Abstract and Keywords
This article examines the unique contribution that analytic ethnography has made and can make to accomplishing two of the key principles of analytical sociology: developing theoretical explanations by identifying mechanisms that connect actors, action, and outcomes; bridging the micro, meso, and macro levels of analysis in those explanations. It first distinguishes ‘analytic ethnography’ from other varieties of ethnography before showing how analytic ethnography has historically developed mechanism-based explanations that go beyond the micro level. It then compares analytic ethnography to analytical sociology in order to highlight the compatibility of the two. Finally, it demonstrates how theoretical integration can be achieved first within analytic ethnography, then between analytic ethnography and analytical sociology, using research on signaling and explanations of outcomes in which signals are the mechanism.
Analytical sociology is designed to bridge the gap between theory and empirical research by incorporating action-based theories into explanations of macro-level phenomena (Hedström 2005). Because understanding macro-level outcomes is the goal, the principles of analytical sociology primarily have been used by mathematically oriented scholars employing statistical analyses, simulation models, experiments, and mathematical models, as most of the chapters in this volume indicate. Ironically, the fundamental principles of analytical sociology are the essence of ethnography. Analytical sociology urges a turning away from variable-centered approaches toward strategies that produce explanations of events and outcomes by revealing the mechanisms that generate the observed relationships. It is designed to examine key social processes and break them down into their primary components, showing how they work together to produce outcomes. The focus shifts from variables to actors, actions, and the ‘cogs and wheels’ that (p. 689) respond to and reproduce social phenomena. Arriving at explanation is the goal. In crucial ways, however, ethnography marches to a different drummer. Moreover, although all ethnography has important characteristics in common, within it are vastly different approaches to the social. No single work can be taken as representative.
How then might we understand ethnography in relation to analytical sociology? This chapter offers no primer on ethnography, nor does it revisit methodological debates or reiterate the instrumental view of ethnography as a supplemental data-collection method for generating hypotheses related to analytical sociology. Instead, I will focus specifically on the unique contribution ethnography has made and can make to accomplishing two of the key principles of analytical sociology: (1) developing theoretical explanations by identifying mechanisms that connect actors, action, and outcomes, and (2) bridging the micro, meso, and macro levels of analysis in those explanations. I will show that analytic ethnography has historically developed mechanism-based explanations going beyond the micro level and has more in common with analytical sociology than is usually assumed. My purpose is to give empirical direction for theoretical integration between the two as a way to advance explanations that link micro, meso, and macro levels of analysis.
Establishing ethnography’s contribution to building mechanism-based explanations requires examining developments in ethnography that probably are unfamiliar to analytical sociologists and other nonpractitioners of the craft. Even ethnographers have not examined these developments in any systematic way. I begin by isolating ‘analytic ethnography’ (Lofland 1971) from the varieties of ethnography practiced, because it provides the primary basis for exchange between ethnography and analytical sociology. Second, and contradicting the conventional view of ethnography as limited to micro-level explanations, I demonstrate how analytic ethnography has developed mechanism-based explanations at all levels of analysis. Third, I compare analytic ethnography to analytical sociology (Hedström 2005) to establish the compatibility of the two, despite different understandings about the appropriate methods for studying the social. Finally, I consider the role analytic ethnography can play in advancing mechanism-based explanations that link actors, actions, and social outcomes. I demonstrate how theoretical integration can be achieved first within analytic ethnography, then between analytic ethnography and analytical sociology. To illustrate both, I consider how the signaling research of Vaughan (1983, 1986, 1996, 2002), Podolny (1993, 2005), Bacharach and Gambetta (2001), and Gambetta and Hamill (2005), authors in this volume using different methods, levels, and units of analysis, might be integrated to elaborate explanations of outcomes in which signals are the mechanism.
(p. 690) 29.1 Analytic Ethnography
What is analytic ethnography? No doubt most ethnographers would argue, and correctly, that all ethnography is analytic. This general point of agreement stands in contrast to a more recent history of contentious debate between ethnographers about the specifics of what ethnography is and how it should be done. The practice of ethnography has diversified into different fieldwork traditions, each with its own approach and epistemological assumptions (see e.g. Adler and Adler 1995; Atkinson et al. 2001; DeVault 2007). To complicate matters further, recent research investigating claims about the proliferation of ethnography found that the words ‘ethnography,’ ‘participant observation,’ and ‘fieldwork’—terms traditionally used by ethnographers to refer to closely related practices—have been used loosely, mainly by people who do not do ethnographic work (Culyba, Heimer, and Petty 2004).
Given this diversity, it is important to begin by locating analytic ethnography within ethnography. For purposes of this chapter, ethnography is defined as participant observation: research conducted by situating oneself in a social setting to observe and analyze individual interaction in order to understand some complex social process, event, activity, or outcome. Developing explanations requires extensive time in the setting—long enough to recognize patterns of interaction and distinctive social practices—so the duration of the observations cannot be known in advance. Explanation proceeds inductively by iteration between theory and data that revises theory. It is more than a method, however. It is a perspective committed to understanding a particular social world in terms of the meanings it has for the people who inhabit it. Because individual interpretation, meaning, and action are shaped by the social, the empirical focus is interaction within its social context. Social context is multilayered; to study it, the ethnographer carves it up according to the people, places, groups, organizations, or institutions relevant to explaining the problem. But context is not just background for some main event. The social context may either provide an explanation of some interaction, or interaction may explain some aspect of social context.
Currently two strands of ethnography with opposing epistemologies are dominant: analytic and critical. Analytic ethnography is our focus. However, critical ethnography has historic importance, not just for the experimental methods and cultural analysis it has produced, but because analytic ethnography has clarified its boundaries and methods as a result of it. By analytic ethnography, I mean an approach to field observations and interpretation of individual interaction that involves careful collection of data and evidence-backed arguments. It relies on systematic methods and standards, assumes that causes and explanations can be found, proceeds inductively to formulate explanations of outcomes, and holds theory and theoretical explanation as core objectives. The analysis developed is (p. 691) conceptually elaborated, based on interrogating the relationship between concepts, theory, and data, and aims for generic explanations of events, activities, and social processes.
In the 1960s and early 1970s ethnographers generally agreed that all ethnography was analytic ethnography, as defined above. Several ethnographers wrote about data-gathering, analysis, and theorizing, laying down a set of principles that became guidelines for other ethnographers. These writers challenged and criticized quantitative methods and grand theory, yet at the same time they adopted many of the tenets of positivism and the goal of developing theory. Their efforts aimed at training new ethnographers but also at gaining legitimacy for qualitative work criticized for its looseness and lack of systemicity in a discipline dominated by quantitative research and scientific methods. Two contributions received a lasting place in the canon: Blumer’s ‘sensitizing concepts’ (1954), and Glaser and Strauss’s grounded theory (1967). Of greatest impact was the Glaser and Strauss book, which was an exegesis on developing general theory by adhering to procedures of scientific method. Glaser and Strauss proposed a system for theory generation that was inductive, evidence-based, and developed by constant comparison between data and the nascent explanation. They identified two types of theory. Substantive theory is an explanation of an individual case that focuses on an empirical area, such as homelessness or patient care, with the goal of explaining the case. Formal theory is an explanation resulting from a conceptual or theory-based inquiry, developed by comparing cases and advancing an explanation that generalizes across cases. Encompassing these developments and affirming rigorous standards, the term ‘analytic ethnography’ first appeared in the literature in Lofland (1971).
