266 ANALYZING DATA
But there is more going on here than might be apparent at first. In a way, Schumacher’s out-
burst fed directly into a more or less deliberate strategy on Adenauer’s part to cast Schumacher as
a nationalist alternative to his policy of cooperation with the “Western” occupying powers (Schwarz
1995). At the same time, Schumacher’s public stand on this occasion paralleled the stance taken
by the Communist Party, which implicitly reinforced the argument often advanced by Adenauer
and other conservatives that all Marxist parties were the same: materialist, nationalist, old-fashioned,
and unfit to govern responsibly. On most occasions, “socialists... were the most effective anti-
communists,” arguing that Marxist thought did not necessarily point in the direction of commu-
nism and providing by their own example an alternative path (Mark 1989); Schumacher was no
exception. But in this instance he made common cause with the more radical parties of the left,
simultaneously undercutting his presentation of social democracy as an alternative to commu-
nism and displaying a hot-headedness that resulted in a suspension from the floor of the Bundestag
(the German parliament).
How are scholars to make sense of this emblematic moment? The configuration of social
resources that made it possible for Schumacher to accuse Adenauer of being a servant of the
occupying powers is not the only configuration that might have existed; nor was Schumacher’s
accusation (or Adenauer’s subsequent response) somehow inevitably produced by that configu-
ration. Instead, the observed patterns were more contingent, depending simultaneously on the
ways that the historical actors interpreted their situations and on the way that subsequent scholars
have interpreted their interpretations.
I suggest that scholars turn to a configurational analysis of social action. Adopting such an
approach would preserve agency at the level of methodology and empirical practice, and not
merely by conceptual or theoretical fiat. Social actors in this conception are like the “cognitive
bricoleur” described by Roy Bhaskar, “the paradigm being that of a sculptor at work, fashioning
a product out of the material and with the tools available to him or her” (Bhaskar 1989, 78).^4 The
work of empirical analysis, then, should involve delineating the resources available and tracing
the ways that they are deployed in practice, while sticking close enough to the data that statements
about available resources have more of an empirical than a conceptual character.
For example, from this perspective, the parliamentary exchange appears as a legitimation struggle.
Both Adenauer and Schumacher were faced with problems of legitimation: In order to subsist in
their positions and to accomplish policy goals, they needed to provide publicly acceptable justifica-
tions for their preferred courses of action.^5 My central analytical claim is that legitimation is con-
strained by the available configuration of publicly shared “rhetorical commonplaces”—those vague
notions that command more or less general assent in the abstract but that stand in need of detailed
specification before they can be determinately linked to specific courses of action (Laffey and Weldes
1997; Ringmar 1996; Shotter 1993b). The work of legitimation involves deploying the available
commonplaces so as to rule out alternative courses of action and to justify one’s own preferred
option. And the stakes are high, inasmuch as the process of legitimation exercises a profound causal
impact on social and political outcomes (P.T. Jackson 2003).
Scholars engaged in such research are faced with what Anthony Giddens would call a “double
hermeneutic” problem (Giddens 1984): They have to interpret what is said by historical actors
while keeping firmly in mind the fact that what they are interpreting are interpretations of situa-
tions that those actors themselves have made. Both of these interpretations feature creative uses
of rhetorical commonplaces. Even if the historical character of the case under investigation pre-
vents these two hermeneutic circles from interacting in a causal or reciprocal sense (Hacking
1999), a thorny methodological problem remains. The challenge is to incorporate the active char-
acter of both streams of interpretation simultaneously, and not to reify one for the purpose of