Gary Fine & James Deegan
This article discusses the role of serendipity in qualitative research. Drawing on ideas and methodological suggestions from a set of classic and recent fieldwork accounts, the authors examine conceptions of serendipity and the ways that these conceptions become embedded in the processes by which we incorporate and embrace the temporal, relational, and analytical aspects of serendipity. The authors reject the perspective that it is the divine roll of the dice that determines serendipity and argue that serendipity is the interactive outcome of unique and contingent "mixes" of insight coupled with chance. A wide range of attempts to make sense of serendipity in sociology and anthropology are provided as exemplars of how planned insights coupled with unplanned events can potentially yield meaningful and interesting discovery in qualitative research.
Three goodly young princes were traveling the world in hopes of being educated to take their proper position upon their return. On their journey they happened upon a camel driver who inquired if they had seen his missing camel. As sport, they claimed to have seen the camel, reporting correctly that the camel was blind in one eye, missing a tooth, and lame. From these accurate details, the owner assumed that the three had surely stolen the camel, and they were subsequently thrown into jail. Soon the wayward camel was discovered, and the princes brought to the perplexed Emperor of the land, who inquired of them how they had learned these facts. That the grass was eaten on one side of the road suggested that camel had one eye, the cuds of grass on the ground indicated a tooth gap, and the traces of a dragged hoof revealed the camel's lameness. (Adapted from The Peregrinaggio  in Remer, 1965)
This exotic tale, told of ancient princes of Sri Lanka, then known as Serendip, 1 inspired Horace Walpole, the English novelist (e.g., The Castle of Otranto), politician, and belle lettrist. In this last capacity, Walpole coined the term 'serendipity" while writing to the British diplomat, Horace Mann, in January 28, 1754. Walpole created serendipity to refer to the combination of accident and sagacity in recognizing the significance of a discovery (Remer, 1965, pp. 6-7).
Accident and sagacity in discovery
Since Walpole's day, the power of serendipity has been rhapsodized by observers of the scientific method (Canon, 1945). Beveridge (l957) claimed:
Probably the majority of discoveries in biology and medicine have been come upon unexpectedly, or at least had an element of chance in them, especially the most important and revolutionary ones. It is scarcely possible to foresee a discovery that breaks really new ground, because it is often not in accord with current beliefs. (p. 31)
Newton's discovery of gravity from an apple falling on his head while he rested beneath a tree is more than a bump in the annals of scientific discovery. This account of Newton's experience privileges chance observation. Whether, however, the story of the discovery of Newton's Second Law is apocryphal, others, including Bequerel's discovery of radioactivity arising from his unwitting tossing of uranium salts into a drawer with photographic materials (Badash, 1965), Tombaugh's discovery of Pluto on the basis of Lowell's flawed calculations (Tombaugh & Moore, 1980), and Penzias and Wilson's (1965) Nobel-winning identification of "cosmic microwave radiation" are accounts grounded on actual happenings. In addition, Kirk and Miller (1986) provide many examples of new discoveries in the biomedical sciences "that occur only in consequence of some kind of mistake" (p. 16), including such schoolboy classics as Fleming's (1946) naming as penicillin the mold that got into his staphylococcus culture and destroyed the bacteria.
The annals of scientific discovery do not merely imply that serendipity is an unusual happening, but that the scientist is "prepared" to make sense of a truer picture of the world, creating a more precise model, such as the discovery by Penzias and Wilson, which is "now considered basic evidence for the truth of the 'standard model' of the universe" (Kirk & Miller, 1986, p. 16). With the wisdom of hindsight, Kirk and Miller remind us that "the first men to hear the echo of the origin of the world thought they were listening to a guano" (p. 16). Our conceptualization, however, goes beyond the banal and well-recognized proposition that chance plays a role in research. Scientists have noted, "chance favors only those who know how to court her" (Charles Nicolle, quoted in Beveridge, 1957, p. 27), and "In the field of observation, chance favors only the prepared mind" (Louis Pasteur, quoted in Beveridge, 1957, p. 34). Consequently, the initial moment of discovery in clinical or field-bound situations (for those fortunate enough to have this experience) invariably evokes the sense that the whole project is turning to dust. The difference lies in being prepared to turn what seems like the ashen remains of a project into a creative opportunity for scientific discovery. In this way, courting serendipity involves planned insight married to unplanned events.