The consensus that all ethnography was analytic ethnography dissolved in the late 1970s. Concurrent with the cultural turn in anthropology, history, and literature, a new genre of qualitative research took hold. Critical ethnography (also known as interpretive or postmodern ethnography) rejected the positivist tenets of objectivism and the search for causes, instead asserting subjectivism, language, and discourse as key to understanding the meanings upon which individual actions were based. The new work was opposed to analytic ethnography’s evidence-based assumptions and theoretical goals, its standard research strategies, and exclusionary research and writing practices (for a review, see Adler and Adler 1995). Critical ethnography eventually achieved an impact on analytic ethnography, visible in the appearance of first-person writing and practices reflecting a new awareness of bias and power relations between ethnographers and the people whose lives they entered. However, its commitment to evidence-based theory remained the same. Analytic ethnographers responded to the challenge by increasing their emphasis on rigor, systematic methods, and their search for generic, data-based theoretical explanations. For example, Katz (1983) reclaimed the quantitative lexicon of reliability, representativeness, reactivity, and replicability by showing how (p. 692) fieldwork embodied these very scientific standards. An all-encompassing textbook of systematic methods for qualitative data analysis appeared (Miles and Huberman 1984).
In 1985 Lofland clarified analytic ethnography in the midst of the alternative critical approaches by then available to ethnographers. He reiterated the goal of general theory: the identification of generic propositions that reflected the social-analytic categories that the data at hand exemplified (1985: 40). Significant historically was his reminder that ethnographers were in the business of science and the production of general explanations that went beyond the individual case. However, Lofland had not gone far enough (Snow, Morrill, and Anderson 2003). The purpose of analytic ethnography was general explanation, but Lofland had said nothing about how to get from generic propositions to theory. Even as analytic ethnography had flourished and systematic methods were deployed, Glaser and Strauss’s emphasis (1967) on going from particulars to a general explanation had apparently been lost. They had distinguished two kinds of explanation: substantive theory, which was achieving an explanation of the empirical case under investigation, and formal theory, or achieving a general explanation that not only encompassed the case at hand but developed by case comparison to generalize to other cases of similar phenomena. As practiced by most analytic ethnographers, however, the constant-comparative method that Glaser and Strauss advocated was reduced to the constant comparison between data and theory, thus dropping the comparison with other cases. Substantive theory, not formal theory, was the usual product (Strauss 1995).
These two trends continue to typify analytic ethnography today: the effort to teach and practice systematic methods has flourished, a most notable contribution being Emerson, Fretz, and Shaw’s Writing Ethnographic Field Notes (1995), and substantive theory is most often the product of an analysis. The result has been a proliferation of research guided by the standards of analytic ethnography, but diverse in many important ways, with little coherence within the specialty. Explanatory theory remains the goal, but analytic ethnographers have different understandings about what theory is, how it is to be achieved, the appropriate unit and level of analysis, and even the meaning of generalizability (Culyba, Heimer, and Petty 2004: 381). These differences have generated intense debates within the community of analytic ethnographers about what ethnography is and how to go about it (e.g. Anderson 2002; Duneier 2002; Newman 2002; Wacquant 2002). However, the heterogeneity within analytic ethnography has had a good result: an extensive body of research exists identifying mechanisms and providing explanations at micro, meso, and macro levels of analysis. The next section demonstrates how, in combination, both substantive theory and formal theory have produced mechanism-based explanations at all levels of analysis, thus establishing the basis for our later discussion of theoretical integration within analytic ethnography and between analytic ethnography and analytical sociology.
(p. 693) 29.2 Varieties of Analytic Ethnography: Linking Action, Mechanisms, and Outcomes in Explanation
Since the 1940s and 1950s, 1950 analytic ethnographers have generated hundreds of mechanism-based explanations at all levels of analysis. Consider the classics. Contradicting the view that ethnography is limited to micro-level explanations, many classic ethnographies produced substantive theories explaining meso outcomes. For example, Whyte’s Street Corner Society (1943) identified reciprocal obligations between individuals as the mechanism explaining the structure of the corner gang, the racket and police organizations, and political organizations in the Italian North End of Boston.
‘Network’ was not yet the accepted terminology when Dalton (1959) identified the mechanisms that produced ‘cliques’ in bureaucratic organizations, then showed how competition, conflict, and cooperation between the cliques were mechanisms producing organization survival. Similarly, Becker (1982) identified cooperative networks of different specializations as the mechanisms producing art, but ‘conventions’—basic knowledge shared by all who participate in the art world—played a role in linking actors and actions to produce the networks. In actor-network theory, the mechanism generating network formation and collective activity was ‘translation’: an entrepreneurial actor defines the problem, interests others, enrolls participants, and mobilizes action (Callon 1986; Latour 1987, 1988). But entities also have agency. Thus, the ‘boundary object’—an idea, a bacteria, a mosquito—could also be a mechanism triggering network formation (Star and Griesemer 1989). Undoubtedly, the best-known substantive theory of a macro outcome is Willis’s classic (1977) explaining the reproduction of social class. He identified the mechanisms by which ‘working class lads get working class jobs.’ By their opposition to school authority and their refusal to master the curricula, they disqualified themselves from educational credentials and reproduced themselves as a working class.
In contemporary ethnographies producing substantive theory, mechanism-based explanations that make micro–macro connections are so numerous they defy examination in a chapter of this length.1 Instead, I turn now to contemporary exemplars of formal theory in order to make a set of more detailed analytical points. Focusing on formal theory has two advantages: first, formal theory is the strongest example because it is based upon case comparisons that show how particular mechanisms have regularly produced similar outcomes over time. Second, it is expedient as a selection criterion because efforts to transform substantive theory into formal theory are few. To draw attention to the production of mechanism-based (p. 694) explanations at a variety of levels of analysis, I present the exemplars in this order: micro, then macro, and finally a micro–meso–macro–explanation. I will first indicate, very schematically, the organizing principles of each exemplar: theoretical orientation, level and unit of analysis, and method. Then I show the mechanism-based explanation as it was developed and elaborated to generalize across cases, successively developing formal theory—one of considerable scope, range, and conceptual complexity, with a degree of specificity that encompasses variability (Strauss 1995: 17).
