From the 14th Floor to the Sidewalk: Writing Sociology at Ground Level


I had difficulty responding to Kai Erikson's essay. I couldn't do so very directly. I found I had to open up and examine a fundamental divergence between the sociology he represents and the sociology I practice. The difference, of course, is one of how sociology is written though in a somewhat different sense of writing than that with which Erikson is concerned. His issues are primarily those of style; mine are of what sociology's lexical practices enable us to say. In the passages Erikson excoriates as examples of bad sociological writing, he focuses on them as problems of style. For me the kind of writing they exemplify represent problems of a different order.

In describing how sociological language works to create at once sociology's landscape and its segregation from ordinary language, Erikson uses a metaphor. He imagines being among people on the sidewalk in a city; there what becomes visible is the particularities of individuals. At street-level among the sidewalkers, “what the eye sees . . . at ground level is an immense scatter of persons who are moving to their own rhythms and living out their own lives.” But traveling up to the 14th floor and looking down, it is possible to see patterns and consistencies, even to make generalizations.

Erikson views the outsiders who are critical of sociological writing as missing the implicit logic of the discipline. Sociology does something distinctive; has its own landscape, its own form of reality. From the 14th floor, looking down on the people in the street, patterns and consistencies may become observable. Hence sociology must have a specialized, even technical, language. At the same time, sociologists can and should find ways of writing that are responsive both to its subject-matter and to those outside the discipline who may want to be informed by what sociologists learn.

If I track the 14th floor metaphor, I run into a major problem. In describing sociological writing, Erikson takes for granted having positioned the knowing subject so that she or he is looking down on the people in the street. He is attentive to how sociological language constitutes its landscape. He emphasizes how sociology depends on its distinctive language to construct a reality that sociologists can have in common but is not easily shared with others. At the same time, he does not address what it is about sociological language that elevates the sociological subject above the crowd. The shift must somehow be performed in language—how else? And though Erikson moves into elegant metaphor, avoiding sociology's plodding nominalizations (I think he takes pleasure in metaphors; he for sure uses them artfully), his metaphors presuppose the 14th floor move. For example, when he suggests that the sociologist looking down on the street from above may be able to discern “tides, currents, forces, pulls” at work among people, he does not attend to how the terms he uses objectify. As metaphors, they are vivid. But as metaphors recognizing patterns observable from above, they also realize a distinctive logic: each has causal power, precludes attributing agency to people, and arrogates the role of agency to an impersonal something. It is this logic that effected, in my view, the translation of the knowing subject to the 14th floor.

In an earlier article of mine, I wrote about the sociological language used in the study of organizations, As can be seen from the following discussion that draws on that earlier thinking, my own critique of sociological language both converges with and diverges from Erikson's. I fasten on a passage that Erikson might well have included among those he assaulted. But my argument focuses on what I can now call the 14th floor effect, that is, the language practices of sociology that achieve the transition from being among people to being above them. My comments are on the following passage from Ronald Jepperson's (1991:145) specification of “institutions”:

Institution represents a social order or pattern that has attained a certain state or property; institutionalization denotes the process of such attainment. By order or pattern, I refer, as is conventional, to standardized interaction sequences. An institution is then a social pattern that reveals a particular reproduction process . . . institutions are not reproduced by “action”—in this strict sense of collective intervention in a social convention. Rather, routine reproductive procedures support and sustain the pattern, furthering its reproduction.

No difficulty, clearly, in including this passage among examples of bad sociological writing. Erikson describes the latter wonderfully:

The problem is that these sentences lumber painfully across the page, bent low by the weight of all the unnecessary syllables and complex locutions they are being asked to carry. They are ponderous, convoluted, hulking, bovine, full of a contrived profundity.

Jepperson's sentences do all this. But I have a different direction:

. . . we can see in Jepperson's definition a struggle . . . to deal with the problem that institutions can exist only in people's activities and yet cannot be reduced to them. No phrase seems quite adequate. Each seems to require elaboration; each elaboration increases ambiguity. Institutions are a social order or pattern; “order or pattern” refers to “standardized interaction sequences;” yet institutions (i.e., order or pattern) are not reproduced by action but by “routine reproductive procedures” which support the pattern. How to distinguish routine reproductive procedures from standardized interaction sequences remains unexplicated. The struggle to conceptualize these objectified forms without reducing them to people's activities is achieved by jumping to the nominalized level of abstraction, bypassing the problem of how they exist. . . . Though institutions and organizations are defined and redefined, the problem of their ontology remains undecided. (Smith 2001:167)

I won't pretend that my own sentences don't lumber, but I quote this analysis because it travels in a different direction than the elevator to the 14th floor. The passage from Jepperson exemplifies more than an issue of writing style. I see him struggling with the problem of how not to write about people, how not to admit them to participation in the sociological event. These are established sociological conventions of writing. Writing in ways that conform with sociology's established lexical practices is difficult to manage, if only because people are always present and active in anything of which sociology writes. They’re always there, waiting to sneak into the sociological text behind the writer's stylistic back. It's difficult to contrive sentences that shut them out.

Few sociological writers are capable of writing as elegantly as Erikson does while at the same time preserving the logic of the 14th floor. And Erikson, as we know from his published studies, has no problem giving people presence in his texts. He is one of the masters of sociological ethnography. But yet, his recommendations return sociology to lexical practices analogous to those used by Jepperson. Though the latter makes no reference to “tides, currents, forces, pulls,” his concepts are equally effective in displacing the presence of people. There are, for example, patterns reproduced by procedures but not by people. Agency has been transfered from people to sociological concepts.

