To start with a humble confession: I am skeptical of the effort to state definitively what is qualitative about qualitative research. Must a field of study reach unanimity on the definition of its master term? If so, then the sociologies of culture, politics, religion, and social movements are all high-risk operations. Aspers and Corte (2019) helpfully acknowledge that the term “qualitative research” emerged as a sub-disciplinary boundary marker, not necessarily a signal of shared epistemological principles. It served, a half century ago, to mark off a turf distinct from that of the statistically driven search for generalizations about populations—sociology’s newly dominant mode. “Qualitative” is a now rusty road sign. The trouble, and intrigue, begin there.

Definitions focus our attention and potentially they help reduce intellectual vertigo. “Qualitative” has worked as a parameter-setting term of convenience—and sometimes we need those. I just am not yet convinced that defining the term more precisely would reduce whatever skepticism some non-qualitative researchers may harbor about this kind of scholarship. There also is the danger that even a carefully considered effort like that of Aspers and Corte would tempt us to truncate or freeze rather than improve research we call qualitative. I do infer that the authors’ work is motivated by two questions I share: What are qualitative researchers trying to do? How should we distinguish strong qualitative studies from weak ones? It is in the larger spirit of the authors’ work, then, to picture an alternative way of addressing those questions—one that would need far more space for a full, systematic justification.

Let’s start from the premise that qualitative is a kind of freeze-frame. It selectively captures and obscures a congeries of distinct research practices that have their own histories, have co-evolved together over time but do not have a simply logical or natural relation to each other. Stare at “qualitative” closely and you see it decomposing into a variety of logics and techniques, some of which will fall out of the frame of any compact discussion of qualitative methodology, including Aspers and Corte’s. This one will be no exception. For brevity’s sake, it will follow the convention of many sociology course curricula and limit the tag “qualitative” to participant-observation and interview research, much as the universe of qualitative research is much bigger and ought not be equated with research techniques alone.

Suppose that we unfreeze “qualitative” and imagine qualitative sociology as an expanding community of inquiry. What participants in this imagined community share is a dedication to interpretive social science. We aim to illuminate meaningful social action. Aspers and Corte (2019, 147) briefly consider and dismiss the possibility that qualitative research should be defined as interpretive research. They point out that quantitative research involves interpretation too; we have to interpret the results of a regression analysis, for example. True, but that is giving up too soon. Any study’s findings need to be interpreted for it to become significant to a scholarly community, but not all studies make it a central goal to represent meanings and meaningful action. I want to propose that we benefit from shifting the emphasis on what qualitative research is toward the question of what constitutes good interpretive social research in sociology. It still makes sense to link qualitative sociological inquiry to the larger, older project of the Geisteswissenschaften. I hope the brief illustrations and bit of terminology offered here can be useful in situations where we rely on or resort to the term qualitative as a tag for a moving, expanding target. First, let’s see what happens if we ponder the authors’ own proposed definition in light of my proposal.

Trying Out a Hunch about a Definition

The definition that Aspers and Corte offer is not wrong in the sense of misrepresenting what many qualitative researchers do. It carries some virtues too. It simply underestimates the context that suffuses those research acts and makes them into meaningful practices as well as practices that explore meaning. In the authors’ definition, qualitative research is “an iterative process in which improved understanding to [sic] the scientific community is achieved by making new, significant distinctions resulting from getting closer to the phenomenon studied” (2019, 155). This sounds close to a restatement of the venerable, constant comparative method that many if not most ethnographers and interview researchers use. That the authors’ carefully coded sample of 89 methodological statements leads to this definition is testament to a certain consistency among scholars who identify as qualitative researchers. Add in the fact that researchers with substantially different goals—Anselm Strauss (Glaser and Strauss 1967; Strauss 1987; Strauss and Corbin 1990) and Michael Burawoy (1998) for example—affirm this iterative process of discovery, and it becomes reasonable to think that qualitative research is a stable thing with a core, not only a name for an evolving collection of moving parts. But let’s look more closely at each part of the definition.

One big reason that self-identified qualitative researchers engage in an iterative process is that we aim to do something different from testing fully elaborated hypotheses at the outset. We suppose that there is something new to see and hear.Footnote 1 But we don’t just go and discover. We keep going back, watching and listening, taking it down, coding provisionally, dumping hunches and launching new ones for the next field visit or the next interviewee: the constant comparative method. We pursue an iterative process in large part because to understand what we are observing, we need to capture other people’s meanings accurately and figure out which meanings are the most relevant for our research. We want to discern how other people make associations—rather than associate their speech and action immediately with our pre-given analytic categories. It takes multiple observations to recognize and distinguish patterns of meaningful action, even apart from the second-order interpretation we do to set those patterns inside an analytic story that will interest our community of inquiry.

