Transcending the ‘End of War’ Debate: Toward a Mechanism-Centered View on the ‘War on War'

Abstract

Has war become obsolete? Some argue that macro-historical social processes are leading war into obsolescence, while for others that pattern is explained by the fact that war is a lingering potential outcome of international politics. I argue that both answers reveal a fundamental problem with the debate about the ‘end of war’. The focus on securing predictions about the end of war keeps analyses trapped in either a teleological or nomothetical linearity and overshadows the varied set of contingent mechanisms that allow for non-violent outcomes to prevail in contentious political episodes. The ‘War on War’ could benefit from analyses that asses shifting configurations of mechanisms in power politics.

Political violence; contentious politics; power politics; critical security studies

The Russian aggression against Ukraine – and its potential escalation into a major power conflict – has brought interstate war once again to the headlines of world politics. However, empirical evidence of a long-term decline in the frequency of interstate war in general¹ – and major power wars, in particular – has inspired a rich debate about the prospects of war becoming obsolete as a social phenomenon. Optimistic views about the prospects of interstate wars going extinct are based on the usual suspects of the liberal peace hypothesis – norms, democratic institutions, market institutions, and socio-economic interdependence ( DOYLE, 1983 ; GAT, 2006 ; MOUSSEAU, 2019 ; MUELLER, 1989 ; O'NEAL and RUSSET, 1999; OWEN, 1994 ; PINKER, 2011 ; RUSSET and O'NEAL, 2001). In contrast, others argue that the costs of war have risen to an unbearable point because the very possibility of war means that countries must continually create capacities and be prepared for it ( COPELAND, 2014 ; LEVY and THOMPSON, 2010) . To be sure, the harsh reality of world politics in the last couple of years has made this debate sound quite foreign. Nevertheless, in this article, I critically examine the debate on the ‘end of war’ to propose a mechanism-centered approach capable of reconciling some of this debate’s insights with the enduring and contingent nature of war.

Joshua Goldstein’s (2011) insightful framing of the decline of war as a ‘war on war’ reveals some crucial layers of the ‘end of war’ debate. The implicit analogy with other diffuse wars, such as the ‘war on drugs’ and ‘war on terror’, brings to the fore the teleology underlying the idea of the ‘end of war’. On the one hand, the analogy shows more explicitly that the idea of eventually extinguishing war in world politics is sustained by a normatively oriented social effort. On the other hand, this focus on the possibility of ending war reveals that centering the debate around ‘whether’ war is obsolete conflates the retrospective assessment of ‘why’ war has not been occurring with prospective speculation about its potential (non-) recurrence. I argue that such conflation is the reason why scholars are bewildered by the resurgence of interstate violence, a conflation that may also get in the way of the very end goal of winning the ‘war on war’.

This problem arises because answers to the question of ‘whether’ war is becoming obsolete are caught in a ‘general linear reality’ trap ( ABBOTT, 1988 ). Those who argue for the obsolescence of war embrace the teleology of the ‘war on war’ endeavor and base their hopeful prospects on the driving forces of sustained progress, through which they explain the decline of war. Countervailing arguments, in turn, rely on trans-historical, nomothetic fixtures of an anarchical international system to sustain that it is the continuous preparation for war and the prevailing latent threat of war that prevents it from happening. Not surprisingly, both sides of the debate can find empirical confirmation. International society has progressively condemned warfare, intensified its efforts in war prevention, and enabled non-violent solutions for political disputes ( GOLDSTEIN, 2011 ; MUELLER, 1989 ; PINKER, 2011 ). However, at the same time, major powers have continuously increased their military capacity and allocated a large share of their budgets to war preparation, thus increasing the costs of engaging in war ( COPELAND, 2014 ; LEVY and THOMPSON, 2010 ). Both sides validate their arguments, and both causes covary with their explananda, but they give competing explanations for them. In that way, the explanatory necessity of their claims could only be addressed by the (inaccessible) counterfactual inexistence of one or the other.

This apparent contradiction reveals the general linear reality trap in which both sides of the debate fall. They both simultaneously superimpose a linear conception of the future into their reading of the past and build their visions of the future on such a reading. This contradiction also exposes the inherent challenge, widely debated by philosophers of science² , of grounding explanation and prediction of social phenomena in invariant covering laws. In granting either a teleological or nomothetic linearity on (partially) ‘constant conjunctions’, both answers to the ‘end of war’ debate misconceive a set of complex unbounded processes for a bounded one, concealing the bundle of conflicting (and often open-ended) mechanisms that have hitherto set it forth. I argue that we need to transcend the ‘end of war’ debate to gain greater insight and better engage with the ‘war on war’.

To that end, I provide heuristics for opening this black box of mechanisms underlying either answer to the ‘end of war’ debate. Mechanism-based explanations focus on unpacking the ‘nuts and bolts’ involved in the process of ‘producing’ singular instances of a phenomenon, instead of their counterfactual necessity (DONNELLY, 2019a; ELSTER, 2007 ). Epistemologically, this approach circumvents the “fundamental problem of causal inference” (KING, KEOHANE, and VERBA, 1994, p. 79) by embracing the overlap between the questions of ‘how’ and ‘why’ social outcomes (like war) are produced. Therefore, a mechanism-centered approach shifts our focus from unverifiable causal claims to constructing a portfolio of contingent open-ended ‘single causes’ that, by analogy, can provide actors with tools to intervene in the social world – to fight the ‘war on war’.

I further argue that addressing the ‘war on war’ with a mechanism-centered approach requires a concept of war open to the intrinsic contingency of social processes. War thus need not be seen as a transhistorical phenomenon, but rather as one that is intrinsically contingent and escapes ‘covering law’ explanations³ . Instead of looking for bold accounts of ‘why’ war does not reoccur and ‘ whether’ it has become obsolete, we should focus on ‘how’ non-military solutions to otherwise militarized outcomes of political contentious episodes have prevailed. As the escalation of the Russian aggression on Ukraine reminds us, interstate wars, and not least major power wars, remain very much in the realm of possibility in international politics, and engaging with that possibility seems crucial for it to remain a statistical noise – or even only a possibility. A mechanism-oriented assessment of the war on war does not overlook the contingency of war but builds on it to better address the persistent contentiousness of world politics.

In the remainder of this article, I argue for such a mechanism-centered assessment of the ‘war on war’ in three main steps. First, I discuss how both answers to the ‘end of war’ question are imbricated in the construal of the social world as a general linear reality, analytically restricting not only our understanding of the process in which war has become an increasingly rare phenomenon in world politics but also our ability to intervene in such a process. Second, building on scholarly debates about explanations in international relations, I discuss the epistemological and pragmatic advantages of using mechanisms to explain and intervene in the social world. Finally, I illustrate how a mechanism-centered approach to the ‘war on war’ can help us transcend the end-of-war debate; I briefly develop an example of a heuristic framework based on the notion of war as a particular form of contentious politics. The conclusion summarizes the argument and contends that we would be in a better position to understand the potential obsolescence of war and act on it if we shifted our focus to mechanisms.

