge production. He argues that coherence facilitates the testing of explanations as well as their comparison. In other words, coherence has long-term implications not only for theory development but also for empirical inquiry (Strevens, 2004). Within this framework, the qualitative dimension of coherence refers not to the number of variables used in an explanation, but rather to its internal structural integrity. Qualitative coherence presupposes that each mechanism addresses a distinct aspect of the phenomenon and does so comprehensively, without functionally overlapping with other mechanisms. This prevents contradictions between mechanisms and ensures that the explanation as a whole constitutes a logically consistent and transparent system that is readily interpretable (Bechtel & Richardson, 2000; Machamer et al., 2000). This logic leads to a somewhat tautological yet substantively important conclusion for the purposes of this article: the greater the internal coherence of the mechanisms within a given -- 10 of 31 -- 11 explanation, the more the explanation satisfies the criterion of coherence. In the case of competing explanations, preference should thus be given to the one whose mechanisms demonstrate the highest degree of internal consistency (Kitcher, 1989; Thagard, 1992). A key advantage of the qualitative dimension is its universality. It enables comparisons across different types of explanations, regardless of the number of variables involved. By focusing on explanatory mechanisms and their mutual coherence, this approach allows for the evaluation of both classical and neoclassical geopolitics, as each relies on specific mechanisms to generate theoretical conclusions. In this context, it may be argued that the task Stefano Guzzini left unfulfilled in his formulation of the geopolitical dilemma can be addressed by applying the qualitative dimension of coherence. It is this dimension that enables a precise assessment of whether, and to what extent, geopolitical explanations are logically structured and epistemically useful. Analytical Use of the Two Dimensions of Coherence In order to assess the extent to which geopolitics meets the criteria of coherence previously defined, it is necessary to conduct a theoretical analysis of existing geopolitical explanations. Accordingly, the following subsection reconstructs the evolution of claims formulated by proponents of classical and neoclassical geopolitics who have attempted to construct explanations of political phenomena1. The analysis will proceed in two stages. In the first stage, three types of geopolitical explanation will be identified. These types differ both in how they conceptualize the role of geography and in the directionality of causal mechanisms linking variables. In the second stage, each type will be evaluated in terms of coherence – first with respect to the quantitative dimension, understood as the number of variables introduced into the explanatory mechanisms. For those types that incorporate more than one variable, the qualitative dimension will also be analyzed, focusing on the nature of the relationships among variables and the degree to which they are integrated into a coherent explanatory framework. Following the individual discussion of each of the types, a comparative analysis will be carried out across three issues. First, the type that most fully satisfies the criteria of quantitative coherence will be identified. Second, a corresponding evaluation will be conducted with regard 1 This review omits geopolitically inclined practitioners who did not develop systematic academic explanations. Consequently, it does not include the works of George Kennan, William Bullitt, Henry Kissinger, or Zbigniew Brzezinski. -- 11 of 31 -- 12 to qualitative coherence. Finally, based on a synthetic assessment of both dimensions, the type exhibiting the highest overall level of coherence will be determined. The significance of this comparative analysis will be discussed in the conclusions. Three Types of Geopolitical Explanations This subsection presents three types of geopolitical explanation, classified according to the role assigned to geography within the structure of explanatory mechanisms – and, in the case of the second and third types, also the role of non-geographical factors. It should be emphasized that this typology is analytical rather than chronological in nature. First Type of Geopolitical Explanation The analysis2 begins with a discussion of the first type of geopolitical explanation, commonly referred to as classical geopolitics3. This subsection examines the contributions of Friedrich Ratzel, Rudolf Kjellén, Halford Mackinder, and Nicholas Spykman. Friedrich Ratzel, widely regarded as the founder of geopolitics as a field of inquiry, is the most prominent representative of this type. Ratzel conceptualized the state as an organism rooted in geographical space, with its political actions shaped by the influence of the natural environment on society. His formulation of geographic laws focused on two key aspects: space and location. The former determined the behavior of groups settled in a given territory, while the latter conferred a specific character upon that space (e.g. Ratzel, 1897; see also Farinelli, 2000). Ratzel’s ideas were further developed by Rudolf Kjellén, who emphasized even more strongly the determinative role of geography. He conceived geopolitics as a ‘science of the state’, in which the natural environment provided the structural conditions for realizing ‘the inexorable laws of progress’, understood as a drive toward expansion and territorial acquisition (Kjellén, 1924; see also Cohen, 2015). Together, Ratzel and Kjellén laid the theoretical foundations of classical geopolitics, in which geography functions as the sole and unidirectional explanatory factor of political phenomena. Their concepts assume that the dynamics of state behavior follow directly from the determining properties of geographical space. 