This "Newtonian" narrative of discovery is consistent with the now dominant social constructivist models in the sociology of science, most notably Thomas Kuhn's Structure of Scientific Revolutions (1962; also see Dean, 1977). Kuhn argued that a scientific breakthrough is unpredictable and not derivative from normal scientific work, but depends on a critical, unexpected insight that leads to a "better" way of understanding empirical relations: a new paradigm. This view of discovery underlines the recognition that scientific work is a "messy" process, but it also asserts that an approximation to truth is possible.
Ethnographers tend, however, to discount this natural science rhetoric of discovery of the "new and better," replacing it with a perspective that presumes that ethnographers are engaged in providing "deeper understandings of things many people are already pretty much aware of" (Becker, 1982 p. x). It is not that they accidently stumble on truth, but that they can find accounts that others find useful in making sense of the world. John and Lyn Lofland (1984) emphasized this point:
Because all social analysis plays off what is ordinarily believed or felt to be known an analysis is interesting only in so far as it departs from what is already "obvious." And because what is seen as obvious is itself changing, what is or will be regarded as interesting is also changing. (p. 127)
We define serendipity, then, as the unique and contingent mix of insight coupled with chance. It is conclusions that are defined as surprising, yet obvious, that contribute to the presentation of "interesting" research (Davis, 1971).
Serendipity in classical fieldwork
The spark for our thinking about the part that serendipity plays in qualitative research originated from readings of a set of classic fieldwork accounts. Our goal was to learn how ideas are developed and presented in qualitative research, and how these ideas become privileged. By exploring the writings of Malinowski (1950), Powdermaker (1966), Wax (1971), and Whyte (1943/1955), among others, we hoped to discover the conditions under which classic status is granted to research documents. And, in fact, we noticed a pattern in these writings in which happenstance was linked to the ethnographer's ability to make sense of seemingly chance events; that is, these scholars were able to "keep their wits about them," finding, in the rush of ongoing events, meanings and opportunities that might escape others.
Qualitative research inevitably contains such "good fortune," but serendipity consists in how we transform our fortune into substantive discovery. Agar (1982), though, brings a Heideggerian sense of "breakdown" to this discussion of the ethnographic process of transforming good fortune into substantive discovery. In a useful concept-splitting exercise, Agar distinguished between occasioned ("those that come up in the course of doing ethnography" [p. 787], core ("those that constitute the focus of an ethnographer's work and eventual report" [p. 788]," derivative ("those that are less important for an ethnographer" [p. 787]), and mandated ("those that one sets out to create" [p. 788]) breakdowns. More simply:
At one extreme, an ethnographer may set out to force a breakdown. At another extreme, unexpected breakdowns may come up and receive less attention - they are occasioned and derivative. However, it is one of the special strengths of ethnography that a breakdown that was originally mandated disappears or becomes derivative, while something that came up serendipitously as an occasioned breakdown moves to the center and becomes core. (Agar, 1982, pp. 787-788)
Agar's work serves as a caution against placing disproportionate emphases on harmony to the exclusion of conflict in ethnographic sense-making. Agar illustrates the generative spiralling of breakdowns and resolutions with reference to Rabinow's (1977) account of his day-to-day experiences in the marketing and commercial center of the town of Sefrou on the eve of neo-colonialism in Morocco. Rabinow discussed the liminal aspects of his close relationship with Ali as follows:
This was the beginning of the dialectic process of fieldwork. I say dialectic because neither the subject nor the object remain static. With Richard or Ibrahim, there had been only minor movement on either wide. But with Ali there began to emerge a mutually constructed ground of experience and understanding, a realm of tenuous common sense which was constantly breaking down, being patched up, and reexamined, first here, then there. (p. 39)