Part of that iterative process, as Aspers and Corte point out, involves making new distinctions. Of course, research that tests hypotheses with statistically measured variables also discovers distinctions; these may become clearer through iteration that happens over the course of separate studies. To be fair, the authors want to reserve “new” for those distinctions that emerge from the observing, coding, and hypothesizing process in one study, not ones that result from applying terms from a pre-existing literature to an unchanging set of data. Steering clear of the epistemological thickets surrounding the question of what is truly “new,” I just point out that qualitative researchers make new distinctions because we want to grasp meaningful action that we otherwise would miss with analyses that bracket meaning. One fine example is sociologist Brown-Saracino’s study (2009) of gentrifiers. It distinguished different, meaningful and consequential social worlds of neighborhood migration and identification that disappear inside purely political-economic concepts which, by themselves, do not sensitize us to meanings.

The notion that distinctions emerge from “getting closer to the phenomenon” plays in the same way. It does describe something that participant-observers and interviewers do on the way to producing research contributions. But getting closer makes sense in the context of a search for meaningful action. When interview research gets “closer” to family life (for example, Blair-Loy 2004; Hochschild 1989), we see how family members arrange couple-hood and work outside the home. If we were not interested in the moral and affective meanings that make “arrangements” into juggling, sacrificing, committing, or finessing, then we would not need anything other than time-use diaries.

Lastly there is the “improved understanding” that the research community gains from closing in on a phenomenon and discerning distinctions. I am glad that the authors do not make improved understanding contingent on particular kinds of conceptual frameworks or epistemological foundations. They don’t instruct, as Glaser and Strauss (1967) did in their foundational statement, that the iterative research process necessarily should lead to “grounded theory” as an end in itself. Aspers and Corte imply and I agree that what counts as improved understanding depends on assessments by communities of inquiry. We can add that the scholarly meanings shared and sometimes contested in those communities inflect each step of qualitative research, from research design to decisions about what counts as a finding or a contribution (Lichterman and Reed 2015; Lichterman 2015).

What Is Good Interpretive Research? Evolving Standards

In that spirit, I want to shift from what is qualitative to the expanding collection of postulates and practices we associate with good interpretive research—again, restricting our discussion to participant-observation and interview studies. These components have somewhat autonomous histories. None are just natural concomitants of a single, qualitative approach. In short, contemporary interpretive social research is guided by three standards, each of which has developed in an expanding arena of dialogue and debate. To meet those standards, interpretive researchers use diverse logics of inquiry to seek out and organize findings that result from distinct observation techniques. Talk of “standards” may sound disciplinarian, restrictive. I stress that I do not mean obeisance to hidebound scholastic traditions or reliance on exactingly precise metrics. Qualitative researchers, like other practitioners of a craft, assess their own and others’ works and are subject to assessment by others; teaching and publication are two among other sites for those judgment calls. What do we go by?

Maybe chief among the standards for qualitative work is interpretive validity. Qualitative researchers aim for illuminating, well-evidenced claims about how other people or collectivities make associations between symbols, people, and practices. In the 1950s and 1960s, researchers understood the search for actors’ meanings within a particular, widely diffused “theory/methods package” (Clarke and Star 2008, 117). That package has habituated several generations of researchers to the idea that participant-observation and interview research naturally proceed in a symbolic interactionist vein with the goal of discovering emergent, locally shared meanings. Many textbooks have packaged qualitative research that way. Since then, studies increasingly show that actors import, appropriate, and innovate with pre-existing meanings and not only generate new ones in interaction. These efforts since the late 1970s have drawn on varied theories of meaningful action and culture. Paul Willis’ (1977) modern classic Learning to Labor, for instance, found the meaning of teenage boys’ classroom defiance not only in a local, well-bounded subculture but a larger, racialized, and gendered working-class culture that, in his account, eventually delivered working-class boys to working-class jobs. My own field work with grassroots environmentalists heard a larger cultural formation, in this instance the American culture of personal authenticity, inflected in a surprisingly political way in local sites (Lichterman 1996). Contemporary qualitative studies use analytic grids that sensitize us to more as well as less institutionalized meanings, more and less subcultural or emergent meanings, expanding the questions we may pursue and the claims we may make.

To achieve interpretive validity, in turn, a qualitative project needs to do well by another standard. I will call it “groundedness,” not to be confused with “grounded theory.” By groundedness I mean the standard that interpretive social researchers share with others in sociology, that claims need to be supported empirically by evidence appropriate for the claim. What counts as good empirical grounding? This is another moving dialogue over time. Social researchers of the “first Chicago school” in the 1920s constructed cases of urban social characters from a great variety of evidence—hearsay from neighbors and police reports as well as conversations with the actors and observations in situ (Platt 1983). By the time the classic study Boys in White (Becker et al. 1961) was published, a different methodological self-consciousness led the researchers to ground their arguments about the everyday meaning of being a medical school student with data from students’ unprompted conversation over lunch in the cafeteria, rather than relying purely on answers to questions the researchers posed to the students.