2 Karl Haushofer has been excluded from this type, following Saul Cohen’s assessment that his work may be better understood as a creative compilation and continuation of Rudolf Kjellén’s ideas (Cohen, 2015: 20). 3 The terms 'first type of geopolitical explanation' and 'classical geopolitics' will be used interchangeably. -- 12 of 31 -- 13 A similar perspective was adopted by Halford Mackinder, although his approach was more regional in focus. In contrast to the abstract formulations of Ratzel and Kjellén, Mackinder concentrated on the geopolitical significance of specific territorial regions. Central to his theory was the so-called ‘pivot area’ – the interior of Eurasia, including Siberia – which he viewed as the axis of global politics. According to Mackinder, any continental power able to control this region (e.g., Russia, Germany, or China) would gain a strategic advantage over maritime powers, thereby enabling it to achieve global dominance. He developed this idea in ‘Democratic Ideals and Reality’, where he articulated his well-known maxim: ‘[w]ho rules East Europe commands the Heartland, who rules the Heartland commands the World-Island, who rules the World-Island commands the world’ (Mackinder, 1904, 1919; Mayhew, 2000). Nicholas Spykman, the architect of Rimland theory, adopted a similar mode of reasoning. Unlike Mackinder, who emphasized the geopolitical centrality of the Eurasian interior, Spykman highlighted the strategic importance of the coastal zones of Eurasia (Western Europe, the Middle East, and South and East Asia). He argued that a state dominating the Rimland would be able to control critical communication routes – from the Mediterranean to the Indian and Pacific Oceans – and thereby exert influence over the entire Eurasian continent. Paraphrasing Mackinder, Spykman famously asserted: ‘[w]ho controls the Rimland rules Eurasia, who rules Eurasia controls the destinies of the world’ (e.g. Spykman, 2011). The theoretical claims advanced by Ratzel, Kjellén, Mackinder, and Spykman constitute the core of classical geopolitics, which is grounded in the assumption that geographical factors operate deterministically. In this framework, political behavior is shaped either by the general influence of geographic space on society (as in Ratzel and Kjellén) or by the properties of specific strategic regions, which determine a state’s capacity to project power over surrounding areas (as in Mackinder and Spykman). Figure 1. The causal mechanism of the first type of geopolitical explanation. Second Type of Geopolitical Explanation -- 13 of 31 -- 14 In the second type, one can identify the presence of a positive feedback between geography and political behavior. This type includes the concepts advanced by Alfred Thayer Mahan and George Renner, both of whom derived their arguments directly from Heartland theory. Alfred Mahan, like Halford Mackinder, viewed Russia as the dominant land power; however, unlike Mackinder, he argued that Russia’s geographical position placed it at a disadvantage, since maritime movement, in his view, surpassed land-based movement in strategic importance. Mahan pointed to the zone between the 30th and 40th parallels of Asia as the key arena of global power competition, where the Russian land empire clashed with British naval power. He argued that the global balance of power could be preserved through an Anglo-American alliance supported by a network of military bases encircling Eurasia (Mahan, 2010). Up to this point, Mahan’s argument remains representative of the first type of geopolitical explanation. However, the innovative nature of his theory becomes apparent when he identifies technology – specifically naval forces – as the key factor enabling spatial influence. According to Mahan, a state's advantage in the struggle for global dominance stems not solely from geographical position, but from the possession of a well-developed naval fleet (Mahan, 2010). Geography does not operate autonomously; its influence must be reinforced by military capabilities that ensure the durability and effectiveness of power projection. Moreover, technology transforms the meaning of geographical space. Without naval power, Mahan’s critical zone would not constitute a contested area. In other words, the geographical component becomes conditionally relevant – its function is dependent on technological support. George Renner also built upon Heartland theory, proposing an expanded conception of strategic space. He argued that air corridors had linked the Eurasian Heartland with a so-called ‘Second Heartland’ located in the Anglo-American region via the Arctic ice fields. As a result, Renner conceptualized an enlarged Heartland extending across the entire Northern Hemisphere. According to his analysis, this region became not only the center of global power, but also acquired strategic superiority due to a network of air, sea, and land routes passing through the polar regions. Within this framework, the Arctic was recognized as a crucial geostrategic point, the control of which would enable global dominance (Renner, 1942, 1944). As in Mahan’s case, Renner’s explanation contains a significant technological component. Geographical dispositions do not operate independently but are activated by technological advancements. Importantly, the Arctic gains strategic significance only when states acquire the technological capabilities necessary to penetrate and exploit it. This is analogous to Mahan’s -- 14 of 31 -- 15 key zone: without air power, the Arctic would remain a relatively insignificant geographical area. In both instances, explanations recognize geography as a relevant causal factor, yet conceptualize technological advancement as an intervening variable that conditions the extent to which geographical imperatives can be translated into state