The ‘end of war’ debate trapped in the cage of ‘general linear reality’

Producing useful explanations for socially relevant phenomena is a common endeavor in distinct branches of social sciences⁴ . The prospective debate on the potential ‘end of war’ has thus been intrinsically associated with its retrospective explanation. To predict ‘whether’ war will not reoccur, one must determine ‘why’ it has not reoccurred. However, this explanation most often involves invariant causal claims about the effects of a certain variable X on the relevant phenomenon Y – in this case, war ( JACKSON, 2017 ; KURKI, 2006 ; LEBOW, 2014 ). These causal claims are formulated to limit uncertainty and built reliable knowledge that Y will follow X in most (if not all) instances ( JACKSON, 2011 , pp. 70-71). To that end, the scientific endeavor of explanation is to ‘empirically’ distinguish efficient causes of the phenomenon from confounding covariates.

However, as widely acknowledged by neopositivists, securing such a distinction is always a tentative enterprise due to the so-called ‘fundamental problem of causal inference’. Simply put, every causal inference relies on the assumption that the observed outcome would not exist had the postulated cause been absent ( BEACH, 2016 , p. 16). As causal processes are factually irreversible, inference about causal claims can be only assessed through a ‘potential outcomes framework’, reliant on a strategy to ‘measure’ such a counterfactual world in a matching control group ( RUBIN, 2005 , 1974 ). Consequently, our ability to intervene in the world becomes predicated on ‘general claims about how the world’ ‘is’, based on the continuous and impossible quest for ‘certain’ ‘knowledge about how the world’ ‘has been’ , and the assumption that this is ‘most likely how the world’ ‘will be’.

As Abbot (1988, pp. 170-171) has put it, this strategy traps our knowledge in this apparently safe world of "general linear reality”, which assumes that “the social world consists of fixed entities (the units of analysis) that have attributes (the variables)” that invariably “cause” one another. By focusing on logically ascertaining and empirically verifying causal claims that “work the same under all conditions” ( TILLY, 2001 , p. 23), we are, by definition, left oblivious to the historical contingency producing these entities, their attributes, and the relations between those attributes. Ultimately, our ability to learn from history is confined to the construction of a linear nomothetical or teleological production of the present and, putatively, of the future. Both answers to the ‘end of war’ debate intrinsically fall under this trap of a ‘general linear reality’. In the remainder of this section, we assess them.

Socio-evolutionary progress and teleological linearity of the ‘end of war’ debate

At the core of the optimistic prospects for the potential obsolescence of war is a teleological notion of social-evolutionary progress, which reduces history to a prelude to the present. Several works on this perspective fall broadly into the so-called ‘liberal peace theory’ ( DOYLE, 1983 ; OWEN, 1994 ). At least since Kant, the dissemination and consolidation of changing attitudes toward war are a central intuitive explanation for both the decline of war and its potential extinction ( BOURKE, 1942 ). This perspective focuses on the individual and collective effects of democracy, the market economy, and cosmopolitan liberal values on the production of sustained peace ( DOYLE, 1983 ; GARTZKE, 2007 ; MOUSSEAU, 2019 ; O'NEAL and RUSSET, 1999; OWEN, 1994) . To be sure, this body of literature provides theoretically sophisticated arguments and a rich inferential debate about these causes and their interactive effects, whose minutia I will not unpack here. And yet it bases its argument on the self-propelling properties of these progressive values and institutions.

Another central perspective on the obsolescence of war tries to bring in contingency by focusing on collective learning. Mueller (1989) , one of the most effective contemporary advocates of this position, argued that the experience of mayhem in contemporary major wars has led leaders in the developed world to be more conscious of the perils of fighting wars against strong enemies and to devise strategies of concertation. Goldstein’s ‘war on war’ focuses more specifically on the system of preventive diplomacy, mediation of conflicts, and peacekeeping established after World War II (GOLDSTEIN, 2011). In an effort to avoid repeating the traumatic events of the first half of the twentieth century, states have built an apparatus in the United Nations (UN) to prevent episodes of contention from emerging and to provide feasible non-military responses to resolve them.

This evolutionist narrative is explicitly expressed in Gat’s work, whose theory of war (and its absence) relies on “an integrated human motivational complex, shaped by evolution and natural selection” (GAT, 2009, p. 571). War becomes obsolete as other means for attaining the desired moral, material, and sexual resources become more efficient ( GAT, 2009 , pp. 592-595). The psychologist Steven Pinker (2011) takes a deeper dive into human nature in search of explanations; in addition to acknowledging new escape valves for our ‘inner demons’, he theorizes about the unleashing of the ‘better angels’ of our nature. For him, human evolution is not only driven by egotistical predation and retaliation: our evolutionary programming is “an open-ended generative system” also equipped with empathy, moral sense, and reason ( PINKER, 2011 , p. 13).

Humans would be leaving war behind in their past, according to Pinker (2011) , because the best features of human nature have been selected through a confluence of social-environmental processes. These processes include the monopolization of the legitimate use of force by the state, greater economic interdependence underpinned by cooperative exchange, feminization of cultural values, and cosmopolitanism. Per Pinker (2011) , these processes make cooperative socialization and non-violent competitive socialization features of greater adaptive fitness.

Hence, although recurring in a modern open-ended version of the old rhetoric of “invoking the first man to settle our disputes for us” ( GELLNER, 1988 , p. 23), Pinker’s explanation (2011) for why war is declining ultimately lies in his close-ended account of the selective properties of the social environment. However, as Levy and Thompson (2010) argue, there are no reasons to believe that the consolidation of the state or world commerce have a deterministic effect on the long peace that Pinker attributes to them ( GLEDITSCH et al., 2013 ). Although the ‘civilizing process’ behind the transformation of human attitudes toward violence seems to have clearer effects on political conflict, these are not as coherently prone to peace as Pinker (2011) assumes ( JACKSON, 2003 ; ZARAKOL, 2014 , 2011 ).

I argue that the fundamental problem with such a perspective is its teleological linearity. Such teleology is intrinsic to the debate in which these scholars are engaged. To justify the claim of the potential ‘end of war’, they conflate highly contingent processes and open-ended mechanisms with one univocal socio-evolutionary process bound to progress. By focusing on supporting this causal claim, they build a linear vision of the future on a linear reading of the past, ignoring the more complex pathways that allowed the still very contentious international politics of recent decades to produce non-violent outcomes.