Claims achieve groundedness through various logics of inquiry. Yes, there is the constant-comparative logic, which helps us support and specify our claims with evidence, whether the claims are thickly interpretive or not. But increasingly in the past two decades, interpretive sociologists are using and innovating other logics. These too involve iterative work but not all the same kind; they bid us organize observations and comparisons differently. Some studies take up emergent clues and embark on a trail of negative comparison cases toward the discovery of a causal relationship (Katz 2001, 2002). Yet it also is possible to use analytic induction’s negative comparison logic in concert with a more theory-driven logic of contrastive explanation: By trying out a succession of contrast cases, the researcher determines that a causal mechanism known from previous research can provide the answer to a question informed by classical theory, such as why do some volunteer projects produce Tocquevillian, virtue-broadening ties while others orchestrated by the same organizations do not (Lichterman and Reed 2015). Earlier statements of the analytic inductive logic may not have envisioned this more theory-driven articulation, and that is the point. We continue innovating and carefully combining logics of inquiry, making it credible to address questions which, at an earlier point in the field’s development, could have sounded too big or “too theoretical” to be grounded convincingly with ethnographic or interview evidence. All the more reason to see standards as evolving criteria, not rigid dictates for what constitutes a good study.

This last example implies a third standard, once suppressed by the traditionally taught notion that qualitative research produces discoveries “from the ground up.” A researcher’s communities of inquiry cultivate what I will call theoretical imagination. They expect a good study to be able to answer the “so what?” question (see Lareau 1989). Some communities expect critical demystification, others want a law-like pattern, or a big gold nugget for our storehouse of enduring facts. These are just three kinds of theoretical contribution—if by theory we mean the conceptual terms we use to discuss the significance of empirical cases (Lichterman and Reed 2015; Lichterman 2017). Communities of inquiry also expect a study’s evidence and theoretical framework to work well together. Review for journal publication regularly demands this nimble dance.

In the past 30 years, an expanding collection of logics has given qualitative researchers new options for theoretical imagination, beyond what the previously more dominant “theory/methods package” allowed. One might use the extended case method, seeking out findings that can work as anomalies that motivate improvements in a pre-existing theory, enabling it to “extend” to an initially confounding case (Burawoy 1998). It is important if perhaps counterintuitive to see that the production of grounded theory depends on theoretical imagination, too, even if its progenitors took a skeptical stance on other kinds of “theory.” Grounded theory is not just the automatic result of iterative observations, counter to the image suggested in its foundational text (Glaser and Strauss 1967). There is still a community of inquiry that vets and decides whether or not a researcher’s new, grounded theory is insightful, applicable to other inquiries, worth the trouble, publishable. The researcher needs theoretical imagination to identify a research literature to which the study at hand can effectively add newly constructed social types or processes—three worlds of neighborhood resettlement rather than simply “gentrification,” for example. The renaissance of interest in abduction (Swedberg 2014; Timmermans and Tavory 2012) offers yet another way to give a study theoretical imagination, by framing findings as the solution to an empirical puzzle that an initial, theoretical picture made significant to begin with (see Lichterman and Reed 2015).

Finally, interpretive sociologists produce findings through observation techniques. This is what “qualitative sociology” conventionally connotes (Swedberg 2021). These, too, have evolved, especially as they interanimate with logics and standards of the craft. In an earlier era, researchers often used participant-observation and interview evidence interchangeably in articles and monographs. Slowly diffusing insights from sociolinguistics, cognitive sociology, the work of Erving Goffman and other sources have made indiscriminate mixing less defensible. More recently, qualitative researchers are agreeing that interview and participant-observation techniques produce different forms of data with different purposes, virtues, and limitations (Lamont and Swidler 2014). Innovation and debate over techniques continue moving and redrawing the “target” of qualitative inquiry. Recently, some researchers ponder whether research on Internet communication should be considered ethnography or textual analysis, and when and where that distinction is still relevant (Hine 2011). An even newer conversation is showing how logics of comparison and explanation that were developed in ethnographic research can work in historical inquiry too (Reed and Lichterman forthcoming).

Conclusion

Suppose we follow my proposal and think “interpretive” when we say “qualitative,” at least when discussing studies that rely on participant-observation or interview techniques. Then we see that qualitative sociology changes as communities of inquiry continue cultivating standards, logics, and observation techniques for investigating meaningful action. That is why it is best to content ourselves with a rough, capacious working definition, a conversation starter that locates the study at hand in a set of research traditions. Then we can go on to say much more about how we did the research, how it meets standards as we define and combine them. Social science is a dialogue, including a dialogue about how we are creating knowledge claims—not a guidebook for producing facts. If that sounds radical it should not; it is just to take seriously the idea that social knowledge is an historically contingent, social construction.