behavior. Consequently, geography loses its deterministic character in shaping political behavior. Furthermore, the strategic significance of a given geographical area emerges only when states reach a sufficient level of technological development to effectively explore and utilize it. This gives rise to a positive feedback between geography and political behavior, in which technology functions as a mediating factor determining both the scope and intensity of geography’s influence on state behavior. This dynamic can be represented as follows: Figure 2. The causal mechanism of the second type of geopolitical explanation. Despite the complications introduced by incorporating technology as an intervening variable, it is important to note that the mechanisms associated with the second type remain internally coherent. These explanations involve two distinct mechanisms. The first accounts for the activation of dispositions associated with a given geographical area by incorporating the influence of a particular factor (naval power or air power) on state behavior. The second explains how a state behavior, enabled by specific technological capacities, assign strategic significance to a particular geographic space. Each mechanism thus corresponds to a different aspect of the phenomenon under examination. The mechanisms are non-contradictory and may coherently function within a single explanatory framework. -- 15 of 31 -- 16 In summary, the second type of geopolitical explanation continues to treat geography as an analytically significant factor; however, its influence is no longer deterministic. Rather, it is conditioned by technological development. In both cases, it is technology that operates as the intervening variable, enabling the activation of a region’s geographic potential and imbuing it with strategic significance. This results in a feedback mechanism in which technology mediates the relationship between geographic space and state behavior. Despite the added complexity introduced by the second explanatory factor, the mechanisms remain coherent: the first describes how technology shapes political behavior within a given geographical context, while the second explains how such behavior, facilitated by technological innovation, confers meaning upon that space. Third Type of Geopoltiical Explanation: General Characteristics The third type is characterized by the incorporation of numerous variables, rendering geopolitics a distinctly multicausal approach. Explanations of this kind are commonly referred to as neoclassical geopolitics4. Although each representative of this type offers a distinct interpretation and independently determines which factors are deemed relevant for explaining political phenomena, an initial collective overview will be provided. Special attention will then be devoted to the concept formulated by Daniel Deudney, which may be regarded as a representative example of this type. The first, and somewhat unexpected, assumption of this type is the claim that ‘[g]eopolitics is not a science of foreign policy, nor state behaviour. It only deals with geographically given obstacles to, and opportunities for, successful policies’ (Scholvin, 2016b: 13). In the view of third type theorists, geopolitics does not perform an explanatory function; it does not explain phenomena but merely describes the spatial conditions that constrain or enable political action. As such, it loses its explanatory power. This redefinition also necessitates a reformulation of geography’s role. Sören Scholvin conceptualizes geographic conditions merely as a set of constraints and opportunities that become relevant only within a specific context (Scholvin, 2016a). Geography ceases to function as a general explanatory variable and instead becomes the physical ‘playing field’ for strategies designed and implemented by international agents (Gray, 1999). In this sense, geography may activate the significance of certain political behaviors but does not determine them. This interpretive shift was initiated by Harvey Starr, who emphasized that ‘[h]ow humans see the 4 The terms third type of geopolitical explanation' and 'neoclassical geopolitics' will be used interchangeably. -- 16 of 31 -- 17 environment – the content of their image of that environmental context – is the central matter of importance in choice’ (Starr, 1991: 4). The relationship between agents and their environment thus depends not on objective spatial conditions, but on how those conditions are interpreted by decision-makers, who alone shape political behavior. This conceptualization constitutes a fundamental departure from the first – and partially from the second – types of explanation. By removing geography’s agency (through its mediation by human perception) and thus challenging the explanatory validity of earlier geopolitical models, this type effectively questions the legitimacy of those explanations and, by extension, the coherence of geopolitics itself. Consequently, in order to preserve the relevance of a reformed geopolitics as a research approach, it became necessary to construct a new explanatory structure. This was accomplished in minimalist terms, with the primary aim of geopolitics now being to examine the interplay between geographical and non-geographical conditions. As a result, the third defining element of neoclassical geopolitics became the specification of a set of factors jointly constituting geopolitical explanations. Alongside geography – treated instrumentally – Scholvin identified technology as a key variable in explaining state behawior (Scholvin, 2016a). However, it should be noted that the role of technology was not entirely new, as it had already featured in the second type. Therefore, further attempts to expand the paradigm necessitated an even broader inclusion of non-geographical variables. In this context, Colin S. Gray highlighted elements such as