Since Malinowski (1950), many fieldwork classics provide evidence of the importance of interpreting and capitalizing on unpredicted, unplanned events. Yet, traditionally, ethnographers were reluctant to discuss their errors and chance occurrences, even when these events proved to be the basis of subsequent insight, perhaps fearing that it would confirm the belief that ethnography was truly dilettantism. Hortense Powdermaker (1966) recognized this absence when she remarked:
Little record exists of mistakes and learning from them, and of the role of chance and accident in stumbling upon significant problems, in reformulating old ones, and in devising new techniques, a process known as "serendipity." A lack of theory, or of imagination, an over commitment to a particular hypothesis, or a rigidity in personality may prevent a fieldworker from learning as he stumbles. (pp. 10-11)
With the growth of the "reflexive turn" in ethnography - what some have labelled the "new ethnography" (Dowd, 1994), the inclusion of occurrences of serendipity in accounts of fieldwork is a battle won long ago, perhaps contributing to the heroic image of the ethnographer who pulls meaning from chaos. We have come to present ourselves as lovers of the play and surprise of research. Although we now have what Atkinson (1990) described as a "mythological corpus" of ethnographers' tales of discovery - frequently in the form of " confessionals" (Van Maanen, 1988), we know little of how serendipity operates in qualitative research. The conceptualization of the dimensions of serendipity must be made more explicit. The question becomes: How do our own lived experiences of insight lead to substantive discovery?
Our goal is to examine not only the inevitable accidents and sagacities that constitute qualitative research - the "disorderly," "messy" features of the research process - but also their substance, method, and significance. Our approach owes much to Hammond's (1967) writing on "chronicling" the unique and contingent "mix" of rational, serendipitous, and intuitive phenomena in the contexts of discovery in social scientific research. "Traces" of serendipity are pervasive in qualitative research; no accounts, however, exist of how qualitative researchers conceptualize serendipity and how these conceptions become embedded in the process by which we incorporate and embrace the temporal, relational, and analytical aspects of serendipity.
Serendipity as controlled chaos
Although discussions of the position of chance findings have often been linked to experimental research, the intersection of insight and unplanned discovery is also at the heart of field research, as examples from sociology and anthropology attest. Indeed, it sometimes appears that qualitative researchers revel in their lack of control over the broad swath of events that they observe and, simultaneously, revel in their perspicuity as observers of the subtle and non-obvious. Lofland and Lofland (1984) wrote:
Naturalists in the social sciences are engaged in a strategy of calculated chaos. They intentionally immerse themselves in the logging of data regarding subjects that are of personal concern to them, a process that initially need have little or no specific social scientific orientation. The theory of the naturalist is that a direction will emerge, will be "discovered." (p. 69)
We believe, along with the Loflands' naturalists, that productive challenges to tacit and taken-for-granted understandings of participants and settings emerge in the initial "open spaces" of fieldwork. Errors, stumblings, and surprises are particularly likely to occur - before our workspace becomes routine. Qualitative researchers are explorers making "first contact" with alien civilizations, courting experience through observing diverse times and places, later reanalyzing and rewriting, hoping for heightened awareness to ignite insight. But this insight is not a treasure at the end of the road for the Princes of Serendip; it is one that unfolds with every twist and turn in the road.
The most influential attempt to apply the concept of serendipity to social scientific theorizing has been the one by Robert Merton (1968, pp. 157-162; see Berardo & Shehan, 1984). As Merton (1962) noted, "There is a rich corpus of literature on how social scientists ought to think, feel, and act, but little detail on what they actually do, think, and feel" (p. 19). Merton (1968, p. 157) provided a systematic attempt to make sense of serendipity in sociology, speaking of the serendipity pattern, whereby unexpected data provide the spark for the creation of theoretical analysis. For Merton (1968, p. 157) three features characterize datum that fit into a serendipity pattern: it must be "unanticipated," "anomalous," and "strategic" (i.e., with implications for the development of theory).
