318 ANALYZING DATA
have a carefully designed research plan, but the plan did not preclude the surprises that waited for
us in the field.
Our research was funded by the National Science Foundation, which requires a detailed pro-
posal.^1 Our first proposal was not funded in part because the research plan was, for NSF review-
ers, inadequately specified, a common problem for interpretive research. The second and funded
proposal described in considerable, even tedious, detail each step in our research process, from
entering the field to collecting and interpreting the narratives. The more step-by-step research
design in the second proposal did not eliminate the surprises and the necessary meanderings and
dead ends that waited for us in the field, but it did provide an essential guide for us as we imple-
mented our research. Counterintuitively, our research plan, which gave structure and direction to
the entire project, also gave us the confidence to allow ourselves to momentarily feel baffled and
lost when in the field. In addition, it provided a common approach as different members of the
research team worked in different field settings.
Nonetheless, our encounters and observations in the field and our close reading of the narra-
tives generated during fieldwork continually stretched our capacity to make sense of all we were
hearing and seeing. Our process of interpretation was more one of coming to grips with surprises
than of confirming expectations, however carefully laid out in our grant proposal. We saw and
heard things we did not expect, such as the police officer deliberately ignoring a direct order to
prosecute a hard-working, illegal Mexican immigrant. We also did not see and hear things we
expected; for example, street-level workers did not reference policy implementation or discretion
in making judgments.
The very richness and complexity of field-based research—the reasons we are drawn to the
field—was unsettling because we heard unexpected claims about street-level work. Also, being
in the field was at times disorienting because the observation of the moment often seemed crucial
to making sense of our subjects and yet it was not automatically connected to what we heard or
saw immediately before or after. Nor was finding compelling patterns of street-level judgments
always made easier in the exchange of ideas between our subjects and us. Giving the subjects of
our inquiry a voice in the interpretive process, a crucial part of our pedagogy of inquiry as de-
scribed below, was sometimes humbling, as one is reminded continuously about the inadequacies
of one’s explanations. Perhaps the most humbling of all is telling someone in the field—in our
case, police officers, teachers, and counselors—of some hard-earned insight and being met with
a polite “that’s obvious” stare.
Moreover, in our desire to forge compelling interpretations, there is a continual risk of assert-
ing more clarity and pattern than actually exist in social life. John Gilliom makes the essential
points: “If we are truly listening to lots of different people, we get lots of different stories, and it
is often impossible—and certainly inappropriate—to assert a universalizing meta-narrative over
their tales. Too much sense-making can be a real problem” (personal communication, August 17,
2001). These cautionary tales notwithstanding, we experienced interpretive, field-based research
as having a pungency and reality—dare we say a “validity”—that is hard to match in more tradi-
tional forms of social research. Close, intense encounters with research subjects in their work-
places provide a solid basis for forming and questioning field-based interpretations.
The research that forms the basis of the interpretive claims rendered in our book and other
writings from this project involved fieldwork at five sites in two states with three different types
of front-line or street-level workers: police officers, teachers, and vocational rehabilitation coun-
selors.^2 The research team was composed of the two authors of this chapter plus our two research
assistants, Trish Oberweis and Suzanne Leland, who made essential contributions to all aspects of
this research. The work resulted in two books, two dissertations, and several articles (Maynard-