EVALUATIVE CRITERIA AND EPISTEMIC COMMUNITIES 109The related move, which strengthens the previous one considerably, is to note the role that
human meaning making plays in action. Human meaning making and beliefs are understood as
“constitutive of actions,” and this view of causality can be used to explain not only individual
actions (see Jackson’s discussion of the beliefs and strategic actions of key German leaders,
chapter 14, this volume) but also the “broad patterns of behavior associated with social move-
ments” (Bevir, chapter 15, p. 285, this volume). In sum, interpretive researchers have a lot to
contribute to understandings of causality in ways that broaden the conceptualization of causality
beyond the variables-based, explanation-prediction, general law model.
Finally, what of generalizability? Campbell and Stanley introduced the cognate term “external
validity” and identified a supposed trade-off in research design: Experiments are strong on inter-
nal validity (causal inference) and weak on external validity (generalizability), in their view,
whereas surveys are strong on external validity and weak on internal validity. Just as they did
with “internal validity,” they offered ways to “design away” “threats” to external validity. Lin-
coln and Guba’s (1985) response to Campbell and Stanley was to argue that such thinking about
the criterion of generalizability misunderstands the research process and the use of research find-
ings. Whether research findings from a particular study should be “generalized” to another set-
ting should, logically, be the responsibility of the person who seeks to “transfer” those findings to
the new setting. (Based on this logic, they rechristened the criterion of “external validity” or
“generalizability” as “transferability.”) The responsibility of the researcher is to provide suffi-
cient “thick description” so that others can assess how plausible it is to transfer findings from that
research study to another setting. It is just such an understanding of, and emphasis on, context—
and of the ways in which context and Sherlock Holmes causality are intertwined—that will en-
able others to build on the research findings they find trustworthy. Consistent with the
methodological ferment occurring in contemporary qualitative-interpretive methods,^31 Adcock
(chapter 3, this volume) lays out yet other ways in which interpretive scholars may relate particu-
lar studies to more general, historical trends and concerns.
CONCLUDING THOUGHTS: EPISTEMIC COMMUNITIES
AND JUDGMENT
“Judgment calls”... refers to all of those decisions (some big, some small, but all neces-
sary and consequential) that must be made without benefit of a fixed, “objective” rule that
one can apply, with precision, like a template or a pair of calipers.... We suggest... that a
set of rules to replace judgment calls not only would be difficult to fashion, but also would
be dysfunctional if we had them.
—Joseph E. McGrath (1982, 13–14)Today, in a number of social science disciplines, the possibility of judgments about research
quality based on “a view from nowhere” (Haraway 1988) has been replaced by the understanding
that such judgments take place within epistemic communities. Cross-epistemic judgments may
sometimes be necessary and even desirable, but, as the thought experiment opening the chapter
illustrated, their legitimacy requires a reviewer with an understanding of alternative research ge-
stalts’ distinctive practices and purposes. Whether within or across epistemic communities, how-
ever, judgment cannot be escaped, and the desire for templates, calipers, and algorithms for
judgment may be indicative of the very human fear of shouldering responsibility for consequences,
big and small. It is perhaps psychologically easier to point to some list of criteria presumed to be
universal than it is to say, “In my judgment, in the context of the practices of this epistemic
community, for such and such reasons, this study is inadequate.”