EVALUATIVE CRITERIA AND EPISTEMIC COMMUNITIES 101Admittedly, drawing on texts is an imperfect indicator of teaching and research practices. I
have no indicators for their course adoption, and even that is a weak indicator of how they are
used in coursework. Indeed, it might be argued that the particular terms I identify are prevalent in
texts solely because of their rhetorical character, rather than because they are the ones most widely
used in empirical research. For these reasons, the list offered is a suggestive, rather than a defini-
tive, representation of usage in contemporary interpretive empirical research. At the same time,
their rhetorical character, the intuitive appeal of a term, is part of what I analyze below on the
grounds that this quality is vital to understanding criteria usage within an epistemic community.^12
Seven terms or concepts emerged from this analysis and, based on the index test, were divided
into “first-order” and “second-order” terms or concepts.^13 The four terms I am calling first-order
are ubiquitous in the criteria literature and are readily found in text indexes: thick description,
trustworthiness, reflexivity, and triangulation. The three concepts I deem second-order—informant
feedback/member checks, audit, and negative case analysis—are widespread in their usage, al-
though they lack the ubiquity of the first four in that no single term has emerged that captures the
amalgam of research practices and expectations each of these concepts represents. Instead, di-
verse terminologies and practices are associated with each of them, and these are my labels for
designating the ideas associated with each concept. For example, one might find “negative case
analysis” in some indexes, “outlier analysis” in others, and “rival explanations” in still others,
depending on disciplinary traditions or idiosyncratic factors. The practices to which these three
terms refer, however, are described in many of the texts, and they can be encapsulated, for present
purposes, in these particular labels.
First-Order Terms: Thick Description, Trustworthiness, Reflexivity,
and Triangulation
“Thick description,” Gilbert Ryle’s term borrowed by Clifford Geertz (1973a) to characterize ethno-
graphic writing, has taken on the aura of a “standard” by which to recognize and judge interpretive
research; it is part of what scholars have come to expect when reading an interpretive study. The
term has come to refer to the presence in the research narrative of sufficient detail of an event,
setting, person, or interaction to capture context-specific nuances of meaning such that the researcher’s
interpretation is supported by “thickly descriptive” evidentiary data. The subtext of this wealth of
detail is the provision of evidence that the researcher was, in the original case of ethnography,
actually present on-site, an eyewitness to the events, setting, and interactions described. The term
can be extended to assessing other methods, such as historical or document analysis, because its
purpose is not an exhaustive listing of details but a nuanced portrait of the cultural layers that inform
the researcher’s interpretation of interactions and events—supporting the researcher’s claim, for
instance, that what she saw was a “wink” and not a “blink,” in Geertz’s famous example, borrowed,
also, from Ryle (Geertz 1973a, 6–7); or, in the case of document analysis, supporting Jackson’s
(chapter 14, this volume) claim that “Western Civilization” (Abendland) had particular, strategic
meanings in post–World War II German politicians’ debates over reconstruction.
“Trustworthiness,” introduced by Lincoln and Guba (1985) and now used widely, is an um-
brella term that captures the broad problem that research needs to be seen as trustworthy—that is
what the concerns for “reliability” and “validity” are all about. It offers a way to talk about the
many steps that researchers take throughout the research process to ensure that their efforts are
self-consciously deliberate, transparent, and ethical—that they are, so to speak, enacting a classi-
cally “scientific attitude” of systematicity while simultaneously allowing the potential revisability
of their research results. As a tool of assessment, it facilitates discussion of criteria for judging the