NEITHER RIGOROUS NOR OBJECTIVE? 85student’s progress in studying a preserved fish, he said, “a pencil is one of the best of eyes” (L. Cooper 1945,
chapter 7. “How Agassiz Taught Professor Scudder”).
- Rigor is also a term in mathematical reasoning, in which it shares with formal logic an attention to
logical processes of deduction, especially in establishing mathematical proof. - The commonsensical understanding of rigor as “stepwise” may derive from this sense. The defini-
tions risk becoming circular in their use of “logic” and “valid,” and so it might be worth reminding ourselves
that logic (from the Greek “logos,” meaning “reason”) is a system of reasoning whose study focuses on the
structure of propositions as distinguished from their content. That is, “formal” logic is the logic of forms or
structures of language. - This formulation invokes the old problem in analytic philosophy of the tension between formal
logic and substantive “truth.” That is, one can construct arguments in which the logic of the forms of
statements is valid—the conclusions do, indeed, derive from the premises—but the content of the state-
ments is patently absurd. - Mark Bevir (2003) has referred to what I take to be the same characteristic as “philosophical rigor.”
By derivation, this definition of structural rigor evolves to a more procedural methodological one, albeit one
closer in sense to the structural definition than to the previous definition: Rigor means adherence to proce-
dures that have been generally accepted as leading to correct conclusions. The defense of interpretive re-
search in these terms is the same as the defense presented there: Good interpretive research adheres to
accepted procedures. The difficulty is that these procedures have not, by and large, been explicated in ways
accessible to those outside of the epistemic communities conducting such research. - Taste in such matters is highly individualistic. My personal list includes the writings of Clifford
Geertz (e.g., 1973a, 1983) and John Van Maanen (e.g., 1978). - I thank Tim Pachirat for enabling me to see these aspects of my argument. As he reminded me,
McCloskey (e.g., 1985) has also noted this in the context of writings in economics. The essays in Simons
(1989) explore various social science fields. The ways in which researchers’ writing itself can constitute the
subject being studied has been treated by Clifford and Marcus (1986), Geertz (1988), Golden-Biddle and
Locke (1993), Van Maanen (1988), and Yanow et al. (1995). - The concept has been debated and discussed at length. See, e.g., Bernstein (1983), Bevir (1999),
Latour (1999), and M. Weber (1946). For a treatment oriented toward language and objectivity, see Lakoff
and Johnson (1980), chapters 25–26. - This is the point that Werner Heisenberg argued against in the context of physics research, named the
“uncertainty principle”: that, in quantum mechanics, the act of measuring the physical characteristics of
one entity (such as the speed of particles in motion) itself potentially alters the behavior of the very things
and processes one is observing, thereby affecting the knowledge the researcher could claim. This was
enacted in human terms in the stage set for the San Francisco production of Michael Frayn’s play Copenhagen
(1998). Heisenberg went to Copenhagen during World War II to visit the senior physicist Niels Bohr, with
whom he had studied prior to the war, for some purpose related to the development of the atomic bomb. No
record of their meeting has, so far, been discovered; the play script speculates as to what transpired. The
scene design has the two men, walking in the woods to escape having their conversation overheard by
Nazis, who, they assume, have bugged Bohr’s home, move around and around in a circular space bordered
by a high wall that serves as the foundation for on-stage bleacher seating. It is as if the audience members
seated in the risers are observing the interaction of two human atoms, whose words and deeds bounce off
each other in unpredictable—and, today, unknown—ways. - For instance, laundry hanging from a line might be “read” for purposes of socioeconomic class or
some other analysis—how many items of what category of clothing or linens, in what condition (having
holes, marked by repairs), and so on—without having to talk to the person who hung the wash or the one
who owns it. - Indeed, in the introduction to the 2000 reissue of the book, Schwartz and Sechrest (2000, xi–xii)
wrote: “As we review the references to Webb et al., we reach the conclusion that it [the book] clarified a
pervasive problem in social science research and that it proposed a plausible remedy.... The problem was
one of validity.” The solution required “a new mind-set in planning and carrying out research,” one that drew
on multiple methods and creativity in going beyond conventional data-gathering techniques. - Kuhn (1970, 5) described research during the “normal science” phase as “a strenuous and devoted
attempt to force nature into the conceptual boxes supplied by professional education.” Latour (1999) takes
issue with the language of “filter,” commonly used in this context, which he sees as implying a passive
acceptance of external reality. I think this problem is resolved by conceptualizing the process as a “sorting”