Fixtures of anarchy and nomothetical linearity in the ‘end of war’ debate

On the other side of the debate, several scholars argue that, paradoxically, the fact that the latency of war is a definitional feature of international relations – with states continually preparing to engage in it – has raised its costs to an unbearable point. War and war preparation have been historically constitutive of the modern (European) state as a hub of authority and resources ( ELIAS, 2000 ; SPRUYT, 1994 ; TILLY, 1992 ). This historical feature has led anarchy-centered accounts to de-historicize war as an inherent attribute of the state system ( HERZ, 1950 ; JERVIS, 1978 ; MEARSHEIMER, 2001 ; WALTZ, 1999 ).

Continuous preparation for war would thus explain the absence of actual wars, due to deterrence. Schelling (1960) argues that the threat of force, rather than its use, has become a central instrument of international politics, especially after strategic nuclear weapons emerged, allowing for mutually assured destruction (MAD). Waltz (1981) argues that MAD substantially diminishes the risk of war, as states would have no reason to seek their survival by creating an ‘’Armageddon. From this perspective, the world has been kept safe because of the maintenance of MAD.

Snyder (1961 , p. 226), however, pondered that the very certainty of MAD carries a paradoxical self-deterring feature, reassuring states that nuclear weapons would not be used unless in response to a first nuclear strike. In that case, MAD has no absolute effect on preventing conventional warfare. In fact, nuclearized states that were not facing MAD have not used ‘nukes’ against non-nuclearized states, and there have been instances where nuclearized states made military movements that could have led to retaliation regardless of MAD ( SNYDER, 1961 ). Fear of nuclear destruction alone does not explain the lack of major power wars since World War II; however, it looms large in every strategic calculation.

Although advocating the central role of MAD, Levy and Thompson (2010) ascribe the decline of war to its progressively high overall costs. Nuclear weapons are not the only deterrent in contemporary warfare: capacities for conventional combat are also increasingly deadly. For Levy and Thompson (2010) , war has become unbearable because its costs are increasing and because its economic rationale has changed. Their argument still emphasizes that states are no longer willing to bear the costs of wars with increasingly destructive potential, but it also acknowledges that cost alone does not tell the whole story. The explanation must incorporate the relative costs of military instruments vis-à-vis other political instruments.

Paradoxically, those who deny the prospect of the ‘end of war’ see deterrence, upheld by the ever-present security dilemma of anarchy, as the ballast of peace. However, situations such as the Russian aggression on Ukraine in 2022 challenge the claim that nuclear weapons – and even MAD – have direct linear effects on war prevention. MAD seems to have protected Russia against direct NATO military involvement in Ukraine, and it has not deterred Russia from waging a war against a state whose partnership with NATO was growing ever closer. MAD also seems to have played a role in encouraging NATO’s response to escalate to the most severe forms of retaliation and involvement in the conflict short of direct military participation. This case resonates with what the literature calls a stability–instability paradox, where MAD prevents strategic conflict but may encourage limited conflict ( JERVIS, 1984 , RAUCHHAUS, 2009).

Therefore, while these perspectives serve as a crucial reminder of the set of structures and configurations that maintain war as a latent potential outcome of international politics, their reading ‘de’-historicizes such a configuration under the guise of a transhistorical law. Furthermore, there are some perverse implications when war reduction is explained by the deterring properties of the ever-growing destructive power of war. Such a view naturalizes war and incentivizes states to continuously mobilize resources toward war as the only possible solution for war prevention, feeding this fundamental material process that sustains the risk of war. Fundamentally, it disregards the social endeavor involved in diverting potential wars to non-violent outcomes. It focuses on fixtures of anarchy that neglect the fundamental contingent role of collective agency in deterring the latency of war from its potential realization.

The linear understanding of the ‘end of war’ and the ‘war on war’

Building the ‘end of war’ debate on this ‘general linear reality’ prism poses some crucial problems for understanding and acting on the ‘war on war’. By focusing on empirically verifying overarching causal claims about the decline of war to assess its potential obsolescence, we remain blind to the historical contingency of such a process. There are two ways in which ‘general linear reality’ prevents us from deriving useful knowledge from contingency and complexity. The first concerns the causal complexity involved in producing every social outcome, not least wars. This causal complexity can take many forms, including the open-ended effects of causes (multi-causality), multiple causes of effects (equifinality), and non-linear interactions among potential causes ( ABBOTT, 1988 ; BEACH 2016 , p. 17; GUZZINI, 2017b; HUMPHREYS, 2017 , pp. 724-725,). Secondly, by demanding ‘empirically measurable counterfactuals’ , efficient causation forgoes the understanding of how any single phenomenon was produced. The only grounds for ‘really’ knowing why any concrete instance of a phenomenon occurs is to subsume it to a general explanation, a covering law whose observable implications have been vetted by systematic analysis of the covariation between putative causes and the outcome – any other explanation “might not be ‘really’ causal” ( JACKSON, 2017 , p. 692).

These two shortcomings combined are a problem if we want to act in the world, as we lose sight of ‘how’ relevant outcomes are produced ( HUMPHREYS, 2017 ; JACKSON, 2017 ). They prevent us from harnessing our understanding of the ‘end of war’ to inform social action designed to advance the ‘war on war’. The debate about the ‘end of war’ is pressing exactly because a substantial part of society wants war to end. War is not a normatively neutral phenomenon; it is an undesirable outcome. Useful explanations of phenomena should inform “concrete action steps […] to solve problems and resolve problem-situations”, that is, to “enhance our practical capacity” to produce a desired rather than an undesired outcome ( JACKSON 2017 , p. 698). I argue that a mechanism-centered approach is fundamental to advancing such an endeavor.

Mechanisms for transcending ‘general linear reality’

Mechanism-centered explanations have gained ground in the discipline of international relations precisely to unpack how causes produce effects ( BENNETT, 2013 ; BENNETT and CHECKEL, 2014 ; GEORGE and BENNETT, 2005 ; WALDNER, 2012 ). The relation between the ‘why’ of causal claims and the ‘how’ of singular explanatory chains, however, is far from trivial. While some scholars argue that the focus should be on mechanisms as intervening variables linking causes and effects, others have sought to abductively identify the mechanisms involved in producing singular social outcomes, whose local causality could be contingently “transferable to other cases” (GUZZINI, 2017b, p. 754)⁵ . Whereas the first tradition remains trapped in the same problems of causal inference, the latter can provide valuable heuristic insight to explain the recurrent production of analogous social phenomena. Yet mechanism-centered approaches have struggled to operationalize the linkage between contingency and generalization, which limits their ability to offer an alternative to the linear reality model and build cumulative knowledge that can travel across cases ( CHECKEL, 2017 , p. 412).