society, culture, ethics, organizational factors (including defense planning and military structure), information, and intelligence activities (Gray, 1999). Phil Kelly incorporated into his analyses the personalities of decision-makers, historical experience, tradition (Kelly et al., 2020), technology (Kelly, 2006), and the level of economic development (Kelly, 2019). Jakub Grygiel emphasized the significance of technology, production, and systems of communication (Grygiel, 2006). As a result, neoclassical geopolitical explanations evolved into hybrid explanations, in which the geographic component became merely one among many variables, often a marginal one. Third Type of Geopolitical Explanation: Daniel Deudney’s Model The fundamental problem with existing formulations of the third type lies in the fact that none of the previously discussed authors has developed a coherent explanation that systematically addresses (1) how the role of geography has evolved, and (2) which factors have assumed its -- 17 of 31 -- 18 explanatory function. The most comprehensive – though, as will be shown, still insufficient – attempt to capture both of these aspects has been made by Daniel Deudney. It is noteworthy that Deudney himself does not describe his approach as geopolitical, instead referring to it as ‘[s]tructural-functional security materialism’ (Deudney, 1997: 93). Only later in his argument does he introduce the concept of geography, treating it as one component of a broader ‘material environment’ in IR. According to Deudney, this environment ‘[i]s composed of the interaction of geography and technology, which are for human practical purposes effectively revelations of natural possibility beyond human control’ (Deudney, 1997: 100). It is this material environment that generates the constraints and opportunities facing human institutions in their pursuit of the long-term and fundamental objective of ensuring physical security. The starting point of Deudney’s model thus echoes a central premise found in the second type – the co-constitution of geography by technology. In the second type, the political relevance of geographic space was activated by technological factors; that is, geographic location acquired strategic value only insofar as technological means allowed it to be incorporated into state strategy. In the third type, however, this relationship is significantly intensified. Technology not only activates but also transforms geography. As Deudney himself notes, the realization that ‘[t]echnology constitutes a change in nature as it presents itself for human purposes … undermined the basic thrust of naturalism that a reality exterior to human intent and design intrudes into human affairs and decisively shapes them’ (Deudney, 1997: 105). In this context, Deudney explicitly refers to the notion of technological determinism, which, in his view, ought to replace geographic determinism. By claiming that ‘[n]ature ... changes via technological development’ (Deudney, 1997: 105), Deudney unequivocally denies geography the status of a causal factor – something that had not occurred in any of the preceding types of geopolitical explanation. In doing so, he aligns with the direction suggested by Sören Scholvin, who argued for reducing geography to a merely descriptive role, devoid of explanatory significance. Other aspects concerning the role of geography and the coherence of Deudney’s model will be addressed following the analysis of the internal structure of his explanation, which takes the following form: -- 18 of 31 -- 19 Figure 3. Daniel Deudney’s model of neoclassical geopolitics as a representative case of the third type of geopolitical explanation. Source: Deudney, 1997: 101. An analysis of the proposed explanation allows for the identification of three key issues characteristic of neoclassical geopolitics from the perspective of the qualitative dimension of coherence. First, the model enables one to trace the influence of geography on political phenomena, while simultaneously demonstrating that its significance is more limited than in the second type. This is due to the fact that the Security Order is shaped by three components: the Material Context (which, as Deudney emphasizes, is ‘only partially given by nature’ – emphasis added), Human Nature, and Problem-Solving Agency. Geography, as part of the first of these categories, becomes merely one among many factors with equal causal status – effectively further weakening its explanatory role. Moreover, the impact of geography is further reduced within the Material Context category itself. The transition from Material Context 1 to Material Context 2 illustrates the process outlined previously, by which geography is transformed by technological factors. As a result, technology gradually marginalizes geography as a structural determinant. This implies that Material Context 2 is even less shaped by geographical conditions than its earlier version, leading to a systematic erosion of geography’s role in subsequent iterations of the model – Security Order 3, 4, and so forth – until its influence effectively disappears altogether. Second, Deudney’s model raises concerns regarding the internal consistency of its mechanisms. The primary issue stems from the absence of any definition for its key analytical categories, namely Human Nature and Problem-Solving Agency. One may assume that these categories -- 19 of 31 -- 20 encompass variables relating to individual behavior and its products – elements also identified by other representatives of neoclassical geopolitics. However, the lack of an enumerated classification of such variables, along with the absence of specified relationships between them, prevents the model from being operationalized or its structure properly assessed. Only through the precise definition of variables within each category, and a clear delineation of their interrelations, would a