Merton, of course, operated from the scientific model described above, which is also implicated in the princes' tale. That is, a real world exists for which clues provide insight. In contrast to a positivist (or postpositivist) view, we suggest that serendipitous insight provides the opportunity for constructing a plausible story. We do not deny the reality of an external world, but only suggest that numerous possible explanations exist and that chance events can be made serendipitous if the event provides the opportunity for story-building. In this way, story-telling is a means, not an end. We use stories in much the same way as researchers might use an illustrative case decorating a statistical study. Our stories are intended as supporting evidence for the paper's conclusions and, it is hoped, permit the reader to experience an abbreviated version of the verstehen and inference processes of the researcher.
In extending Merton's serendipity pattern to the doing of ethnographic work, we focus on the opportunities that chance provides. Specifically, we wish to examine a trio of analytically distinct components of research: temporal serendipity (happening upon a dramatic instance), serendipity relations (the unplanned building of social networks), and analytic serendipity (discovering concepts or theories that produce compelling claims). Each depends on the readiness to seize upon chance events; that is, the unstructured, inductive quality of fieldwork often provides leeway to incorporate the power of serendipitous findings into the core of a research report.
Being exposed to a particularly dramatic event can, at times, transform a mundane ethnography into a classic. Missing the same event, without depicting equivalently dramatic occurrences, can make an ethnography seem pedestrian. We enshrine the power of "being in the right place at the right time," even while realizing the number of potentially right places and times. In a "classic" ethnography, perhaps two or three events define the ethnographical work and make it memorable to its reading audience. The observer cannot choose in advance to witness an event; his or her presence is, in part, a function of the decision of the observer to judge "where the action is."
Even though events are unpredictable, types of events tend to occur regularly at particular times (Melbin, 1987), and we depend on this patterned quality of events to permit us to make temporal choices about when to observe. If one hopes to observe barroom brawls, one would be wise to observe on Saturday evenings, rather than on Tuesday afternoons. If one wishes to examine preadolescent friendships and insults, one would be advised to watch recess, rather than mathematics class (Deegan, 1993; Fine, 1987), For instance, in Fine's study (1987) of Little League baseball, he learned that games between two highly competitive teams were more likely to generate conflict than games between teams that did not care whether they won or lost. Criminal activity or abusive behavior, while not predictable, are also not temporally or spatially random (Klockars, 1974). The prepared participant observer hoping to maximize the chances of obtaining memorable data selects just the right time and just the right place.
Of course, serendipity presumes not only that one will observe memorable events, but that they will be recognized as significant when they occur and will be shaped into powerful narratives for jaded readers. What marks the profound observer from the casual one is the ability to see a pattern or implication that has gone unnoticed and, having exposed it, to find it in other social settings.
Despite our wistful hope that readers will examine our writings with diligent care, most ethnographies are only known by a few dramatic scenes. Street corner society (Whyte, 1943/1955) is known as much as anything, among all the material recounted, for the bowling scene and the multiple voting at election time. Readers who examine the layered meanings of Whyte's work discover not only a portrayal of a bowling scene, but also the sociological themes of social differentiation and reproduction embedded in the "culture of bowling." Similarly, as recounted in The interpretation of cultures, Clifford Geertz (1973, pp. 413-414) decided to be present at that now celebrated Balinese cock-fight because that was where the action was. He was able to witness a dramatic police raid of the event that he subsequently employed as a lens to understand Balinese character and cultural traits. The academic memory of that dusty cock-fight is now indelibly linked to Geertz, as tightly as any of his theoretical contributions.
Much of what we readers are exposed to, we rapidly discard, and writers must present narrative accounts of such indelible clarity that we are not likely to miss their import. Indeed, for Geertz and Whyte startling fact became rich narrative. Those with such raw materials are the lucky ethnographers. Discovering novel data sources is part of the serendipity pattern (Snow & Anderson, 1993, p. 323).
In preparing for this paper, we tested ourselves by thinking about our memories of a set of ethnographies and their authors. For many of these ethnographers, we could only recall the group observed and a few theoretical and empirical generalizations. We could recall specific events for only a minority of works. Most of these events were unpredictable occurrences that an observer might easily have missed if tired or nursing a cold.
