Publication Type: Conference Paper/Unpublished Manuscript
Review Method: Peer Reviewed
Abstract: The paper contrasts the Hempelian deductive-
nomological model of explanation with the
explanation of social phenomena via reference
to causal mechanisms. It defines causal
mechanisms and explores the implications of
using them for explanatory purposes, addressing
issues such as microfoundationalism,"as if"
assumptions, reference to unobservables, and
"infinite regress" to finer levels of detail.

Publication Type: Conference Paper/Unpublished Manuscript
Abstract: The appeal to ideas as causal variables and/or constitutive features of political processes increasingly characterises political analysis. Yet, perhaps because of the pace of this ideational intrusion, too often ideas have simply been grafted onto pre-existing explanatory theories at precisely the point at which they seem to get into difficulties, with little or no consideration either of the status of such ideational variables or of the character or consistency of the resulting theoretical hybrid. This is particularly problematic for ideas are far from innocent variables – and can rarely, if ever, be incorporated seamlessly within existing explanatory and/or constitutive theories without ontological and epistemological consequence. We contend that this tendency along with the limitations of the prevailing Humean conception of causality, and associated epistemological polemic between causal and constitutive logics, continue to plague almost all of the literature that strives to accord an explanatory role to ideas. In trying to move beyond the current vogue for epistemological polemic, we argue that the incommensurability thesis between causal and constitutive logics is only credible in the context of a narrow, Humean, conception of causation. If we reject this in favour of a more inclusive (and ontologically realist) understanding then it is perfectly possible to chart the causal significance of constitutive processes and reconstrue the explanatory role of ideas as causally constitutive.

Publication Type: Conference Paper/Unpublished Manuscript
Review Method: Peer Reviewed
Abstract: The following is a draft chapter from a manuscript entitled Peacekeeping and the
Peacekept: Maintaining Peace After Civil War. The book examines peacekeeping in civil
conflicts in the post-Cold War era. It combines quantitative and case study analysis to examine
several empirical questions: where do peacekeepers get sent? do they make peace more likely to
last, all else equal? and if so, how do they do so – what are the causal mechanisms linking the
presence of peacekeepers to more enduring peace? It is this last question that the chapter
presented here addresses.
In an earlier theory chapter, I note that despite a now rather vast literature on
peacekeeping, we do not have a strong causal theory of how exactly peacekeepers make a
difference. Many studies discuss the functions of peacekeeping (monitoring, interposition, etc.)
without spelling out explicitly what the presence of peacekeepers changes from the perspective
of the belligerents themselves, the “peacekept” that would make them less likely to return to war.
I argue that civil war may resume through four possible causal pathways (or some combination
thereof): outright aggression; security dilemmas spirals driven by fear and uncertainty; accidents
or “involuntary defection” by rogue elements within either side; political exclusion that drives
the politically losing side back to war. The theory chapter spells out specific causal mechanisms
through which peacekeepers might interfere with any of these pathways, thereby making peace
more likely to last. The following chapter examines three cases, covering attempts to maintain
peace both with and without peacekeepers, to evaluate whether and how these causal
mechanisms operate.