I follow Jackson’s (2017) argument that, even if we refrain from ascertaining and verifying the generality of causal claims, we still draw on their theoretical abstractions to produce explanations. First and foremost, these generalizations allow for a ‘problem-situation’ to be classed as a phenomenon somehow similar to other instances of a given explanandum Y. These generalizations are the existing constitutive and causal theories about phenomena analogous to Y. With this previous knowledge about the production of an abstract class of Y, one can trace the contingent production of that specific outcome. This is not to reduce the explanation of a singular outcome to an ‘if-then’ law about Y but rather to build on available theoretical knowledge about analogous phenomena, to select “salient factors in the part of the causal net” that are contingently producing such events ( HUMPHREYS, 2017 , p. 721)⁶ . “That generality is what bridges the gap between individual instances, and makes the explanation useful,” by allowing for a tentative “manipulability” over the outcomes we want to affect ( JACKSON, 2017 , p. 703).

The explanatory power of mechanisms thus relies on the analytical dialogue between the contingency of local causation and the abstractions afforded by homology and analogy with phenomena of a similar nature and in dialogue with the heuristics of theories about these phenomena⁷ . This is a perspective similar to Beach’s systemic view of mechanism as “a series of parts composed of entities engaging in activities” that come together to produce analogous outcomes (BEACH, 2016, p. 18). Donnelly (2019a), surveying the philosophy of biology more deeply, adds that processual mechanisms should be understood as “productively organized entities and activities”⁸ . The key logic is to go beyond entities and activities and highlight the ‘configuration of their organization’⁹ . Processual-relational approaches in the field of international relations and elsewhere have theorized the emerging properties of organizational configurations, which should also be captured by a mechanism-centered approach¹⁰ .

In this sense, mechanisms are always embedded in a constitutive theory and scientific ontology about the social production of a given phenomenon, but they are not limited by their intrinsic simplifications. Scientific ontologies provide the meta-theoretical lens for any explanation, the abstractions we use to term the entities, activities, and relations in the social world ( JACKSON and NEXON, 2013 ). Constitutive theories mobilize this ontology to identify the specific configuration of entities, activities, and relations that allow us to class a problem-situation and explain how this configuration shapes the production of outcomes of interest ( NORMAN, 2021 ). While these abstractions do not allow for the sequence of events that can lead to an outcome to be deducted, they are necessary to identify contingent causal pathways in single cases and learn from the contingency of similar problem-situations.

The dialogue between abstraction and contingency may be seen as grounded in homologies and analogies. Homologies focus on entities “having the same or similar relation, relative position, or structure”¹¹ , that is, their being organized in a similar configuration. Causal processes are homologous insofar as similar entities are performing similar activities through similar configurations¹² . Analogies entail “highlighting correspondence or partial similarity” so that a phenomenon appears “comparable to something else in significant respects”¹³ . It is thus the homological production of analogous outcomes that allows for explanatory insights to travel across contexts¹⁴ .

I would argue that mechanisms – through analogy and homology – allow us to escape the cage of the ‘general linear reality’ by drawing on abstraction and generalization pragmatically. Through analogy, we can identify a problem-situation as being similar to a class of existing phenomena. Through homology, we translate the theoretical knowledge we already have about the configurations of actors and activities that explained the production of outcomes analogous to those in the context of the case under analysis. Both analogy and homology involve a dialogue with previous empirical and theoretical knowledge about a given class of phenomena.

Figure 01 captures this idea schematically. As shown in (A), a ‘general linear reality’ perspective has a causal logic in which the effects pass from a cause X to an outcome Y through an equally invariant mechanism. The mechanism-centered perspective, however, overcomes the self-imposed limitations of linear perspectives by focusing on the many combinations of the multiple configurations of entities and activities – the mechanisms – that have produced other instances of Y. It uses a simplified abstraction to assign an outcome of interest to the same class of phenomena and identify the configurations of entities, activities, and relations producing this outcome. Previous explanations can be contrasted with this ontological framework to then organize contingency into an explanation. Hence, a mechanism-centered approach can rely on the generality of causal claims without losing non-linearities and other causal complexities deriving from each singular instance of the given phenomena, as it abdicates determining the degree of generality of those causal claims.

Figure 01
Mechanisms for transcending 'general linear reality'

By analogy with similar phenomena, the mechanism-centered approach reifies Y as the productive activity of entity E, which is equally reified through a given constitutive theorization. Hence, in drawing on homologies with previous productive accounts of Y, mechanism-centered analysis pulls the contingent threads of its explanations by breaking with such a descriptive causality. A prosaic example might be helpful: if we want to know ‘why a car has stopped’, by analogy, we use the idea of family resemblance to class our ‘explanandum’ with other phenomena, and so we mobilize a constitutive theory of what a ‘car stopping’ is, a type of entity E (a car) and its activity Y (stopping). To actually explain the motion of a car in a useful way, we need to unpack the entity itself as a set of intrinsic (i.e., the car’s mechanics) and extrinsic (i.e., the driver’s intentions, traffic signs, the pedestrian crossing) configurations of entities and activities that are contextually relevant for the car to stop or some alternative outcome. The connection between these entities and activities would be drawn in a dialogue with existing theories about the phenomenon at hand, and it would involve the use of abstract homologies with other configurations of entities and activities that we believe can lead to stopping – or not – a car. Analysts can, therefore, draw on such generality to build “some general statement [X] connecting factors with the outcome [Y]” ( JACKSON, 2017 , p. 702).

In other words, by using analogy and homology between a given explanandum and other existing phenomena, scholars can avoid the contingency of single-case causation and draw on existing theoretical knowledge to make sense of such contingency. Theories are thus seen not as positive assertions about a linear world, waiting to be falsified, but as generalized abstractions of a non-linear world, which just need to be sufficiently useful and accurate to explain the case under study. Multiple existing explanations can still be juxtaposed through abduction in a given case in order to infer the best local explanation, without making any claim about the general validity of theories¹⁵ . Grounded theory may be built on this locally identified casual pathway insofar as novel mechanisms challenging or refining existing explanations are identified, thus expanding our portfolio of mechanism-centered accounts of analogous phenomena¹⁶ .