coherent theoretical analysis be feasible. Assuming, however, the most common definitions – Human Nature as ‘[t]he natural ways of behaving that most people share’ (Cambridge Dictionary), and Problem-Solving Agency as the ability of finding solutions to problems encountered in life (Robertson, 2017) – one can identify a significant inconsistency in the construction of explanation. If Problem-Solving Agency entails the ability to transform external stimuli into intentional political behaviors, it should function as an intervening variable – one that enables the translation of Human Nature and the Material Context into a specific Security Order. Yet, in Deudney’s model, Problem-Solving Agency is treated as an independent, co-equal category, which is analytically unjustified. Agency does not exist independently of a concrete problem context; rather, it constitutes a latent potential that becomes activated under specific external constraints (e.g. Acharya, 2018; Hudson, 2005; Kuus, 2008). For this reason, the model’s principal flaw lies in its logical inconsistency: Problem-Solving Agency should not be conceptualized as an independent causal factor but rather as a mediating mechanism between the Material Context and Human Nature, enabling their transformation into a particular political outcome. The causal logic that should structure Deudney’s model may be illustrated as follows: -- 20 of 31 -- 21 Figure 4. A reformulated Deudney’s model partially consistent with the qualitative dimension of coherence. Nevertheless, even if Deudney’s explanation were to be reformulated in line with the proposal outlined in Figure 4, a third issue emerges regarding the delineation of the scope of specific mechanisms. Although the causal relationships between the Material Context and the Security Order, as well as between the Human Nature and the Security Order, are formally distinguished, their substantive content remains undefined. In other words, Deudney does not specify which particular aspects of political behavior are determined by the Material Context and the Human Nature, nor how these two components influence the formation of the Security Order. This shortcoming originates in the problem previously identified in the analysis of the qualitative dimension of coherence. The absence of clearly defined and operationalized variables within the categories of Material Context and Human Nature renders them analytically unusable. As a result, Deudney’s model lacks analytical coherence: despite the formal separation of distinct mechanisms, it fails to offer insight into how these mechanisms function in practice. Ultimately, the model contributes little in terms of explanatory value, as it does not provide tools for either empirical verification or theoretical reconstruction of the processes it purports to explain. -- 21 of 31 -- 22 In sum, the third type of geopolitical explanation proves problematic in terms of both dimensions of coherence. This is particularly evident in the case of Daniel Deudney’s model, which redefines the role of geography. Having been stripped of its explanatory power, geography is reduced to one among many (unspecified) components of the Material Context – its influence increasingly displaced by technology, leading to the marginalization of geography in shaping the Security Order. While the model formally incorporates three causal factors – Material Context, Human Nature, and Problem-Solving Agency – its overall construction also lacks coherence. First, the absence of definitions and operationalization for these categories precludes the identification of specific variables or the elucidation of their mutual interactions. Second, Problem-Solving Agency is treated as an independent category within the model, even though, from a logical standpoint, it should function as an intervening variable, mediating the effects of the Human Nature and the Material Context. Third, the model fails to delineate the respective domains of each mechanism offering no clarification as to which dimensions of the Security Order are shaped by each causal factor. Consequently, the model delivers neither analytical nor explanatory utility, as it does not permit empirical testing or theoretical reconstruction of the political phenomena it seeks to explain. A Comparative Critique of Three Types of Geopolitical Explanations The analysis of the three types reveals a gradual reduction in the explanatory significance of geography in the analysis of political phenomena. In the first type, geography functions as the primary determinant of state behavior. In the second type, it is subordinated to technology, which acts as an intervening variable activating the strategic relevance of particular spaces and generating a positive feedback between geography and political behavior. In the third type, geography is reduced to a merely descriptive role, supplanted by other factors that effectively eliminate its explanatory value. As a result, the role of geography in neoclassical geopolitical explanations has been substantially diminished in favor of alternative variables. As exemplified by Deudney’s model, however, the scope and nature of these alternative variables remain unclear, as they are neither precisely defined nor properly operationalized. Type Role of geography Number of variables First Geographical determinism One (geography) Second Positive feedback between geography and political phenomena, technology as an intervening variable Two (geography, technology) -- 22 of 31 -- 23 Third The reduction of geography to a descriptive function, the loss of its explanatory power Many (undefined number of variables) Table 1. Overview of types of geopolitical explanation This section compares the three types of geopolitical explanation in terms of: (1) quantitative coherence; (2) qualitative coherence; and (3) which type best satisfies both dimensions of theoretical