Publication Type: Proceeding

Publication Type: Proceeding
Abstract: The notion that important elements of political action may be “socially constructed” has received growing attention in political science since the 1980s. Meta-theorizing and empirical work on culture, ideas, norms, and identity has garnered increasing space in top journals. “Constructivist” theorists of various types have won book contracts from university presses and faculty positions at prestigious universities.
Despite rising recognition of ideational or constructivist scholarship, however, its status in the discipline remains ambiguous. Instances where empirical claims by non-ideational theorists entertain ideational alternative explanations are the tiny exception rather than the rule. Empirical ideational arguments, by contrast, still achieve high-level mainstream publication only given elaborate and explicit competition with non-ideational claims. Much of their apparent rise has also come not in old mainstream venues, but in the emergence of newer journals which are less insistent that constructivists engage elaborate contests with non-ideational theories along orthodox methodological lines.
This is probably a typical pattern for new schools of thought that attempt over decades to be admitted to central disciplinary debates (at least in the post-1945 world of proliferating journals). But if we turn from publication patterns to the internal logic of competition between ideational and non-ideational scholarship, it becomes clear that we are not just looking at the slow movement of a new approach into mainstream debates. On both sides of this line—and especially amongst constructivists themselves—we find scholars arguing that the appropriate end-result will not be engagement and competition but the consolidation of separate-but-equal realms of inquiry.
Two related notions suggest that constructivists and non-constructivists are most interested in distinct discussions. The first traces most famously to Max Weber, who taught that arguments that invoke meaning engage something other than explanation. Weber posited a difference between an argument’s “adequacy on a causal level”—its explanatory force—and the kind of understanding, or “adequacy on the level of meaning,” to which ideational scholarship aspires. He thought we could capture causality in action (being confident that under certain conditions, certain people would do certain things) without understanding the significance of what people were doing as they saw it. Later scholars expanded on Weber to put ideational work in its own interpretive or “hermeneutic” category, setting it off from the causal dynamics that non-ideational explanations of action ostensibly share with the natural sciences.
This move is tied to the notion that much ideational scholarship asks “how” or “what” questions in a “constitutive” mode, creating a division of labor with the “why” questions posed by explanatory work. Culture, norms, ideas, identities, and other socially-constructed elements define certain realities and imbue them with meaning in inseparably constitutive ways. Explanatory scholarship plays out the more mechanistic causal workings within that context. We need constitutive scholarship, for example, to see how the norm of sovereignty constitutes the state. In constructivist claims this is not a separable, temporally-sequential, causal-explanatory relationship. The very minute that people accepted norms of sovereignty they looked around and saw states. Explanatory approaches can analyze cause and effect dynamics within that socially-constructed reality.
This article argues that neither rationale for separate realms of inquiry stands up to scrutiny. Weber created the “understanding” category by defining explanation as a purely correlational exercise. Following Hume, he included only arguments that infer causal laws from “cross case” patterns without elaborating the causal mechanisms of action. While some non-constructivist scholars admittedly retain this classic Humean definition of causality and explanation, diverse theorists like Jon Elster, Daniel Little, and Henry Brady and David Collier have all argued for definitions of explanation that include at least some attention to “within case” causal mechanisms. Causal mechanisms in explanations of human action necessarily make claims about mental processes and “understanding” (with the very rare exception of instinctual psychological arguments that short-circuit cognition entirely). If we accept this view of explanation—and I submit that it fits not just with a great deal of sophisticated philosophical literature but also common sense—then interpretive arguments make claims over the same pathways as objectivist or “rationalist” ones. They compete directly (and may also be combined) in trying to theorize and document the mechanisms by which people arrived at certain actions.
Moreover, constitutive dynamics do not conjure up a weird mode of relations distinct from the sequential, causal creation of the present from the past that characterizes the rest of our universe. Constitutive relationships are themselves explicable, and also always carry causal-explanatory implications for action. The state and norms of sovereignty may co-exist inseparably, but at some point some mechanism brought about the first state-sovereignty system. In principle we should be able to explain how this happened. On the other end of the explanatory equation, using the label “constitutive” for a norm or idea clearly implies that there was a substantial difference between the actions we see and how people would have acted without the norm in place. For most views of causality and explanation (including even many Humean ones), that makes the norm a “background cause” of the actions—much like a variety of more conventional causes like market positions, possession of certain resources, levels of education, and so on.
If these points make sense, why have so many scholars so readily separated constitutive and causal argument for so long? In my view their discomfort in seeing constitutive background causes as explanatory flows from the logically unavoidable role of contingency in such arguments. Full-fledged constitutive arguments necessarily incorporate creative or accidental leaps across contingency. If the norm of sovereignty were fully explicable from pre-existing conditions, this derivative, inevitable by-product of something else would not deserve the heady label “constitutive.” The whole point of calling it “constitutive” is that many courses of action were objectively possible until, for underdetermined reasons, partly-free agents constructed a particular set of interpretations around themselves. Thus social construction constituted one kind of arena and actors from a wider range of available options. Understandably, perhaps, theorists of all inclinations have sensed this underlying contingency and have been reluctant to qualify arguments founded on initial indeterminacy as “explanations.”
Once we accept a definition of explanation that includes causal mechanisms, however, we encounter strong reasons to accept as explanations some arguments that include explicitly underdetermined, contingent leaps. If explanation of action is not simply a search for patterned regularities but also about uncovering the process by which actions come about, it is difficult to rule out that the real process by which some actions come about might include genuine contingency. Only if we insist on a totally deterministic view of history is it then reasonable to exclude from “explanation” arguments that include some delineated contingency. We must at least allow that the claims that best capture how and why something happened—and how they fit (or do not) into patterned regularities—might include some range of contingent leaps.
I refer to this sort of explanation that underlies constructivist scholarship as “particular explanation.” It is “particular” in explaining certain actions as the result of earlier contingent developments that we would not expect to turn out the same way even under identical conditions. Particular explanations are built on things that did not have to happen the way they did according to some correlational pattern, even probabilistically—but because they did, other things followed (deterministically or probabilistically) and those other things can be explained as the consequences. In other words, they focus on the causal consequences of resolved contingencies. The deepest claim of all constructivist scholarship takes this format. It argues that some range of historical outcomes was open until people embedded themselves, through their own creative or accidental actions, in distinct new socially-constructed dynamics. Constructivists have tended to focus on socially-constructed dynamics once they are in place, which is why they have the impression of dealing with constitutive logics that are inseparable from certain actions and not classically causal. But their arguments directly imply this sort of causal-explanatory claim over a longer time frame.
The overall argument, then, is that closure of the Weberian divide, the possibility of direct engagement between constitutive and causal claims, and the vocabulary of particular explanation all follow from a definition of explanation that a great many political scientists are surely already willing to accept. We are all asking questions that connect to debates over the same explanatory territory. For mainstream non-ideational scholars, this cuts away some of the main reasons why they have felt comfortable with research designs that do not include ideational alternatives. For theorists who are interested in social construction, the upshot is that they must continue to expand their engagement with non-ideational claims. At the very least, the first generation of constructivists has established as abstractly plausible that social construction is an irreducible cause of major variation in politics. Only if all theorists engage empirical arguments for and against this possibility can we proceed in a remotely scientific way.