This dialogue between specific contingency and general abstraction is represented schematically in Figure 01 . As previously mentioned, pathway (A) represents a theory X, establishing a sequence of abstract mechanisms leading to an outcome Y. As shown in the middle of Figure (B), to explain a given case, one must break with the reification that is built in Y and, by homology with both the constitutive and causal explanations in X, unpack the concrete configurations of entities and activities producing that instance of Y. Over time, this constant dialogue between contingency and abstraction allows for homologous causal pathways attributed to analogous phenomena to accumulate and be refined so they can amplify and streamline our knowledge about the configurations of mechanisms that may interact to produce any singular instance of Y. As illustrated by the bottom diagram (C) in Figure 01 , instead of competing to explain the overall variation (and perhaps no single instance) of Y, the mechanism-centered perspective invites scholars to draw on the general insights derived from the ontological and causal claims of theories to assess the configurations of intrinsic and extrinsic entities and activities that ‘inefficiently’¹⁷ produce ‘specific occurrences’ of a class of phenomena, such as war.

Instead of judging the general validity of competing causal claims, such a mechanism-centered assessment provides scholars and practitioners with a portfolio of causal explanations they can pick and recombine – within the limits of their ontological commitments – to make sense of the world they want to understand or in which they want to intervene. Of course, the very production of these mechanism-centered causal explanations is intrinsically bounded by the constitutive theorization of the phenomenon, the configurations of relations that comprise it, as well as by how these configurations constrain entities and their activities. Turning to the subject of this article, to identify homological mechanisms from analogous phenomena, one must rely on existing theorizations of war as the abstract activity of an entity. In the next section, I build on a definition of war as a contingent potential product of contentious politics among states to illustrate how it can offer heuristics for a mechanism-centered understanding of the ‘war on war’.

The ‘war on war’ in the non-linear world of ‘contentious’ world politics.

In the mechanism-centered framework just outlined, the first step to understanding the mechanisms of the ‘war on war’ is to understand the class of analogous phenomena we want to explain and in which we want to intervene: war. Although theoretical disputes are fierce when scholars discuss the general causes of war or why a given war broke out, they are considerably milder when it comes to defining war. Few would challenge the simplest dictionary definition of war as a “state of armed conflict” among polities¹⁸ . Of course, analytically, that does not say much. While such a dictionary definition identifies the relevant entities producing war (states), it gives little insight, descriptively, into the activity of producing war itself. Clausewitz’s classic definition of war “as a continuation of political intercourse, carried on with other means” (CLAUSEWITZ, 2008, p. 87), on the other hand, already hints at this productive logic of identifying relational systems of entities and activities. It recognizes that states engage in a series of activities, ‘political intercourses’, that, under certain conditions, can produce a specific violent outcome: armed conflict. Moreover, operational definitions for comparative statistical purposes usually establish a threshold of a thousand battlefield casualties¹⁹ , emphasizing the level of violence produced with such an outcome. War is thus a highly violent product of politics among states.

To turn these loose definitions of war into an explanation of war or non-war outcomes, we must identify an ‘ontological or constitutive theory’ of its production. Charles Tilly’s contentious politics framework may offer the grounds for exemplifying such a mechanism-based explanation. It explains collective violence among social groups as a particular manifestation of contentious politics (TILLY, 2003, pp. 04-11). Contentious politics, in turn, is conceived as interactions among “makers of claims”, whose claims, “if realized, affect the interests of at least one of the (other) claimants” (McADAM, TARROW, and TILLY, 2001 , p. 05). Taken as a particular form of political violence, interstate war might be conceived as a particular outcome of contentious politics. In this perspective, (modern) warfare is conceived as an instrument of political mobilization that prevailed in the configuration of the European state system and then became a common means of political socialization as the international system became globalized ( DUNNE and REUS-SMIT, 2017 ; MAcDONALD, 2014 ; SCHENONI, 2021 ; TILLY, 1992 ).

Instead of overlooking historical contingency, a productive definition of war builds it in and then consciously abstracts from it. In the case of war, the state is reified as a recognized claimant of sovereign authority over an indivisible population and territory, while its wars are reified as a means available to further contentious politics through the violent collective mobilization of military resources. In such a conception, interstate conflict is a highly contingent phenomenon, one among many other potential outcomes of contentious politics. This theoretically informed definition is a necessary but insufficient step toward explanation. By definition, states are constitutive of interstate war. Every single war is produced by states, but states engage in contentious politics also by other means.

As outlined in the last section and illustrated in Figure 02 , tracing back the explanation of a phenomenon means breaking with its definitional reification. More specifically, the grammar of any X causal claim about war usually involves breaking any ‘k’ number of warring states into internal and external configurations of concrete entities and activities. Any explanation of why a given war occurred would then draw on homologies between the war-prone problem-situation under study and previous knowledge about configurations of entities and activities producing war in analogous contentious situations. Such a focus on homologous concatenations of mechanisms also allows us to consider alternative configurations that could have led to alternative outcomes. It gives a sense of ‘manipulability’ on which concrete action can be grounded to achieve outcomes other than war ( JACKSON, 2017 ).

Figure 02
Mechanism-centered approach to war as contentious politics

These homologies and analogies allow for a dialogue between the contingent context of a given contentious war-prone situation and existing theories, the abstractions that can help make sense of that context. For example, consider two strong competing explanations for the decline of war: the theories of democratic peace and nuclear deterrence. Both theories can be seen through a rationalist ontology that can underpin a contentious politics explanation. Instead of focusing on whether democratic states fight each other or whether MAD prevents the escalation of conflicts, we should concentrate on the entities and activities that could produce alternative outcomes in situations of contentious politics. For instance, causal explanation in democratic peace theory involves breaking a set of states into their governments and constituencies, which are then further broken into concrete entities (i.e., branches of government, domestic pressure groups), whose activities (i.e., lobbying, protest) produce a non-war outcome in a contentious situation between two or more democratic states ( FEARON, 1994 ; SCHULTZ, 2001 ; TOMZ, 2007 ; WEEKS and TOMZ, 2013 ). Causal explanation in deterrence theory, on the other hand, means breaking the state into a set of decision-making stakeholders who exchange credible threats and costly signals to deescalate conflict ( FUHRMANN and SECHSER, 2014 ; POWELL, 1990 ; SCHELLING, 1960 ; SNYDER, 1961 ).

Of course, in putting into practice the heuristic potential of mechanisms, we must delve much more deeply into the entities and activities of any relevant explanation about the production of war as a potential outcome of contentious politics. Furthermore, one does not need to rely on the tacit rationalism underlying the contentious politics framework. The same can be done with other constitutive theories on the production of war outcomes that are grounded on different social world ontologies. The heuristics, however, remains the same as that which was exemplified in the previous paragraphs. Instead of empirically ascertaining the generality of the explanations about the production of war and non-war outcomes, scholars should revisit these explanations to build a portfolio of mechanism-centered explanations of why international politics produce war or some non-violent outcome. By breaking causal claims into mechanisms, configurations of entities, activities, and relations, scholars provide heuristics to explore the open-ended processes shaping world politics in a non-deterministic fashion and understand how the war on war has been won at every battle.