29.2.1 Katz: a theory of situational transcendence
Theoretical orientation; unit and level of analysis; method: A phenomenologist and symbolic interactionist, Katz’s theoretical orientation directs his attention to actors, mechanisms, and outcomes, all at the micro level. The causes of the outcomes to be explained can be found in aspects of the immediate situation. Explanation, for Katz, resides in uncovering the meanings grounded in the subject’s experience of the moment. Emotion is the mechanism linking actors and actions to outcomes. Emotions are an actor’s way of sensing the meanings currently present in the immediate situation. Conditions antecedent to the situation, such as social class, gender, and race, are irrelevant ‘background’ factors that ‘fail to address the lived experience’ of social life. Thus, his unit of analysis is individual interaction, isolated from individual social location, previous experience, history, culture, or structure. His method is to develop formal theory by cross-case comparison: analyzing similar phenomena in different social settings.
Mechanism-based explanation: In Seductions of Crime: Moral and Sensual Attractions of Doing Evil (1988) Katz develops a formal theory to explain all varieties of crime. He compares six cases: each chapter builds an explanation of one type of crime from many examples by combining ethnography, interviews, and published historical, biographical, and research accounts. In each, the circumstances of the situation challenge the actor’s sense of identity and place in the world. The resulting emotion drives the criminal act such that at the moment of its commission the crime becomes morally justified—a sensible, satisfying, compelling action that allows the offender to transcend the situation.
In violent crimes, emotions—humiliation, degradation, frustration—produce a ‘socially transcendent rage’ that compels the person to act. Committed in defense of social identity, the crime at the moment of commission restores the offender’s sense of self in relation to others. Reinforcing Katz’s claim for a theory explaining all kinds of crime, he varies social class, showing how middle-class and upper-class people are seduced by the situation in property crime and white-collar crime. For example, college students shoplift not for material reasons but because of the power of the object, its availability, the ease of taking it, and the emotional thrill of transcending (p. 695) the mundane by getting away with a forbidden act. Thus, based on these cross-case comparisons, Katz stakes his claim that his theory generalizes to the crimes of all social classes.
29.2.2 Burawoy: the extended-case method
Theoretical orientation; unit and level of analysis; method: A Marxist, Burawoy’s theoretical orientation directs his attention to macro-level outcomes: specifically, reproduction or change of economic systems. Taking factory jobs, he studied the micro processes of four large-scale historical transformations: decolonization, the transition to organized capitalism, the Soviet transition to socialism, and the transition from socialism to market capitalism. The mechanism responsible for linking actors, action, and outcomes was the social organization and regulation of work. His unit of analysis was the shop floor, analyzed within history and the economic forces of production. He used the extended-case method (Burawoy 1998), so called because a case (1) extends into history and into the future, as each day in the field shows the evolving situation, (2) extends from micro processes to macro-level forces, and (3) extends theory by comparison with other cases.
Mechanism-based explanation: In Manufacturing Consent (1979) Burawoy found that the possibility of working-class revolutions was not constrained by the superstructure, as Marxists posited, but in the workplace. The factory coordinated the interests of workers and management in the pursuit of profit so that consent, not dissent, was the product. The organization and regulation of work was the mechanism that produced the consent of workers to work. The trade union gave legitimacy to contracts and protected the rights of individuals, eliciting consent to the factory order. The piece-rate system guaranteed a minimum wage and job security. Together, they gave workers incentive and opportunity to establish counternorms to protect themselves from some of the deprivations of the job. Consequently, at the micro level work was constituted as a game in which workers cooperated to ‘make out,’ which, ironically, resulted in the factory meeting its production goals. In this way, advanced capitalism reproduced itself.
Using comparisons to proceed from this substantive theory toward formal theory, Burawoy investigated whether work was organized and regulated differently in Eastern Europe, so that dissent rather than consent was the product. Ethnographies of factories in different political economies explored variation within and between capitalism and state socialism (Burawoy 2009). The comparisons affirmed that the organization and regulation of work was the mechanism that produced either dissent or consent of workers to work, and thus the development (or not) of counterpower movements to effect social change.
(p. 696) 29.2.3 Bourdieu: a theory of practice
Theoretical orientation; unit and level of analysis; method: Bourdieu was eclectic, combining Marx, Weber, and structural anthropology with attention to the body and practices in everyday life. His focus was the macro level: power, domination, and reproduction or change of social hierarchy. The mechanism linking actors and actions to outcomes was habitus: a subjective set of dispositions comprised of lasting patterns of thought, perception, and behavior (Bourdieu  1990). Habitus is acquired by individuals or groups through internalization of culture or social structures such as class, family, occupation, education. They become embodied, providing a set of practical skills, enacted without thinking. His method for developing formal theory was cross-case analysis: analyzing similar phenomena in different social settings. Sallaz and Zavisca (2007) trace its evolution chronologically: social structure and social action from the Kabyle; the reproduction of inequality in education in France; social inequality in cultural production and consumption; the theory of the state through language, education, and housing. His unit of analysis varied, depending on the case he was examining. An experienced anthropologist, even when relying primarily on survey research his ethnographic insights dominate the analysis (Wacquant 2004b).
Mechanism-based explanation: In Distinction ( 1984), Bourdieu shows how an individual’s cultural practices can either maintain or increase position in the system of class domination. Class is enacted in tastes for food, ritual practices of serving, the formality of the meal, preferences for glassware and crockery, and the use of the body in consumption. Taste itself becomes embodied, its enactment ‘revealing the deepest dispositions of the habitus’ ( 1984: 190). For Bourdieu, the dynamic relationship between habitus, field, and capital is essential to explaining persistence or change in structures of inequality. The field is a highly competitive site structured by power relationships, where people jockey for resources, status, and control over outcomes. Forms of capital—economic, social, cultural, symbolic—are a determinant of success for individuals. In response to the conditions they encounter, actors internalize sets of dispositions and actions (habitus) consonant with the norms of the field, thereby tending to reproduce its power structure. In Distinction, class-cultural practices constitute a system of status signals, differentiating the working class from the bourgeoisie, thereby limiting mobility, maintaining class position, and reproducing inequality.
In European scholarship applying his theory to other cases, it has been verified frequently. In American sociology, however, field, habitus, and capital have seldom been joined in a single analysis (Sallaz and Zavisca 2007; Emirbayer and Johnson 2008). Instead, each concept has had a separate path of elaboration (see e.g. Martin 2003). Habitus may be the concept least investigated empirically, and rarely as the linking mechanism between micro and macro levels of analysis that Bourdieu (p. 697) intended. (For American exceptions see Vaughan 1996; Hallett 2003; Wacquant 2004a; Desmond 2007; Kahn 2008.)
