The example comes from a relatively recent work by prominent second-generation Constructivist Audie Klotz, a chapter in her co-edited volume Qualitative Method in IR: A Pluralist Guide (Klotz and Prakash 2008). The chapter advises students on how to “do” case selection in qualitative work, which, tellingly, Constructivism is typically but quite incorrectly conflated with (see Rousseau 2006; Jackson 2011). We may, of course, be biasing our generalizations of second-generation Constructivism by using this chapter, our own form of selection bias. Nonetheless, we believe it represents an instructive example of second-generation Constructivism, one that was utilized in notable (and critical) characterizations of the “norms” literature in the late 1990s (see Frost 1998). It also represents the difficulties a generation (and members of its cohort) faces in attempting to “mellow” down its world (and methodological) views.
Klotz explicitly claims that her chapter is a more mellowed (2008, 48) take on the issue of methods than she espoused in 1995 (Klotz 1995). However, we also see a multitude of exclamation points throughout the chapter coloring her advice to students on how to do qualitative methods. Klotz uses her previous work on norms and apartheid as a foil through which to instruct readers on how to do Constructivist case studies and how Constructivism itself can thereby be “tested.” While finding theoretical claims from first-generation Constructivists “useful,” Klotz tells us that such work was really nascent Constructivism, because it “did not offer a specific theory to test. Indeed, it resisted the whole endeavor of testing theories in the conventional sense!” (Klotz and Prakash 2008, 51-52). Klotz argues that her book, via good case selection, actually did do this - testing Constructivism vis-a-vis Realism and Liberalism. Klotz argues, further, that a good qualitative study can even “disprove” a theory
So, was this a progressive development in the “research program” of “Constructivism?” Did the second-generation develop Constructivism to the point where it could be tested? In defense of first-generation Constructivists such as Onuf, Klotz provides us a reason for why Constructivism itself was not tested. It was not, as Onuf (and now others, particularly Patrick Jackson) would advise us, because the mind-world monist philosophical ontology underpinning Constructivism does not lend itself to testing (Jackson 2011, 113). No, it was because, as Klotz notes, the “value of case studies depends on the status of the theory that underpins it . . . in the late 1980s Constructivism had not been articulated to the point where it could have been tested” (2008, 52-53).
In Klotz’s more mature view, then, it is about status and articulation that makes a perspective testable, not the incommensurability of perspectives that they offer via the mind-world philosophical assumptions they make. In other words, this was not a progressive development in the research program of Constructivism. Instead, it was a writing against the first generation within a context - a milieu - of neopositivist expectation for what counts as “good” research. The style of the writing was also important, done with a confidence and certainty that post-second-generation Constructivists - the present authors included - would themselves write against. Onuf’s words in his Tinos lectures are important here:
The importance I attached to social construction in World of Our Making
prompted me to adopt a label - constructivism - already in use in social theory but not IR. Such a label implies a philosophical stance identified with Kant and frequently called constructivist. I have found these philosophical implications intriguing, perhaps to disadvantage in a world where philosophical realism prevails in many variations . . . Yet to ignore those implications allows the casual appropriation of the constructivist label by scholars who are not Kantian constructivists. In most cases they think that an interest in collective identity and informal rules (dubbed norms) instead of the legal rules means that they are somehow no longer liberal institutionalists. They are wrong.
(Onuf 2012, 19)
Our aim here is not to cast aspersions on any particular member of second- generation Constructivism. The second generation did IR scholarship, particularly but not exclusively in the United States, a great service by putting the social, culture, and hence moral and ethical aspects of international politics at the forefront of the field’s agenda. They are without question to be congratulated. Moreover, in so doing the second-generation Constructivists opened the way in the past decade for Constructivist-trained scholars, such as the current authors, to speak to others (including prospective departments of employment). Their work thus deserves our attention and praise. But such work is not all there is or was to Constructivism, as often appears to be the case when scholars cite “Constructivism” and claim that it all says what that second generation (especially) had to say.
Thus, Barder and Levine have it right about the 1990s Constructivists, namely that they carried a methodological position within a “historical moment” (of that decade) which instead “took on millennial significance” (Barder and Levine 2012, 603). The methodological position meant that the world “out there” could only be understood as being more closely approximated to the way the second-generation saw the world. The best way to “get their empirics on” with the other “isms” was to test the latter with their version of Constructivism, against a “world” that was developing in ways favorable to their Constructivist, and largely progressive, sentiments.
But as Jackson has so persuasively argued (2011, 113), these calls to test Constructivism, of Onuf or others, is really “absurd.” The second-generation had not discovered a way to test Constructivism. For Jackson:
Constructivism has often been charged by critics with failing to elucidate empirically testable propositions about world politics . . . in other words, constructivists are charged with failing to subject their scientific ontologies to the kinds of evaluations that are only meaningful within a philosophical ontology of mind-world dualism.
Barder and Levine note that other Constructivists “sustain a kind of humility that has eluded the [second-gen-Constructivist] theorists of the via media” (2012, 604). We would agree, but their argument needs to be pushed further. Thus, it is not just a political persuasion, but also an assumption, one that did not obtain for the first and the post-second generations, of a mind-world divide and a stylistic certainty on being right. It is that condition that defines the second-generation of Constructivism. For the post-second generation of Constructivists such as Barder and Levine, and many others collected in this volume and beyond, the appropriate stance towards our scholarship is precisely the opposite: mind-world monism as a methodological position and a “chastened” view of theory (see Levine 2012).
If this is the case, then it should direct our attention to what was done to work such as WOOM, and why, only now, we seem to be revisiting that work with a different, more appreciative, but still critical eye to restore it and do something else with it that makes more sense. In this way, Barder and Levine are also correct about this past decade, that the 2000s were a “moment of truth” for second-generation Constructivism. But they are so much more, for if generations are impacted by events “out there” as well as the writing against a previous generation “in here” (i.e., within the field of IR itself), then we might notice how events in the 2000s also pushed the post-second generation to write against the second generation as well. Thus the “mood” (see Klingberg 1952) of the 2000s that shifted this generation to a more circumspect, radically contingent, skeptical, even pessimistic, power-centric understanding of the world than their post-Cold War predecessors. This might also explain the affinity that exists between Onuf’s version of Constructivism and the version those working within Constructivism in the past decade have used to understand the world. The next section takes on this suggestion in relation to the recently proclaimed turns to “practices” and “relations.”