coherence. Geopolitics and the Quantitative Dimension of Coherence The theoretical analysis yields two key conclusions, one for each dimension. First, with regard to the quantitative dimension, there is a clear evolution in the number of explanatory factors across the three types of geopolitical explanation. In the first type, geography remains the sole variable; in the second, technology is introduced as an intervening variable, generating positive feedback between spatial conditions and state behavior. In the third type numerous non- geographical variables are incorporated, resulting in a multicausal framework. As a result, explanations of the third type meet the criteria of the quantitative dimension of coherence to the least extent, whereas classical geopolitical explanations meet them to the greatest extent. Geopolitics and the Qualitative Dimension of Coherence It is necessary to examine geopolitics also from the perspective of the qualitative dimension of coherence. In the case of geopolitical explanations formulated within the third type, a second issue arises: their explanatory mechanisms lack coherence. To demonstrate this, it is essential to focus on the key conclusions drawn from the analysis of the explanatory mechanisms in the second and third types5. Explanations of the second type involve two mechanisms. The first explains the activation of the dispositions of a given geographical area through the influence of a specific factor (naval power or airpower) on state behavior. The second explains how a state's behawior shaped by particular technological capabilities assigns strategic significance to a given geographical space. Each mechanism thus addresses a distinct aspect of the phenomenon in question. These mechanisms are therefore non-contradictory and can be integrated into a single explanation. 5 Given that the first type is coherent by virtue of its deterministic relationship between geography and state behavior, it will not be analyzed. -- 23 of 31 -- 24 In contrast, the coherence of explanatory mechanisms in third type of explanation raises serious concerns. The number of variables increases substantially, yet remains undefined, making their identification and systematization more difficult. Among proponents of the third type, there is no consensus regarding which variables should be incorporated into explanatory mechanisms. Two extreme positions can be identified: Sören Scholvin’s minimalist view, limited to the role of technology, and Colin Gray’s maximalist approach, which calls for the inclusion of variables such as society, culture, ethics, organization, information, and intelligence. This lack of a unified framework results in theoretical fragmentation, leading to underdefinition and a significant reduction in the coherence of explanations characteristic of neoclassical geopolitics. This problem is particularly evident in the explanation proposed by Daniel Deudney. Despite his aim to theoretically structure neoclassical geopolitical explanations, Deudney’s model fails to overcome the aforementioned issues. First, the model illustrates the systematic decline of geography’s explanatory role in the analysis of political phenomena – both through its reduction to just one of many variables within the Material Context, and through its processual marginalization via technological influence. The successive transformation of geography in each iteration of the model results in the progressive erosion of its impact, ultimately eliminating its relevance in shaping the Security Order. Second, the explanation proposed by Deudney raises serious concerns with respect to the qualitative dimension of coherence, primarily due to the imprecise definition of its key analytical categories: Human Nature and Problem-Solving Agency. The absence of operationalization and the failure to specify the interrelations between these categories prevent a consistent theoretical analysis. Moreover, assuming a standard understanding of both concepts, their placement within the model appears flawed: Problem-Solving Agency, understood as the capacity to transform external stimuli into political action, should serve as an intervening variable rather than as an independent and coequal component. As a result, the model suffers from internal logical inconsistency, as agency becomes detached from the context in which it is meant to be activated – significantly reducing its explanatory utility. The third major issue with Deudney’s explanation is functional in nature and concerns the lack of clear delimitation regarding the scope of each explanatory mechanism. Although causal relationships between Material Context and Security Order, as well as between Human Nature and Security Order, are formally indicated, the explanation does not specify which particular aspects of political behavior these categories encompass or how they influence the phenomena under analysis. This gap stems from the earlier failure to operationalize variables within both -- 24 of 31 -- 25 categories, which undermines their analytical applicability. Consequently, the model loses analytical coherence: despite its formal structure, it does not offer tools for empirical application or theoretical reconstruction of the processes it aims to explain, thereby diminishing its explanatory relevance. The progressive marginalization of geography, uncertainty regarding the number of explanatory variables, lack of clear definitions of core analytical categories, and the absence of specification concerning the scope of mechanisms all significantly hinder the conceptualization of the third type in a way that ensures coherence. As a result, neoclassical geopolitical explanations fail to meet the criteria of the qualitative dimension of coherence – whereas explanations of the first and second types do fulfill this criterion. Which Type of Geopolitical Explanation Demonstrates the Highest Degree of