29.2.4 Vaughan: analogical theorizing
Theoretical orientation; unit and level of analysis; method: Analogical theorizing generates micro-, meso-, and macro-level explanations. It is influenced by sociological theory indicating that action is situated in layered structures, culture, and history, which affect interpretation, meaning, and action at the local level. Meso-level structures—neighborhoods, complex organizations, groups—transform historical institutional imperatives, shaping cultural understandings to fit the local situation. Thus, the unit of analysis is individual actors and interaction, situated within meso- and macro-level contexts. Analogical theorizing is based on Simmel’s formal sociology, which urges sociologists to extract form from content to find common processes and structures in similar activities across different social settings. It calls for cross-case (hospitals, schools, families) rather than same-case (all hospitals) comparison. Lovers break up as con men cool out marks as bosses fire employees as children terminate monopoly games (Vaughan 1986: 188–195).2
A case is chosen as a possible example of some phenomenon. The starting theory or concept is used heuristically, as a tool for sorting the data to identify both analogies and differences. Procedures are in place to guard against force-fitting the data to the theory (Vaughan 1992, 2004). The result may confirm, refute, or elaborate the starting theory, which, in its modified form, guides analysis of the next example. Additional cases are chosen that vary in size, complexity, and function, intentionally producing data at different levels of analysis in order to develop a micro–meso–macro explanation. Findings of either reproduction or change are possible. Although the emphasis is on identifying micro–meso–macro connections, analogical theorizing is theoretically and methodologically neutral, and thus could be used with qualitative and quantitative research at any level of analysis, as I will show at pp. 702–6.
Mechanism-based explanation: Vaughan’s ‘The Social Organization of Dissent’ investigated ‘individual organization members who resist authority, speaking out against illegal, illegitimate, or immoral actions in or by the organization to which they belong, alerting other actors who might help effect some change’ (unpublished manuscript, ch. 5). Comparing the prison snitch, corporate whistle-blowers, and dissent against sex-discrimination in hiring, sexual harassment in the workplace, and domestic violence showed the mechanisms that produced or repressed dissent across cases. In the prison, the workplace, and the family, the same bundle of five mechanisms appeared. Social location, power-dependence relations, norms of loyalty, organizational retaliation, and social support interacted: willingness to dissent (p. 698) varied as these conditions varied. For example, social support, when present, was an ‘enabler’ of dissent, when absent, a ‘silencer.’ Macro-level contingencies, such as structured gender inequalities, and legal and normative standards also affected the production of dissent.
To sum: Analytic ethnography has a long history of building mechanism-based explanations linking actors, actions, and outcomes at all levels of analysis. Close examination of formal theory exemplars showed that analytic ethnography comes wrapped in very different packages (see Table 29.1). The heterogeneity in analytic ethnography is not an indicator of an undisciplined specialty, however. Rather, each exemplar differs in the theoretical orientation of the author, demonstrating that mechanism-based explanations have been achieved using very different theoretical approaches. Each has been driven by its author’s theory of how society works, which in turn affects the unit of analysis, the level of analysis, and the level of explanation that the ethnographer produces.
The same is true for substantive theory. Since each ethnographer’s process is ordered by some overarching theory or theories about how the world works, and since theoretical orientation determines the level of explanation, analytic ethnographers have produced mechanism-based explanations at every level of the multi-layered social context. Not all analytic ethnographers are committed to explaining macro-level outcomes. However, the examples of substantive theories and formal theories, in combination, show how the varieties of analytic ethnography bridge these layers, filling in the intersection of social processes and structures at micro, meso, and macro levels of analysis. This ability of analytic ethnography to identify mechanisms that link actions to outcomes at all levels of analysis is central to understanding how it meshes with the goal of analytical sociology to develop explanations that bridge the distance between theories of macro–micro relations and research that does the same. To establish the basis for such integration, next I examine the compatibility of these two approaches, usually viewed as oppositional and incompatible.
29.3 The Compatibility of Analytic Ethnography and Analytical Sociology
Recall that analytical sociology is designed to escape the limits of variable analysis in quantitative sociology by examining key social processes and breaking them down into their primary components, showing how they work together to produce social outcomes. Providing plausible causal accounts of why things happen is a goal,
Table 29.1 Formal theory mechanism-based explanations
Unit of Analysis
Level of Explanation
Social Organization of Work
Reproduction or Change: Economic System
Marx, Weber, Anthropology, Body, Practice
Position of Individual in Field or Social Space
Reproduction or Change: Hierarchy, Domination
Position of Individual in Meso-Structure; Macro-Context
Reproduction or Change
(p. 700) despite internal disputes about the meaning of explanation. Examining analytic ethnography in relation to analytical sociology’s more specific principles establishes the compatibility of the two approaches by showing how, using very different methods, analytic ethnography uses the same principles to arrive at explanations of the social. Those principles are explanation, dissection and abstraction, precision and clarity, and a theory of action (Hedström 2005).
29.3.1 Principles: similarities and differences
Explanation: The position of analytical sociology is that the best explanation for the social sciences is mechanism-based. Mechanisms are the ‘constellation of entities and activities, typically actors and their actions, that are linked to one another in such a way that they regularly bring about the type of phenomenon we seek to explain’ (Hedström 2005: 2). As we have seen, mechanisms also are the core concern of analytic ethnography. Empirical work breaks down social processes into key components to see how they effect social outcomes. Explanation requires detailed description: the two go hand in hand (Katz 2001b). Detailed description renders mechanisms observable. They may be micro-, meso-, or macro-level, or some combination.
Dissection and abstraction: For analytical sociology, dissection and abstraction are tools for decomposing complex social processes in order to identify the mechanisms that drive social outcomes. Dissection isolates the constituent entities; abstraction selects from them those most essential to understanding the phenomenon of interest. Analytic ethnography relies upon the same sorting and sifting practices to identify mechanisms that produce social regularities. Observations are recorded in detailed field notes, often supplemented by interviews, documents, and/or archival materials. Close, careful attention to data and its accurate depiction of social life is the essence of explanation. It is the complexity and amount of descriptive data, often difficult to sort into patterns, that makes dissection and abstraction essential tools for explanation. After gathering descriptions, the focus shifts to the analytic: reorganization of data into explanatory lines through iterative processing of theory and data. Description is made useful by conversion into theoretical specification. This process Katz describes as going from ‘how’ to ‘why’ (2001b). ‘Abstraction,’ however, applies in a second way for those analytical ethnographers who strive for formal theory. After isolating the essential patterns and deriving substantive theory, they abstract further from the substantive explanation to build formal theory generalizing beyond the individual case.
Precision and clarity: For analytical sociology, achieving precision and clarity includes identifying details and conveying them through operational and conceptual definitions that can be easily grasped (Hedström 2005). Otherwise mechanisms remain obscure and explanation is not achieved; the relationship between one (p. 701) activity, event, or phenomenon and another will not be conveyed. For analytic ethnography, precision and clarity also depend on moving from the detailed data to clear operational and conceptual definitions. Precision and clarity are obtained by (1) accurate description so that both patterns and exceptions are clear; (2) analytic induction, a research logic for data analysis that requires continually adjusting a developing explanation to take into account anomalies that contradict hypotheses until a perfect relationship between data and theory is reached (Lindesmith 1947; Katz 2001a); (3) constant comparison between theory and data (Glaser and Strauss 1967) in the individual case and in some instances cross-case comparison also, depending on the researcher’s interest in formal theory or substantive theory; and (4) presenting results that provide strong empirical evidence to support analysis and conclusions.