Conclusion: in search of mechanisms for winning the ‘war on war’ at every episode of contention

Deciding ‘whether’ war is becoming obsolete requires determining ‘why’ it has been in decline since the end of World War II. One way to explain this trend is through the inexorable evolutionary process toward greater sociability and rationality, in that case, war is on the path to obsolescence. Another way to address this issue is to focus on the latency of war, which pushes states to continuously raise the costs of war and paradoxically, would prevent war from happening. I have argued that, by seeking to categorically define ‘whether’ war is obsolete and will not reoccur, both perspectives overly seek to categorize the causes of war decline (the ‘why’), conflating complex social processes with teleological or nomothetical driving forces that make us oblivious to the more nuanced insights on ‘how’ that has come to reality in different contentious events.

As the growing great power rivalry between the United States and China and the invasion of Ukraine remind us of the persistent peril of war in world politics, I argue that we get a more revealing assessment of this historical trend of war decline by diverting the question ‘from why’ war has declined ‘to how’ states achieved alternative outcomes to contentious politics in specific episodes of contention. To be sure, human endeavor seems to be at the heart of every battle won in the war on war, as it has created conditions for peaceful outcomes to thrive where they had not in the past. Yet for such an agency to guide additional steps on the ‘war on war’, we must understand the effective (inefficient) contingency of this agency instead of getting lost in either teleological or nomothetical disputes around general causal claims.

In this work, I sought not only to challenge the value of determining whether war is becoming obsolete but also to offer heuristics to transcend the self-imposed limitations of the ‘end of war’ debate. I argue that, instead of conducting empirical analysis to verify and judge competing causal claims, a mechanism-centered approach to the ‘war on war’ builds on the analytical heuristics underpinning these theoretical claims to ground our ability to understand and intervene in the social world. Seeing war as an outcome of contentious politics is an example of such a dialogue between empirical contingency and theoretical abstraction since such a perspective emphasizes the contingency of war as one possible outcome of contentious episodes among states while it recognizes the recurring agonism of world politics. Instead of just probing the deterring effects of the high costs of war or the causal efficiency of liberal forces of progress, we need to examine configurations of entities and activities that have kept war solely as a potential outcome. A similar heuristic can be applied to other ontological commitments and constitutive theories of war. As the theoretical generality of constitutive and causal explanations is converted into identifiable concrete actors and activities, we can consciously use abstraction and reification to gain insight into the contingency of real-world phenomena.

By further investigating the patchwork of shifting configurations of mechanisms in episodes of political contention among states, along with their multiple effects on violent and non-violent power politics, scholars would offer a pragmatic way of understanding the current battlefield of our noble quest for peace. In doing so, one ought to acknowledge the equifinality and multi-causality producing any single instance of war or peace, avoiding the risks of naturalizing the present and losing sight of the challenges ahead. By moving away from essentialist and deterministic—teleological or nomothetical—causal claims, scholars can embrace the contingency of the social world; and that, to borrow Tilly’s (1996 , p. 1606) cautionary words, can help avoid leaving social scientists baffled by forthcoming plot twists in world politics.

Acknowledgments

I would like to thank the Editorial Team at the BPSR as well as two anonymous reviewers for their constructive engagement with this work. I am also very grateful to Adam B. Lerner, Cristiana Maglia, Jaakko Heiskanen, and Stefano Guzzini for their comments on previous drafts. The research leading to this publication was funded by CAPES ‘Doutorado Pleno no Exterior’ nº 88881.128839/2016-01.