Coherence? The preceding analysis raises a fundamental question: which type of geopolitical explanation is the most coherent? This question can be addressed by drawing on the two dimensions of coherence discussed earlier. From the quantitative perspective, the greater the number of explanatory variables included in a given model, the smaller the relative share of geography as an explanatory factor becomes. In such cases, geography ceases to serve as a dominant explanatory principle and is reduced to one among many variables influencing political phenomena, thereby weakening the geopolitical character of the explanation. Consequently, the most coherent explanations are those of the first type, in which geography functions as the sole causal factor. Explanations of the second type are slightly less coherent due to the addition of technology as an intervening variable, although the overall structure of the explanation remains relatively consistent. The least coherent are explanations of the third type, where geography is further marginalized in favor of numerous undefined or vaguely specified variables. This observation is reinforced when considering the qualitative dimension. As demonstrated earlier, third type of geopolitical explanation suffer from internal inconsistency – both due to the undefined scope of included variables and to the unclear causal relationships between analytical categories. In contrast, second type of explanations, like those of the first type, exhibit internal consistency. The second type contain two distinct mechanisms, each explaining a different aspect of the phenomenon: one accounts for the role of technology in activating the geopolitical relevance of a given space, while the other explains how political actions assign strategic significance to that space. The first type, by comparison, rely on a single, unidirectional mechanism in which geography directly determines political behavior. Despite -- 25 of 31 -- 26 their simplicity and deterministic character, these explanations remain coherent insofar as they avoid internal contradictions and do not introduce an excess of variables that might undermine the logic of the explanation. If one accepts that a coherent geopolitical explanation should fulfill the requirements of both dimensions of coherence, then only first type of geopolitical explanations fully meet these criteria. As such, they may be considered the most strictly ‘geopolitical’ in nature. While the second type do satisfy the criterion of qualitative coherence, they fail to meet the demand for variable minimization, rendering them hybrid rather than strictly geopolitical. The third type, by contrast, lack coherence in both dimensions, thereby undermining their value as geopolitical explanations altogether. In sum, the more geopolitical reasoning has diverged from the classical, deterministic approach, the more it has lost its internal coherence and, with it, its justification for remaining a distinct analytical framework within IR theory. Conclusions The aim of this article was to conduct a theoretical analysis of classical and neoclassical geopolitical explanations through the lens of two dimensions of coherence: the quantitative and the qualitative. Drawing on Stefano Guzzini’s dilemma, according to which geopolitics ‘[i]s either distinct and determinate but wrong or indistinct, indeterminate and correct but trivial’(Guzzini, 2017: 428) – the article examined how the reconceptualization of the role of geography and the inclusion of non-geographical variables have affected the coherence of geopolitical explanations. For the purpose of the analysis, three types of geopolitical explanation were identified and evaluated. The analysis demonstrated that the further geopolitics has moved away from classical formulations, the less it has met the criterion of coherence. Only first type of explanations satisfy the requirements of both dimensions of coherence; the second type meet only the qualitative criterion, while the third type satisfy neither. This conclusion entails two significant implications. First, it concerns the status of geopolitics as a research approach. If the objective is to preserve geopolitics within the scope of IR, it must be emphasized that only classical geopolitics fulfills the definitional criteria necessary to qualify as fully geopolitical. It is the only geopolitical approach that meets both coherence criteria. However, this recognition should be viewed as diagnostic rather than prescriptive. Given the -- 26 of 31 -- 27 prevailing consensus among IR scholars that rejects determinism as a valid foundation for academic explanation, this effectively implies the exclusion of classical geopolitics – and thus geopolitics as a whole – as a research approach capable of offering academically legitimate explanations of IR. Second, the article’s proposal to apply two dimensions of coherence demonstrates its potential as an analytical tool. It may be used to assess the coherence of other theories or approaches within IR, particularly those, such as neoclassical realism, that have been criticized for lacking a consistent explanatory structure (e.g. Legro & Moravcsik, 1999). References Acharya, A. (2018). Constructing Global Order: Agency and Change in World Politics. Cambridge University Press. Bartelborth, T. (1999). Coherence and Explanations. Erkenntnis, 2–3, 209–224. Bechtel, W., & Richardson, R. (2000). Discovering Complexity: Decomposition and Localization as Strategies in Scientific Research. The MIT Press. BonJour, L. (1985). The Structure of Empirical Knowledge. Harvard University Press. Cohen, S. (1982). A new map of global geopolitical equilibrium: a developmental approach. Political Geography Quarterly, 1(3), 223–241. Cohen, S. (2015). Geopolitics: The Geography of International Relations. Rowman & Littlefield. Deudney, D. (2000). Geopolitics as Theory: Historical Security