Operational definitions and conceptual definitions are integral to every analysis, although their use and centrality in the written analysis may vary. First, concepts may be used to set up the research problem by defining the social setting conceptually and stating the theoretical and empirical boundaries for investigation. Second, a known concept may guide the data analysis, its development, and theory elaboration. Third, a new concept may be created as an outcome of the data analysis. Finally, several concepts may be joined in a formal theory that initially frames the data analysis, then is revised as a result of the findings.
A theory of action: The stated goal of analytical sociology is not just achieving explanation, but achieving explanation that links actors, actions, and outcomes. This requires a theory of action that explains what actors do. Analytic ethnography also shares this goal. However, the two approaches disagree on methodological strategies and theoretical assumptions for achieving a theory of action. Ironically, this difference turns out to be the crux of a productive relation between them.
Most often in analytical sociology the empirical focus is on the structures of interaction that are the outcome of individual interaction, not the actors, actions, and interactions themselves (but see Gambetta, this volume). The theory of action precedes the analysis, guiding it by positing the mechanisms likely to have brought about the outcomes to be explained. Consequently, making the connection between actions and outcomes relies on a rational-choice assumption. Because the focus is structures of interaction, analytical sociology differentiates itself from other rational-choice-based approaches; it is ‘structural individualism,’ not methodological individualism (Hedström 2005: 5 n. 4). About actors and actions, the question asked is ‘why, acting as they do, they bring about the social outcomes that they do’ (2005: 5). Social outcomes are defined as changes in collective properties brought about by the structures of interaction. Thus, the outcomes to be explained are structural, either at the meso or macro level of analysis. Often, social change is the outcome to be explained.
In contrast to analytical sociology, actors, action, and interaction are directly observed in analytic ethnography. The assumption is that individual choice is (p. 702) rational, but the social context shapes what is perceived as rational at a given moment. Although this constitutes the major divergence between analytic ethnography and analytical sociology, it also is the most important basis for exchange between them. Because the social context is multilayered, analytic ethnographers have directly linked actors and actions to micro, meso, and macro outcomes, as we have seen. Research has verified both reproduction and change. The contributions analytic ethnography has made to mechanism-based explanations across levels of analysis are the foundation for theoretical integration: building mechanism-based explanations that connect actors and actions to social outcomes within analytic ethnography and between analytic ethnography and analytical sociology. The next section will illustrate this potential.
29.4 Theoretical Integration: Signals as Mechanisms
Analogical theorizing has a role to play in advancing mechanism-based explanations both within analytic ethnography and between it and analytical sociology. Eclectic in orientation, it is a tool for theory-building by case comparison regardless of qualitative or quantitative method. The premise behind cross-case rather than same-case comparison is that theory and concepts developed at one level of analysis can be used heuristically to examine social relations at another. This is so because all socially organized forms have structures and processes in common—division of labor, stratification, conflict, hierarchy, cooperation, culture, socialization (Vaughan 1992). Therefore, theoretical integration could be achieved two ways.
One would be separate projects within analytical sociology and analytic ethnography to advance mechanism-based explanations at the same level of analysis. In analytical sociology, for example, research could investigate how the same mechanisms (e.g. those associated with networks) operate across different substantive problems, or how different mechanisms are involved in producing the same (or similar) social outcomes. In analytic ethnography, cross-case comparison at the same level of analysis could begin to move substantive theory toward formal theory. For example, research using actor-network theory has never been integrated to identify similarities and differences found. Theoretical integration at the same level of analysis is an important precursor to the second means of achieving theoretical integration: comparing research on similar phenomena, in different social settings, at different levels of analysis. The result would be the development of mechanism-based explanations linking actors to outcomes that (p. 703) span micro, meso, and macro levels of analysis, both within analytic ethnography and between it and analytical sociology. I will demonstrate this second approach by showing how signals as a mechanism can be found in multiple contexts at multiple levels of observation. First, the signaling examples will show how theoretical integration across levels of analysis can be achieved within analytic ethnography (Vaughan 1983, 1986, 1996). Then, to show theoretical integration between analytic ethnography and analytical sociology, I extend my findings on signals to research by analytical sociologists (Podolny 1993, 2005; Bacharach and Gambetta 2001; Gambetta and Hamill 2005).
In substantively unrelated research—corporate malfeasance (Vaughan 1983), the deterioration of intimate relationships (Vaughan 1986), and NASA’s flawed decision to launch the space shuttle Challenger (Vaughan 1996)—signals were a mechanism contributing to harmful outcomes. Comparison was not part of the original design; I did not suspect that the cases had analogous patterns nor that signals would be a part of the explanation. The misconduct case was not an ethnography but important because signaling theory (Spence 1974) became part of its sociological explanation, which then led to its ethnographic elaboration in the two subsequent cases. Because the three cases were organizations varying in size, complexity, and function, each case produced data at different levels of analysis. The result was theoretical integration across levels of analysis: a mechanism-based theory of mistake and unanticipated consequences that connected micro, meso, and macro levels of analysis.
In the first case (Vaughan 1983), a prominent drugstore chain made over 50,000 false Medicaid claims, submitted them on computer tapes, and received reimbursement from the state welfare department. The scheme went on for three years. Why were the false claims not detected? The explanation lay in the characteristics of organizations, how they exchange information, and how information (signals) was interpreted. Spence (1974) explained how organizations discriminated among multiple job applicants in hiring decisions. To reduce costs, firms used a shortcut method that relied on signals as indicators of quality and competence. Various signals (education, recommendation letters) thus became the basis of decision-making. Spence’s theory fit my data: many claims were submitted; the welfare department’s computerized screening system checked for predesignated signals to separate good claims from bad. Because signals varied, they could be manipulated, making fraud possible. Discovery was prevented by ‘structural secrecy’—the nature of organizational exchange, structure, and division of labor interfered with the sending, receiving, and interpretation of information—so the fraud continued. It was discovered accidentally during investigation of another matter.
Intimate relationships are the smallest organizations we create. Their endings are not sudden breaks but gradual transitions during which identity and social networks are transformed (Vaughan 1986). The main pattern is that one person, (p. 704) whom I called the initiator, becomes unhappy and begins leaving the relationship socially and psychologically long before the other person. How was it possible for one person to get so far away without the other noticing and acting? Surprisingly, the uncoupling process was analogous to the fraud case: both had long incubation periods that were littered with warning signs that were either misinterpreted or ignored. This case provided micro-level data that added to the structural explanation in the first case by revealing the process of sending, receiving, and interpreting signals. The initiator’s signals were embedded in patterns of information that diffused their import: they appeared mixed, weak, and routine. In addition, the partner had a positive view of the relationship, and thus focused on the positive signals, not the negative. Structural secrecy again played a role. The relationship’s patterns of exchange, structure, and division of labor affected the partner’s interpretative work. Typically, initiators only sent an undeniable strong signal when they were ready to go. By that time, the relationship was difficult—if not impossible—to salvage.