References

• ABBOTT, Andrew (1988) Transcending general linear reality. Sociological Theory . Vol. 06, Nº 02, pp. 169-186.
• BEACH, Derek (2016), What are we actually tracing? Process tracing and the benefits of conceptualizing causal mechanisms as systems. Qualitative & Multi-method Research . Vol. 14, Nº 01/02, pp. 15-22.
• BENNETT, Andrew (2016), Do new accounts of causal mechanisms offer practical advice for process tracing? Qualitative & Multi-method Research . Vol. 14, Nº 01/02, pp. 34–39.
• BENNETT, Andrew (2013), The mother of all isms: causal mechanisms and structured pluralism in International Relations theory. European Journal of International Relations . Vol. 19, Nº 03, pp. 459-481.
• BENNETT, Andrew and CHECKEL, Jeffrey T. (eds) (2014), Process tracing: from metaphor to analytic tool, strategies for social inquiry. Cambridge: Cambridge University Press. 342 pp..
• BERTILSSON, Thora Margareta (2004), The elementary forms of pragmatism: on different types of abduction. European Journal of Social Theory . Vol. 07, Nº 03, pp. 371–389.
• BOURKE, John (1942), Kant's doctrine of ‘Perpetual Peace’. Philosophy. Vol. 17, Nº 68, pp. 324-333.
• CLAUSEWITZ, Carl von (2008), On War . New Jersey: Princeton University Press. 752 pp..
• COPELAND, Dale C. (2014), Economic interdependence and war . Princeton: Princeton University Press. 504 pp..
• CHECKEL, Jeffrey T. (2017), Constructivism at mid-life: theory, methods and the turn to mechanisms. Cooperation and Conflict . Vol. 52, Nº 03, pp. 411-413.
• DONNELLY, Jack (2019a), Explanations, mechanisms, and causes: towards explanatory pluralism in international studies. Paper prepared for presentation at Columbia University.
• DONNELLY, Jack (2019b), Systems, levels, and structural theory: Waltz’s theory is not a systemic theory (and why that matters for International Relations today). European Journal of International Relations . Nº 03, pp. 904-930.
• DOYLE, Michael W. (1983), Kant, liberal legacies, and foreign affairs. Philosophy and Public Affairs . Vol. 12, Nº 03, pp. 205-235.
• DUNNE, Tim and REUS-SMIT, Christian (2017), The Globalization of International Society . Oxford: Oxford University Press. 544 pp..
• ELIAS, Norbert (2000), The civilizing process: sociogenetic and psychogenetic investigations. London: Blackwell. 592 pp..
• ELSTER, Jon (2007), Explaining social behavior: more nuts and bolts for the social sciences. Cambridge: Cambridge University Press. 484 pp..
• FEARON, James D. (1994), Domestic political audiences and the escalation of international disputes. American Political Science Review . Vol. 88, Nº 03, pp. 577-592.
• FUHRMANN, Matthew and SECHSER, Todd S. (2014), Signaling alliance commitments: hand-tying and sunk costs in extended nuclear deterrence. American Journal of Political Science . Vol. 58, Nº 04, pp. 919-935.
• GARTZKE, Erik (2007), The Capitalist Peace. American Journal of Political Science . Vol. 51, Nº 01, pp. 166–191.
• GAT, Azar (2009), So why do people fight? Evolutionary theory and the causes of war. European Journal of International Relations . Vol. 15, Nº 04, pp. 571-599.
• GAT, Azar (2006), War in human civilization . Oxford: Oxford University Press. 840 pp..
• GELLNER, Ernest (1988), Plough, sword and book: the structure of human history. Chicago: Chicago University Press. 288 pp..
• GEORGE, Alexander L. and BENNETT, Andrew (2005), Case studies and theory development in the Social Sciences . Cambridge: MIT Press. 352 pp..
• GLEDITSCH, Nils Petter; PINKER, Steven; THAYER, Bradley A.; LEVY, Jack S., and THOMPSON, William R. (2013), The Forum: the decline of war. International Studies Review . Vol. 15, Nº 03, pp. 396-419.
• GLEDITSCH, Nils Petter; WALLENSTEEN, Peter; ERIKSON, Mikael; SOLLENBERG, Margareta, and STRAND, Håvard (2002), Armed conflict 1946-2001: a new dataset. Journal of Peace Research . Vol. 39, Nº 05, pp. 615-637.
• GOLDSTEIN, Joshua (2011), Winning the War on War: the decline of armed conflict worldwide. New York: Dutton. 400 pp..
• GUZZINI, Stefano (2017a), Afterword: International political sociology, or: the social ontology and power politics of process. In: Routledge handbook of international political sociology . Edited by GUILLAUME, Xavier and BILGON, Pinar. Milton Park Abingdon: Routledge. Pp. 366-375.
• GUZZINI, Stefano (2017b), Power and cause. Journal of International Relations and Development . Vol. 20, Nº 04, pp. 737-759.
• HERZ, John H. (1950), Idealist internationalism and the security dilemma. World Politics . Vol. 02, Nº 02, pp. 157-180.
• HUMPHREYS, Adam R C. (2019), Realism, empiricism and causal inquiry in International Relations: what is at stake? European Journal of International Relations . Vol. 25, Nº 02, pp. 562-587.
• HUMPHREYS, Adam R C. (2017), Causation, complexity, and the Concert: the pragmatics of causal explanation in International Relations. Journal of International Relations and Development . Vol. 20, Nº 04, pp. 717-736.
• HUMPHREYS, Adam R. C. (2010), The heuristic application of explanatory theories in International Relations. European Journal of International Relations . Vol. 17, Nº 02, pp. 257-277.
• JACKSON, Patrick Thaddeus (2017), Causal claims and causal explanation in international studies. Journal of International Relations and Development . Vol. 20, Nº 04, pp. 689-716.
• JACKSON, Patrick Thaddeus (2011), The conduct of inquiry in International Relations: Philosphy of Science and its implication for the study of world politics. Milton Park Abingdon: Routledge. 288 pp..
• JACKSON, Patrick Thaddeus (2003), Defending the West: occidentalism and the formation of NATO. Journal of Political Philosophy . Vol. 11, Nº 03, pp. 223-252.
• JACKSON, Patrick Thaddeus and NEXON, Daniel H. (2013), International theory in a post-paradigmatic era: from substantive wagers to scientific ontologies. European Journal of International Relations . Vol. 19, Nº 03, pp. 543–565.
• JACKSON, Patrick Thaddeus and NEXON, Daniel H. (1999), Relations before States: substance, process and the study of world politics. European Journal of International Relations . Vol. 05, Nº 03, pp. 291-332.
• JERVIS, Robert (1984), The illogic of American nuclear strategy. Ithaca: Cornell University Press. 204 pp..
• JERVIS, Robert (1978), Cooperation under the security dilemma. World Politics . Vol. 30, Nº 02, pp. 167-214.
• KING, Gary; KEOHANE, Robert O., and VERBA, Sydnei (1994), Designing social inquiry . New Jersey: Princeton University Press. 264 pp..
• KURKI, Milja (2006), Causes of a divided discipline: rethinking the concept of cause in International Relations Theory. Review of International Studies . Vol. 32, Nº 02, pp. 189-216.
• LEBOW, Richard (2014), Constructing cause in International Relations . Cambridge: Cambridge University Press. 206 pp..
• LEVY, Jack and THOMPSON, William (2010), Causes of War . Oxford: Willey-Blackwell. 288 pp..
• MAcDONALD, Paul K. (2014), Networks of domination: the social foundations of peripheral conquest in International Politics. Oxford: Oxford Univeristy Press. 336 pp..
• MAOZ, Zeev; JOHNSON, Paul L.; KAPLAN, Jasper; OGUNKOYA, Fiona, and SHREVE, Aaron P. (2019), The dyadic militarized interstate disputes (MIDs). Dataset Version 3.0: Logic, characteristics, and comparisons to alternative datasets. Journal of Conflict Resolution . Vol. 63, Nº 03, pp. 811-835.
• McADAM, Doug; TARROW, Sidney, and TILLY, Charles (2001), Dynamics of contention . Cambridge: Cambridge University Press. 412 pp..
• MEARSHEIMER, John J. (2001), The tragedy of great power politics . New York: Norton. 561 pp..
• MOUSSEAU, Michael (2019), The end of War: how a robust marketplace and liberal hegemony are leading to perpetual World Peace. International Security . Vol. 44, Nº 01, pp. 160-196.
• MUELLER, John (1989), Retreat from Doomsday: the obsolescence of Major War. New York: Basic Books. 336 pp..
• NEXON, Daniel H. (2009), The struggle for power in early modern Europe: religious conflict, dynastic empires, and international change. Princeton: Princeton University Press. 408 pp..
• NORMAN, Ludvig (2021), Rethinking causal explanation in interpretive international studies. European Journal of International Relations . Vol. 27, Nº 03, pp. 936–959.
• O'NEAL, James and RUSSET, Bruce (1999), The Kantian peace: the pacific benefits of democracy, interdependence, and Iiternational organizations, 1885–1992. World Politics . Vol. 52, Nº 01, pp. 01-37.
• OWEN, John M. (1994), How liberalism produces democratic peace. International Security . Vol. 19, Nº 02, pp. 87-125.
• PINKER, Steven (2011), The better angels of our nature: why violence has declined. New York: Viking. 802 pp..
• POULIOT, Vincent (2014), Practice tracing. In: Process tracing: from metaphor to analytic tool. Edited by BENNETT, Andrew and CHECKEL, Jeffrey T.. Cambridge: Cambridge University Press. pp. 237-259.
• POWELL, Robert (1990), Nuclear deterrence theory: the search for credibility. Cambridge: Cambridge University Press. 240 pp..
• RAUCHHAUS, Robert (2009), Evaluating the nuclear peace hypothesis: a quantitative approach. Journal of Conflict Resolution . Vol. 53, Nº 02, pp. 258-277.
• RUBIN, Donald B. (2005), Causal inference using potential outcomes. Journal of the American Statistical Association . Vol. 100, Nº 469, pp. 322-331.
• RUBIN, Donald B. (1974), Estimating causal effects of treatments in randomized and nonrandomized studies. Journal of Educational Psychology . Vol. 66, Nº 05, pp. 688-701.
• RUNHARDT, Rosa (2016), Tracing the productive continuity of social mechanisms. Qualitative & Multi-method Research . Vol. 14, Nº 01/02, pp. 22–28.
• RUSSET, Bruce and O'NEAL, James (2001), Triangulating peace: democracy, interdependence, and international organizations. New York: Norton. 394 pp..
• SARKEES, Meredith Reid and WAYMAN, Frank (2010), Resort to War, 1816-2007 . Washington: CQ Press. 577 pp..
• SCHELLING, Thomas C. (1960), Strategy of conflict . Cambridge: Harvard University Press. 328 pp..
• SCHENONI, Luis L. (2021), Bringing war back in: Victory and state formation in Latin America. American Journal of Political Science. Vol. 65, Nº 02, pp. 405-421.
• SCHULTZ, Kenneth A. (2001), Looking for audience costs. Journal of Conflict Resolution . Vol. 45, Nº 01, pp. 32-60.
• SNYDER, Glenn H. (1961), Deterrence and defense: toward a Theory of National Security. Princeton: Princeton University Press. 306 pp..
• SPRUYT, Hendrik (1994), Institutional selection in international relations: state anarchy as order. International Organization . Vol. 48, Nº 04, pp. 527-557.
• SWEDBERG, Richard (2016), Before theory comes theorizing or how to make social science more interesting. The British Journal of Sociology . Vol. 67, Nº 01, pp. 05-22.
• TILLY, Charles (2003), The politics of collective violence . Cambridge: Cambridge University Press. 290 pp..
• TILLY, Charles (2001), Mechanisms in political process. American Review of Political Science . Vol. 04, pp. 21-41.
• TILLY, Charles (1996), To explain political processes. American Journal of Sociology . Vol. 100, Nº 06, pp. 1594-1610.
• TILLY, Charles (1992), Coercion, capital, and European States, AD 990-1992 . Oxford: Blackwell Publishers. 288 pp..
• THOMAS, Gary (2010), Doing case study: abduction not induction, phronesis not theory. Qualitative Inquiry . Vol. 16, Nº 07, pp. 575–582.
• TOMZ, Michael (2007), Domestic audience costs in international relations: an experimental approach. International Organization . Vol. 61, Nº 04, pp. 821-840.
• WALDNER, David (2012), Process tracing and causal mechanisms. In: The Oxford Handbook of the Philosophy of Social Science. Edited by KINCAID, Harrold. Oxford: Oxford University Press. Pp. 65-84.
• WALTZ, Kenneth N. (1999), Globalization and governance. PS: Political Science and Politics . Vol. 32, Nº 04, pp. 693-700.
• WALTZ, Kenneth N. (1981), The spread of nuclear weapons: more may be better: Introduction. The Adelphi Papers . Vol. 21, Nº 171, pp. 01-38.
• WEBER, Max (1958), Science as a vocation. Daedalus . Vol. 87, Nº 01, pp. 111-134.
• WEEKS, Jessica L. P. and TOMZ, Michael (2013), Public opinion and the democratic peace. American Political Science Review . Vol. 107, Nº 04, pp. 849-865.
• ZARAKOL, Ayse (2014), What made the modern world hang together: socialisation or stigmatisation?. Intenational Theory . Vol. 06, Nº 02, pp. 311-332.
• ZARAKOL, Ayse (2011), After defeat: how the East larned to live with the West. Cambridge: Cambridge University Press. 312 pp..

