COMPARATIVE-HISTORICAL SOCIOLOGYand deployed, but for the most part as interpretive
and organizing frames, or as lenses through which
history is understood and represented. Interpretive
comparisons often are thought impressive, but
sometimes of questionable validity due either to
their self-validating logic or to their lack of explicit
scientific criteria for evaluating the truth content
of the interpretation (Bonnell 1980; Skocpol 1984b;
Tilly 1984; Stryker 1996; Mahoney 1999).
Holistic comparisons usually are tied to par-
ticular theories and are used when a ‘‘social whole’’
such as a world system (Wallerstein 1974) is
methodologically posited. Conceptualizing the en-
tire world system in this manner suggests that
there is but one theoretical unit of analysis, and
that unit is the world system. What are considered
to be separate ‘‘units of analysis’’ or ‘‘cases’’ in
most comparative-historical strategies are, in the
holistic methodological frame, really only interre-
lated and interdependent historical realizations of
a singular emergent process or social system. Wheth-
er nation-states, regions, or cultures, these ‘‘mo-
ments’’ therefore are not the discrete and inde-
pendent units demanded by analytic formalism.
Analysts using holistic historical comparisons in
this manner often depend on an ‘‘encompassing’’
functional logic that explains similarities or differ-
ences among parts of a whole by the relationship
the parts have to the whole (Tilly 1984). Thus,
Wallerstein’s (1974) comparisons explain the dif-
ferential development of temporally and spatially
specific (though interdependent) economic units
in the world system—the core, periphery, and
semi-periphery—by their differential positions and
roles in the world economy. Such explanations are
often circular in their reasoning and always diffi-
cult to test, but they may serve as useful illustra-
tions of the workings of a social whole or of the
inner logic of a theory (Bonnell 1980; Skocpol 1984b).
Philip McMichael (1990) suggests ‘‘incorpo-
rated comparison’’ as an alternative to purge holis-
tic comparison of its mechanistic and self-validat-
ing logic. Here a kind of social whole is posited,
but it does not preexist and regulate its parts, as
does the world system. Instead, it is analytically
‘‘reconstituted’’ by conceptualizing different his-
torical instances as interdependent moments,
which, when cumulated and connected through
time and space, ‘‘form’’ the whole as a general but
empirically diverse historical process. Thus the
very definition and selection of ‘‘cases’’ becomes
the object of theorizing and research and not its
point of departure or merely the vehicle conveying
data for analysis. Karl Polanyi’s ([1944] 1957) analy-
sis of the emergence and decline of laissez-faire
capitalism is a compelling example of this form of
holistic comparison.Individualizing interpretive analysts use histori-
cal comparison to demonstrate the historical and
cultural particularity of individual cases. Analysts
in this tradition often choose research questions
and cases on the basis of their recurrent moral
and historical significance rather than scientific
representativeness or comprehensiveness. They
also tend self-consciously to eschew general theo-
ry, quantification, and the methodology of causal
analysis. Rather, interpretation is grounded in rich
historical narration, in anthropological sensitivity
to the import of cultural practice, in the culturally
embedded meanings in social action, and in the
use of persuasive concepts that crystallize or
resonate with significant social values and themes
(Bendix 1963; Geertz 1973; Bonnell 1980; Tilly
1984; Skocpol 1984b). Bendix’s ([1956] 1974) re-
search on work and authority in industry is an
exemplar of this strategy, and Daniel Goldhagen’s
(1996) recent study, Hitler’s Willing Executioners, is
another much more controversial example.Casual Interpretivism. Historical comparisons
rooted in causal interpretivism synthesize aspects of
both analytical-formal and interpretive-historical
comparisons. Inspired in part by Weber’s (1949)
formulation of ‘‘causal interpretation’’ (and an-
thropologist Clifford Geertz’s [1980] similar idea
of ‘‘interpretive explanation’’), this strategy is
stamped by explicit causal reasoning, the potential
for explanatory generality across comparable in-
stances, and the use of methodological procedures
allowing for strict replication, on the one hand,
and attention to historical narrative, cultural par-
ticularity, and subjective meaning, on the other.Causal interpretivism has been most fruitfully
developed and applied in the analysis of historical
narratives (Sewell 1996; Griffin 1992, 1993; Quadagno
and Knapp 1992; Somers 1998). Narratives are
‘‘sequential accounts’’ organizing information in-
to ‘‘chronological order to tell stories about what
happened.. .’’ and why it happened as it did, and,
as such, are ‘‘tools for joining sequentiality, contin-
gency, and generalizability’’ (Stryker 1996, p.305,
307). Narrativists conceptualize historical time so