Materialism. European Journal of International Relations, 6(1), 77–107. Deudney, D. (1997). Geopolitics and Change. In M. Doyle & G. Ikenberry (Eds.), New Thinking in International Relations Theory (pp. 91–123). Routledge. Dittmer, J. (2014). Geopolitical assemblages and complexity. Progress in Human Geography, 38(3), 385–401. Farinelli, F. (2000). Friedrich Ratzel and the nature of (political) geography. Political Geography, 19(8), 943–955. Fettweis, C. (2015). On Heartlands and Chessboards: Classical Geopolitics, Then and Now. Orbis, 59(2), 233–248. -- 27 of 31 -- 28 Fetzer, J. (2000). Science, Explanation, and Rationality: Aspects of the Philosophy of Carl G. Hempel. Oxford University Press. Flint, C. (2022). Putting the “geo” into geopolitics: a heuristic framework and the example of Australian foreign policy. GeoJournal, 87(4), 2577–2592. Godfrey-Smith, P. (2003). Theory and Reality: An Introduction to the Philosophy of Science. The University of Chicago Press. Gottman, J. (1951). Geography and International Relations. World Politics, 3(2), 153–173. Gray, C. (1996). The Continued Primacy of Geography. Orbis, 40(2), 247–259. Gray, C. (1999). Inescapable geography. Journal of Strategic Studies, 22(2–3), 161–177. Gray, C. (2012). Geopolitics and Deterrence. Comparative Strategy, 31(4), 295–321. Grygiel, J. (2006). Great Powers and Geopolitical Change. The Johns Hopkins University Press. Guzzini, S. (Ed.). (2012). The Return of Geopolitics in Europe? Social Mechanisms and Foreign Policy Identity Crises. Cambridge University Press. Guzzini, S. (2017). Militarizing politics, essentializing identities: Interpretivist process tracing and the power of geopolitics. Cooperation and Conflict, 52(3), 423–445. Hudson, V. (2005). Foreign Policy Analysis: Actor-Specific Theory and the Ground of International Relations. Foreign Policy Analysis, 1, 1–30. Human Nature, Cambridge Dictionary, https://dictionary.cambridge.org/dictionary/english/human-nature Kelly, P. (2006). A Critique of Critical Geopolitics. Geopolitics, 11(1), 24–53. Kelly, P. (2019). Rescuing Classical Geopolitics. Geopolitics, History, and International Relations, 11(1), 41–58. Kelly, P., Barrera, G., & Jewett, J. (2020, October). Classical Geopolitics Continues under Attack. https://www.vision-gt.eu/wp-content/uploads/2020/10/AD_19_2020.pdf Kitcher, P. (1989). Explanatory Unification and the Causal Structure of the World. In P. Kitcher & W. Salmon (Eds.), Scientific Explanation (pp. 410–505). University of Minnesota Press. -- 28 of 31 -- 29 Kjellén, R. (1924). Der Staat Als Lebensform. Kurt Vowinckel Verlag. Kristof, L. (1960). The origins and evolution of geopolitics. Journal of Conflict Resolution, 4(1), 15–51. Kuhn, T. (1970). The Structure of Scientific Revolutions. University of Chicago Press. Kurki, M. (2008). Causation in International Relations: Reclaiming Causal Analysis. Cambridge University Press. Kuus, M. (2008). Professionals of geopolitics: Agency in international politics. Geography Compass, 2(6), 2062–2079. Lakatos, I. (1978). The methodology of scientific research programmes. Cambridge University Press. Legro, J., & Moravcsik, A. (1999). Is Anybody Still a Realist? International Security, 24(2), 5– 55. Machamer, P., Darden, L., & Craver, C. (2000). Thinking About Mechanisms. Philosophy of Science, 67(1), 1-25. Mackinder, H. (1904). The Geographical Pivot of History. The Geographical Journal, 23(4), 421–437. Mackinder, H. (1919). Democratic ideals and reality: a study in the politics of reconstruction. Constable and Company. Mahan, A. (2010). The Influence of Sea Power upon History, 1660-1783. Cambridge University Press. Mayhew, R. (2000). Halford Mackinder’s “new” political geography and the geographical tradition. Political Geography, 19(6), 771–791. Murphy, A., Bassin, M., Newman, D., Reuber, P., & Agnew, J. (2004). Is there a politics to geopolitics? Progress in Human Geography, 28(5), 619–640. Owens, M. (2015). In Defense of Classical Geopolitics. Orbis, 59(4), 463–478. Popper, K. (2002). The Logic of Scientific Discovery. Routledge. Ratzel, F. (1897). Politische Geographie. R. Oldenbourg Verlag. -- 29 of 31 -- 30 Renner, G. (1942). The Nature of Geographical Ideas. Teachers College Record: The Voice of Scholarship in Education, 43(8), 1–8. Renner, G. (1944). What the War Has Taught us About Geography. Journal of Geography, 43(9), 321–330. Robertson, S. (2017). Problem Solving: Perspectives from Cognition and Neuroscience. Routledge. Scholvin, S. (2016a). Geographical conditions and political outcomes. Comparative Strategy, 35(4), 274–283. Scholvin, S. (2016b). Geopolitics: An Overview of Concepts and Empirical Examples from International Relations. https://www.files.ethz.ch/isn/196701/wp91-Geopolitics.pdf Sober, E. (1975). Simplicity. Clarendon Press. Sober, E. (2015). Ockham’s Razors: A User’s Manual. Spykman, N. (2011). Heartland and Rimland. In R. Kasperson & J. Minghi (Eds.), The Structure of Political Geography (pp. 168–176). Routledge. Starr, H. (1991). Joining political and geographic perspectives: Geopolitics and international relations. International Interactions, 17(1), 1–9. Starr, H. (2013). On Geopolitics: Spaces and Places. International Studies Quarterly, 57(3), 433–439. Strevens, M. (2004). The Causal and Unification Approaches to Explanation Unified-Causally. Noûs, 38(1), 154–176. Teschke, B. (2006). Geopolitics. Historical Materialism, 14(1), 327–335. Thagard, P. (1989). Explanatory coherence. Behavioral and Brain Science, 12(3), 435–502. Thagard, P. (1992). Conceptual Revolutions. Princeton University Press. van Fraassen, B. (1980). The Scientific Image. Clarendon Press. Walsh, D. (1979). Occam’s Razor: A Principle of Intellectual Elegance. American Philosophical Quarterly, 16(3), 241–244. -- 30 of 31 -- 31 Wight, C. (2006). Agents, Structures, and International Relations: Politics as Ontology. Cambridge University Press. -- 31 of 31 --