• 1
  See ( GLEDITSCH et al., 2013 ) for a detailed discussion on the empirical basis for such a claim. Although the question of since when has war been declining is important, the main focus here is on how scholars addressing war decline explain such a trend and which are the analytical and pragmatic implications of these explanations.
• 2
  See Jackson (2011 , pp. 01-16), for a summary of the debate.
• 3
  As for Tilly (2001 , p. 23), covering laws are overarching explanatory generalizations that “work the same under all conditions”, in his critical reference to Hume-Hempel’s causal model of explanation. In the remainder of this article, all references to “covering laws” and “causal claims” are used interchangeably as in ( JACKSON, 2017 ).
• 4
  See ( JACKSON, 2011 ; WEBER, 1958 , pp. 16-23,).
• 5
  See also (DONNELLY, 2019a; NEXON, 2009 , pp. 13-17; POULIOT, 2014 ; TILLY, 2001 ).
• 6
  See also ( HUMPHREYS, 2019 ).
• 7
  See ( HUMPHREYS, 2010 ).
• 8
  See Donnelly, 2019a, p. 06. Website.
• 9
  See Donnelly, 2019a.
• 10
  (GUZZINI, 2017a; NEXON, 2009 ; POULIOT, 2014 ; TILLY, 2001 ). See also ( JACKSON and NEXON, 1999 ).
• 11
  Oxford English Dictionary. Available at ˂ https://www.lexico.com/definition/homology ˃.
• 12
  See (DONNELLY, 2019b).
• 13
  Oxford English Dictionary. Available at ˂ https://www.lexico.com/definition/homology ˃.
• 14
  This is close to a combination of the strategies for identifying the “productive continuity” of mechanisms that Runhardt (2016) calls the ‘regularities and causal process observations’. Although her work compellingly argues that the first is not necessary and the second is insufficient for causal inference, I believe her critique only holds if one seeks to infer efficient causality in a ‘potential outcomes framework’, against which I argue here. See also ( BENNETT, 2016 , pp. 37-39).
• 15
  See Bertilsson (2004) and Thomas (2010) .
• 16
  See Swedberg (2016) .
• 17
  See Lebow (2014) .
• 18
  Oxford English Dictionary: Available at ˂ https://www.lexico.com/definition/war ˃ . This of course disregards armed conflicts within and across states, as does the debate around the “end of war” ( GLEDITSCH et al., 2013 ).
• 19
  See ( GLEDITSCH et al., 2002 ; MAOZ et al., 2019 ; SARKEES and WAYMAN, 2010 )

Edited by

Revised by Karin Blikstad

Publication Dates

• Publication in this collection
  15 May 2023
• Date of issue
  Apr 2023

History

• Received
  07 Apr 2022
• Accepted
  05 Oct 2022