The Challenger research (Vaughan 1996) inquired into the organizational contribution to the technical failure. Why, for years before the accident, had NASA repeatedly launched despite recurring anomalies on the O-rings, the shuttle part that failed? The analysis revealed a now familiar pattern: the accident was preceded by a long incubation period filled with warning signals that were either misinterpreted or ignored. This case produced data that gave a micro-, meso-, and macro-level explanation showing signals as a mechanism producing the outcome. At the micro level, as launch data accumulated O-ring anomalies occurred in a pattern of information indicating they were operating as expected: signals appeared mixed, weak, and routine. At the meso level, structural secrecy created by organizational exchange, hierarchy, division of labor, and evidentiary rules obscured the seriousness of these early warning signs. At the macro level, decisions to continue launching were reinforced by production pressures generated by the White House, Congress, and international relations, resulting in a frame of reference focusing on the good signals, not the anomalies. The result was ‘the normalization of deviance’: technical anomalies that outsiders viewed after the accident as clear signals of danger were viewed as normal and acceptable to insiders as launch decisions were being made.
The cross-comparison demonstrates theoretical integration within analytic ethnography: it led to a Bourdieusian theory of signals and interpretive work that included micro-, meso-, and macro-level factors (see Vaughan 2002). It verified signals as a mechanism in producing mistakes in organizations that varied in size, complexity, and function. It exposed the dynamics of signals as a mechanism and the process of sending, receiving, and interpreting them. It showed that signals can be interpreted differently, their meaning affected by their embeddedness in a stream of information, organization characteristics, history and previous understandings, actions by actors in the institutional environment, and the social location of the (p. 705) interpreter. Because the meaning of a signal varies, depending on the social context and the worldview of the interpreter, falsification, mistake, and unintended consequences can occur. Further, the comparison shows that the ethnographic project is uniquely positioned to identify mechanisms linking actors to macro-level outcomes because the relation between micro, meso, and macro levels of analysis can be—and is—observed, traced, and documented in what people say and do in local situations. For example, a NASA engineer’s statement, ‘No one has to tell me schedule is important when I see people working 24 hours a day, 7 days a week,’ captures the impact of NASA’s competitive institutional environment on organizational practices and thus on the individual engineers.
To consider theoretical integration with analytical sociology, we extend the above findings about signals to the research of Podolny and Gambetta. The signaling comparison is especially apt because it shows how a theory developed in a mathematically oriented rational-choice specialty (economics), can be elaborated into sociological explanations of signals as a mechanism operating in different social settings, across levels of analysis. Podolny (1993, 2005) conceptualizes status as a signal of quality and investigates market mechanisms for sustaining the status-ordering in a market. He hypothesizes that rational consumers will rely on status as an indicator of quality in order to reduce their uncertainty in markets. Making rational-choice assumptions and working at the macro level, he examines status dynamics in industries, comparing investment banking, wine, shipping, venture capital, and semiconductors (Podolny 2005). One finding is that signals are inseparable from context: reputational information is situated in the pattern of ties among a group of actors.
Bacharach and Gambetta (2001) reconceptualize signaling theory to analyze trust games. Using ethnography and interviews to understand how taxi drivers establish a customer’s trustworthiness, Gambetta and Hamill (2005) examine customers’ mimicry and falsification of signals and taxi drivers’ rational and nonrational reasoning (unfounded beliefs, unfounded stereotypes) in identifying and interpreting signals to resolve uncertainty about trustworthiness. They alter the rational-choice assumption that incentives in the form of rewards and penalties motivate the customer; instead Gambetta defines trustworthiness as the product of customer attributes conveyed to and interpreted by the taxi driver as signals. Consistent with signaling theory in economics, costs figure into the trust game: reliability of a signal is based on calculating the cost of attaining it legitimately.
The above summaries of Podolny, Gambetta, and Vaughan suggest how integration of theory and research on signals between analytic ethnography and analytical sociology might proceed. Analogical comparison is justified because, uniformly and consistent with use of the concept in economics, Podolny, Gambetta, and Vaughan all treat signals as information, examining how decisions are made under conditions of uncertainty, and all found signals operating as a mechanism. In contrast to its use in economics, in all three signals are inseparable from context. However, (p. 706) it is the differences between the three signaling projects that promote theory elaboration and integration between the two disciplinary specialties. The research by Podolny, Gambetta, and Vaughan could be subject to systematic in-depth comparison, searching for analogies and differences, that begins to articulate a general theory of signaling that merges levels of analysis. Space prevents such a comparison here, but we can identify some directions that would advance mechanism-based explanations.
Whereas typically analytical sociology examines structures of interaction rather than interaction itself (Hedström 2005), analytic ethnography supplies data about interaction within social context. Variation in signals sent over time, their embeddedness in information streams, institutional and organizational effects, and the social location of the interpreter show the dynamics and uncertainty that underlie the production of trust and the status-ordering of a market. What is the relevance of mixed, weak, and routine signals or the sequential interjection of new signals in a social context over time for the research of Podolny and Gambetta? Both Gambetta and Vaughan raise the possibility of mistake and unintended consequences. How might this affect market stability? What relevance might the normalization of deviance have for both trust relations and market stability, in light of uncertainty and variation in signals and their interpretation? All three projects are either explicitly or implicitly about risk, so another possibility is to incorporate signals as mechanisms into a theory of risk that stands up across the three substantive problems summarized here. Above are only a few of the possible ways that analytic ethnography and analytical sociology could be joined. My purpose has been to be suggestive, rather than definitive. Nonetheless, the examples indicate the potential for theoretical integration both within analytic ethnography and between it and analytical sociology, thus advancing the development of mechanism-based explanations that span micro, meso, and macro levels of analysis.
The purpose of this chapter has been to demonstrate the unique contribution that analytic ethnography has made and can make to accomplishing the two key principles of analytical sociology: (1) building mechanism-based explanations that connect actors, actions, and outcomes, and (2) bridging the micro–meso–macro levels of analysis in those explanations. I began by defining ethnography and locating analytic ethnography within it as an evidence-based, theory-building enterprise. Then, using examples from substantive theory and formal theory, I showed the contribution analytic ethnographers have made to identifying mechanisms (p. 707) that connect actors, actions, and outcomes at all three levels of analysis. Next, to establish the empirical basis for theoretical integration between the two specialties, I showed their compatibility, describing how they achieve the same principles. The single divergence—a theory of action—is crucial to exchange between the two. Finally, I addressed the critical issue of theoretical integration. Demonstrating analogical theorizing as a useful tool, I showed how analytic ethnography produced an analysis of signals as a mechanism that incorporated micro, meso, and macro factors to explain outcomes. Finally, I extended those results to signaling research in analytical sociology to show how theory building might occur.