Sociological discourse (and I'm using the term discourse here much as Michel Foucault defined it) imposes an interpretive order on what people may have said to a researcher or what a researcher may have observed. Sociology relies on a language capable of dominating ordinary language and of selecting fragments from what people were saying to fit its concepts. Translating the logic of that language into metaphor does not change it. The presence of people as subjects speaking for themselves and from the actualities of their everyday lives is displaced. What people have to say is subdued very generally to the status of instances or expressions of the sociologist's interpretation.

The barrier to communication between sociologists and nonsociologists goes deeper than our competence as writers. Jeff Alexander (1989) has argued that the discipline of sociology does not hold an empirical ground in common (p. 21); it must rely on traditional theory for its very existence as an intellectual community (p. 27). In my earlier article referred to above, I suggest that sociology's lexical practices disconnect its texts from the actualities of which they claim to speak. The referential power of terms such as institution, forces, structure, rule, bureaucracy, and so on is ambiguous. The discipline constitutes its own form of reality without having undertaken the Galilean responsibility of anchoring statements about the world in observation. Or better, of making statements about the world that are entailed by observation.

For the last 20 or 30 years, I've been working with colleagues on an alternative sociology we now call institutional ethnography (Campbell and Gregor 2002; Smith 1987, 2005, 2006). Until Erikson supplied his brilliant metaphor, I hadn't quite realized just how radically it diverges from 14th floor sociology. In a sense, institutional ethnography is specifically designed to avoid the 14th floor and to remain among the people “at ground level.”

Institutional ethnography takes up and works from and with people's everyday experience; it recognizes that our everyday doings are coordinated with those of others in relations of which we are generally only marginally, if at all, aware. The everyday is deeply penetrated, organized, shaped by social relations that coordinate people's local doings with those of others, both those immediately present and others elsewhere or elsewhen. Take the sidewalk scene that Erikson draws for us. Of course, people's movement is ongoingly coordinated. Ethnomethodology has been interested in this kind of ordering (Garfinkel 2002). But institutional ethnography directs attention to how people's activities on the street (to follow Erikson's metaphor) are coordinated in social relations beyond it. If we start looking around, we could find many and points of entry into the maze of translocal coordination. People's hours of work, for example, are coordinated and turn up on the street very visibly in rush hour; there are proprieties and styles of dress and demeanor that reference the standardized images originating in the mass media (standards that are definitely not attained by the panhandler if indeed he's allowed to sit there against the wall of the bank with his paper cup set out in front of him); there's the materially realized conventions that divide sidewalk from street, and the travel of sidewalkers and traffic—a difference maintained by the city's engineering department and policed by the police as well as by informal appeals by drivers and pedestrians to rules that presumably are written somewhere in some legislation. City government also ensures that the sidewalk isn't clotted with waste plastic and paper, that the wastewater runs off when it rains, and so on. Someone sets down a dollar in the panhandler's cup and he moves off quietly to buy a coffee, entering thereby the complex relations constituted by the exchange of money and commodities. The people on the sidewalk are in various ways active in relations that are not fully visible on the street and are generally taken for granted. However, their ongoing interactions as sidewalkers are coordinated among themselves; they are also coordinated translocally, that is, across particular local settings where others are or have been at work. And no doubt there are other invisible presences that haven't occurred to me.

Institutional ethnography starts from people's everyday local experience and explores the translocal that is present in and organizes their everyday. Of course, it doesn't and can't explore everything. It begins always with what is of concern. It is a sociology that proceeds by inquiry; it relies on the local actualities of people's doings and how they’re coordinated to discover how those relations relevant to the issues of concern are being put together. The primary research focus may be much as in traditional sociological ethnography. The researcher is an observer, probably a participant observer of people's everyday lives and how they’re coordinated locally, making the relevant extra-local relations visible as they enter into the coordinating of people's work. See Tim Diamond's participant observation of the work of being a nursing attendant in old people's residences. His Making Grey Gold is a model. Or the focus may be on exploring how the institutional is organized so that it enters into people's everyday lives as it does. See Ellen Pence's examination of the workings of the judicial process of domestic abuse cases from the standpoint of the safety of women who have been abused. Research proceeds by inquiry; it explores and discovers, learning first from those whose standpoint in the everyday organizes the research direction and then moving, given the economy of research, to exploring and discovering how the present-in-the-everyday extra-local relations are themselves actually being put together. It includes the textual mediation of people's local doings as integral to the ethnography. Because it takes up what people are doing as individuals but always under the aspect of how their doings are coordinated, it doesn't get stuck with particularities. Because it has developed ways of discovering the translocal relations in people's everyday lives and of tracking their organization, it is also exploring the relations and organization of what might be otherwise called power, governance, and so on. The products of research are not, however, case studies because they are opening up ethnographically social relations that are generalized and generalized across local settings and particular individuals.

A sociology that stays with people's everyday life experience, as they know and report it (including how they read and take up the texts that enter into the organization of their work) does not even have to attribute agency to people. People never lost it and hence don't have to rely on the sociologist to replace it conceptually—as is proposed, for example, by Anthony Giddens. True, the research may be technical and at the research stage may not be readily translatable to outsiders, but what has been learned can be made readily accessible to nonsociologists. Institutional ethnography's research findings offer to people something like a map that extends what they know of their everyday world. The cartographic method may be technical and inaccessible to those unused to it, but it produces the equivalent of maps that people can take into their everyday worlds and read to find their direction.