My point has been that analytic ethnography has more in common with analytical sociology than is usually assumed, has consistently and reliably identified mechanisms that link actors to outcomes at all levels of analysis, and that theoretical integration both within analytic ethnography and between it and analytical sociology is a productive way to advance mechanism-based explanations that link micro, meso, and macro levels of analysis. Obstacles exist, however. Currently, there is lack of integration both within analytical ethnography as well as between it and analytical sociology. Little effort has been put to accumulating results within each specialty in a systematic way. For many scholars of both specialties this integrative direction requires altering a commitment to substantive theory and thus how research is conducted. Typically, the emphasis is on moving forward with the next substantive project, not integrating the findings from the last into the next, so formal theory is not the outcome. Moreover, integrating micro, meso, and macro levels of the social in explanations has not been a common goal. This condition has structural origins. Although theory classes instruct that the essence of society is the nexus of structure and agency, our tendency is to specialize in a method that best captures either the macro or micro level of analysis but not both, due to department specialty, mentor influence, personal capability and preference, and the desire to get a job in a discipline where specialization is rewarded. Although research methods classes encourage mixed methods, how to combine them to make empirical connections across levels of analysis is not typically taught.
Many scholars bridge the micro–macro gap by framing micro-level projects in literature describing relevant macro-level factors or the reverse, but empirical demonstrations of the connection between the two are more rare and tend to be qualitative. An obstacle specific to ethnography is that ethnographers work, often alone, on distinctive problems in unique settings that make finding generalities across projects difficult. If achieving integration within each approach is complex and challenging, how then might we pursue a productive relationship between analytic ethnography and analytical sociology in making micro–macro connections? Although the structural obstacles described above suggest that ultimately bridging the gap between theories of macro–micro relations and research depends upon structural change in the discipline, the example of signals as mechanisms suggests that crossing intradisciplinary boundaries to bridge the micro–macro gap (p. 708) may be easier than imagined. Most promising in the signaling example is that three sociologists independently read across disciplinary boundaries to discover and use an economic theory that helped explain their data, which led to examination of signals as a mechanism in multiple contexts at multiple levels of analysis in sociology. Even if general theory remains elusive, the exchange of ideas, theories, and concepts between analytic ethnography and analytical sociology would be productive. Some of the central interests of analytical sociology are networks, informal rules and social norms, and the desires, benefits, and structured opportunities of individuals (Hedström 2005). These all can be and have been studied by analytic ethnographers. Network ties, for example, can be examined using individuals, families, organizations, or nation-states (or subunits within them) as the units of analysis. Dalton’s intra- and interorganizational ‘cliques’ (1959), Latour’s actor-network theory (1987), Becker’s conventions (1982), and the concept of boundary object (Star and Griesemer 1989) hold hypotheses for network research in analytical sociology.
It is this bridging ability of analytic ethnography that is the essence of theoretical integration, both within the specialty and between it and analytical sociology. Analytic ethnography cannot explain large-scale social outcomes, as does analytical sociology, but it has developed mechanism-based explanations of both reproduction and change at the macro level. Also, it can make micro–meso–macro-level connections and has accumulated data at all levels of analysis (see note 1). Analytical sociology examines the structures of interaction. The strength of analytic ethnography is revealing the complexity of actors, actions, and interactions—the mechanisms behind the mechanisms that contribute to the macro-level change and stability that are the focus of analytical sociology. Theorists have long argued that the social is explained by the intersection of micro- and macro-level factors. Mechanisms are the linchpin of that relationship. The examples of mechanism-based explanations presented here allow a more empirically grounded consideration of what micro–macro integration might entail and how we might proceed. Further, the similarities and differences between analytical sociology and analytic ethnography described in this chapter suggest that a productive exchange relationship would make more complete explanations possible.
Adler, P., and Adler, P. (1995), ‘The Demography of Ethnography’, Journal of Contemporary Ethnography, 24: 3–29.Find this resource:
Anderson, E. (2002), ‘The Ideologically Driven Critique’, American Journal of Sociology, 107 (6): 1533–50.Find this resource:
Atkinson, P., et al. (2001) (eds.), Handbook of Ethnography (Thousand Oaks, Calif.: Sage).Find this resource:
Auyero, J. (2005), Routine Politics and Violence in Argentina (New York: Cambridge University Press).Find this resource:
Bacharach, M., and Gambetta, D. (2001), ‘Trust in Signs’, in K. S. Cook (ed.), Trust in Society (New York: Russell Sage).Find this resource:
Becker, H. S. (1982), Art Worlds (Berkeley, Calif.: University of California Press).Find this resource:
Blumer, H. (1954), ‘What is Wrong with Social Theory?’, American Sociological Review, 19: 3–10.Find this resource:
Bourdieu, P.  (1984), Distinction (Cambridge, Mass.: Harvard University Press).Find this resource:
——— (1990), The Logic of Practice (Stanford, Calif.: Stanford University Press).Find this resource:
Bourgois, P. (1995), In Search of Respect (New York: Cambridge University Press).Find this resource:
Burawoy, M. (1979), Manufacturing Consent (Chicago, Ill.: University of Chicago Press).Find this resource:
(1998), ‘The Extended Case Method’, Sociological Theory, 16: 4–33.Find this resource:
(2009), The Extended Case Method: Four Countries, Four Decades, Four Great Transformations, and One Theoretical Tradition (Berkeley, Calif.: University of California Press).Find this resource:
Callon, M. (1986), ‘Some Elements of a Sociology of Translation’, in J. Law (ed.), Power, Action, and Belief (Keele: Methuen), 196–229.Find this resource:
Culyba, R. J., Heimer, C. A., and Petty, J. C. (2004), ‘The Ethnographic Turn’, Qualitative Sociology, 27 (4): 365–90.Find this resource:
Dalton, M. (1959), Men Who Manage (New York: J. Wiley).Find this resource:
Desmond, M. (2007), On the Fireline (Chicago, Ill.: University of Chicago Press).Find this resource:
Devault, M. L. (1991), Feeding the Family (Chicago, Ill.: University of Chicago Press).Find this resource:
(2007), ‘Knowledge From the Field’, in C. Calhoun (ed.), Sociology in America: A History (Chicago, Ill.: University of Chicago Press).Find this resource:
Diamond, T. (1992), Making Grey Gold (Chicago, Ill.: University of Chicago Press).Find this resource:
Duneier, M. (2002), ‘What Kind of Combat Sport is Sociology?’, American Journal of Sociology, 107 (6): 1551–76.Find this resource:
Eisenhardt, K. M. (1989), ‘Building Theories from Case Study Research’, Academy of Management Review, 14 (4): 532–50.Find this resource:
Emerson, R. M., Fretz, R. I., and Shaw, L. L. (1995), Writing Ethnographic Fieldnotes (Chicago, Ill.: University of Chicago Press).Find this resource:
Emirbayer, M., and Johnson, V. (2008), ‘Bourdieu and Organizational Analysis’, Theory and Society, 37 (1): 1–44.Find this resource:
Espiritu, Y. (2005), Home Bound (Berkeley Calif.: University of California Press).Find this resource:
(p. 710) Gambetta, D., and Hamill, H. (2005), Streetwise (New York: Russell Sage).Find this resource:
Glaser, B. G., and Strauss, A. L. (1967), The Discovery of Grounded Theory (Chicago, Ill.: Aldine).Find this resource:
Goffman, E. (1961), Asylums (Garden City, N.Y.: Anchor).Find this resource:
Grasmuck, S. (2005), Protecting Home (Piscataway, N.J.: Rutgers University Press).Find this resource:
Halle, D. (1984), America’s Working Man (Chicago, Ill.: University of Chicago Press).Find this resource:
Hallett, T. (2003), ‘Symbolic Power and Organization Culture’, Sociological Theory, 21 (2): 128–49.Find this resource:
Hedström. P. (2005), Dissecting the Social (Cambridge: Cambridge University Press).Find this resource:
Kahn, S. (2008), The Production of Privilege, Ph.D. thesis (University of Wisconsin).Find this resource:
Katz, J. (1983), ‘A Theory of Qualitative Methodology’, in R. E. Emerson (ed.), Contemporary Field Research (Boston, Mass.: Little, Brown), 127–48.Find this resource:
(1988), Seductions of Crime (Chicago, Ill.: University of Chicago Press).Find this resource:
(2001a), ‘Analytic Induction’, in N. J. Smelser and P. B. Baltes (eds.), International Encyclopedia of the Social and Behavioral Sciences (Oxford: Elsevier).Find this resource:
(2001b), ‘From How to Why on Luminous Description and Causal Inference in Ethnography (Part 1)’, Ethnography, 2 (4): 443–73; (Part 2), Ethnography, 3 (1): 63–90.Find this resource:
Latour, B. (1987), Science in Action (Cambridge, Mass.: Harvard University Press).Find this resource:
(1988), The Pasteurization of France (Cambridge, Mass.: Harvard University Press).Find this resource:
Lindesmith, A. R. (1947), Opiate Addiction (Bloomington, Ind.: Principia).Find this resource:
Lofland, J. (1971), Analyzing Social Settings (Belmont, Calif.: Wadsworth).Find this resource:
(1985), ‘Analytic Ethnography’, Journal of Contemporary Ethnography, 24: 30–67.Find this resource:
Martin, J. L. (2003), ‘What Is Field Theory?’, American Journal of Sociology, 109: 1–49.Find this resource:
Miles, M. B., and Huberman, A. M. (1984), Qualitative Data Analysis (Beverly Hills, Calif.: Sage).Find this resource:
Morrill, C. (1995), The Executive Way (Chicago, Ill.: University of Chicago Press).Find this resource:
Newman, K. (2002), ‘No Shame’, American Journal of Sociology, 107 (6): 1577–600.Find this resource:
(2004), Rampage: The Social Organization of School Shootings (New York: Basic).Find this resource:
Podolny, J. M. (1993), ‘A Status-based Model of Market Competition’, American Journal of Sociology, 98: 829–72.Find this resource:
(2005), Status Signals (Princeton, N.J.: Princeton University Press).Find this resource:
Sallaz, J. J., and Zavisca, J. (2007), ‘Bourdieu in American Sociology, 1980–2004’, Annual Review of Sociology, 33: 21–41.Find this resource:
Salzinger, L. (2003), Genders in Production (Berkeley, Calif.: University of California Press).Find this resource:
Sherman, R. (2007), Class Acts (Berkeley, Calif.: University of California Press).Find this resource:
Snow, D. A., Morrill, C., and Anderson, L. (2003), ‘Elaborating Analytic Ethnography’, Ethnography, 4 (2): 181–200.Find this resource:
Smith, D. L. (2005), Institutional Ethnography (Lanham, Md.: Alta Mira).Find this resource:
Spence, A. M. (1974), Market Signaling (Cambridge, Mass.: Harvard University Press).Find this resource:
Star, S. L., and Griesemer, J. R. (1989), ‘Institutional Ecology, “Translations”, and Boundary Objects’, Social Studies of Science, 19 (3): 387–420.Find this resource:
Strauss, A. L. (1995), ‘Notes on the Nature and Development of General Theories’, Qualitative Inquiry, 1: 7–18.Find this resource:
and Corbin, J. A. (1994), Basics of Qualitative Research (Newbury Park, Calif.: Sage).Find this resource:
Vaughan, D. (1983), Controlling Unlawful Organizational Behavior (Chicago, Ill.: University of Chicago Press).Find this resource:
(1986), Uncoupling (New York: Oxford University Press).Find this resource:
(p. 711) (1992), ‘Theory Elaboration’, in C. Ragin and H. S. Becker (eds.), What is a Case? The Foundations of Social Inquiry (New York: Cambridge University Press), 173–202.Find this resource:
(1996), The Challenger Launch Decision (Chicago, Ill.: University of Chicago Press).Find this resource:
(2002), ‘Signals and Interpretive Work’, in K. A. Cerulo (ed.), Culture in Mind (New York: Routledge).Find this resource:
(2004), ‘Theorizing Disaster: Analogy, Historical Ethnography, and the Challenger Accident’, Ethnography, 5 (3): 313–45.Find this resource:
Theorizing: Analogy, Cases, and Comparative Social Organization (Chicago, Ill.: University of Chicago Press), in progress.Find this resource:
Wacquant, L. (2002), ‘Scrutinizing the Street’, American Journal of Sociology, 107: 1468–532.Find this resource:
(2004a), Body and Soul (New York: Oxford University Press).Find this resource:
(2004b) (ed.), ‘Pierre Bourdieu in the Field’, Ethnography, special issue, 5 (4).Find this resource:
Whyte, W. F. (1943), Street Corner Society (Chicago, Ill.: University of Chicago Press).Find this resource:
Willis, P.  (1981), Learning to Labour (New York: Columbia University Press).Find this resource:
Zerubavel, E. (2007), ‘Generally Speaking: The Logic and Mechanics of Social Pattern Analysis’, Sociological Forum, 22 (2): 131–45.Find this resource:
(*) For helpful comments on an earlier version of this chapter I thank Karen Barkey, Howie Becker, Larissa Buchholtz, Diego Gambetta, Jack Katz, Debra Minkoff, Loic Wacquant, and the editors of this book.
(2.) Several well-known analytic ethnographers have developed theory by cross-case comparison without describing their method (see e.g. Goffman 1961; also Katz 1988). Of those ethnographers who have, none has acknowledged the role of analogy (Eisenhardt 1989: 540–1; Strauss and Corbin 1994; Morrill 1995; Strauss 1995; Snow, Morrill, and Anderson 2003; Zerubavel 2007).