Who one knows is, within bounds, a matter of chance. A rich corpus of contemporary field research has addressed relational and personal processes in fieldwork (Douglas, 1972; Emerson, 1983; Johnson, 1975; Klockars, 1974; Peshkin, 1986; Wax, 1971). While there are categories of individuals with whom one is relatively likely to establish relations, our specific contacts cannot be predicted in advance. One's true love was selected from a pool of eligibles, most of whom one never gets to compare. Ethnography is preeminently a methodology that depends on relationships. The relationship between the researcher and the subject of the research is less significant in experimental, survey, or historical research, while little could be more central in ethnographic research.
The direction of a course of analysis and the research questions asked can be influenced by the alliances a researcher makes in the early stages of a project. It is not sufficient that one makes contact (good fortune), but one must also be able to capitalize on this contact (serendipity). Central to Whyte's (1943/1955) study was his friendly field partnership with Doc and his boys. Doc grasped his role as a sponsor and gatekeeper who facilitated Whyte's rapport with other community members. Whyte's alliance with Doc led to others in the community who were largely ignored as well as to those who were disliked in the community (Fine, 1993). Ultimately, this friendship with Doc led to a greater understanding of community patterns that might have remained opaque, 2 and we recognize such alliances by enshrining the concept of "key informant." But how did this "key informant" relationship develop? In most cases, it did not develop through a conscious selection by the researcher. This image may be cynically romantic (as in narratives of strategically manipulating one's future spouse to fall in love), but the reality often seems a good deal messier - based on happenstance, luck, or mistaken identity.
Admittedly the competent ethnographer will do whatever possible to foster productive or desirable relationships, even providing services to create a favor bank. These plans, however, often go awry, if the ethnographer is too demanding. In fact, the events that cement friendship are often those that are unpredictable. For instance, David Snow and Leon Anderson (1993) reported that their acceptance by key members of the homeless community they studied was possible because one of the two researchers was arrested and jailed for vagrancy. They had "unwittingly negotiated an important rite of passage" (p. 322). Geertz (1973, p. 416) reported similar acceptance for having fled the cock-fight with the rest of the village. Chance events provided opportunities for these researchers to demonstrate trust - although in other circumstances, instances of failure to demonstrate that trust also occur and may even typify the researcher (see Mitchell, 1993).
Though we can find variables that contribute to the likelihood of a social tie emerging and flourishing, relationships develop from a situationally mysterious process in which physical presence, chance remarks, and unplanned actions connect persons to each other, leading to other relationship opportunities. For instance, when Edna Salamon (1989, p. 2) appeared for a scheduled interview with the manager of a London escort service, he assumed that she was searching for employment. When he learned she was not, he was so amused (or perhaps embarrassed) that he consented to let her study his agency, which specialized in homosexual liaisons.
The ethnographer depends on the knowledge and commitment of his or her key informants. The interests, friends, and opportunities of the key informant become in some measure those of the researcher. If these interests and contacts are similar to those that the researcher had intended to study, the research can proceed as planned. If the interests and contacts diverge, the researcher will either have to alter the focus of the research, or, potentially more damaging to the study, collect thin data. Alternately, one must separate oneself from the informant and begin again.
Fortunately, for the doing of qualitative research (as with a marriage market), several potential key informants exist and each can contribute to a successful project. Research can be conducted with many good contacts, rather than with a unique, heroic one. The question is not about establishing relations with only the right person, but rather whether the researcher can make use of the relationships. To each contact the researcher must communicate persuasively. Powdermaker's (1966, p. 139-40) account of her first days in the rural community of Indianola, Mississippi, is a compelling description of a fieldworker using her wits to build rapport. Following a "flowery" introduction from Mr. Green (the Rockefeller Foundation representative for Negro education) to the leading citizens of Sunflower County at the court house, Powdermaker decided that the less she said at the time, the better. Powdermaker (1966) proved adroit in dealing with the fear and hostility of the leading citizens of Indianola in this initial encounter:
The first question: Are you interested in changing the status quo? My answer: "I'm here to study it", which happened to be the truth. "Are you going to publish newspaper articles when you leave," accompanied by statements about Yankee journalists who spend a few days in the south then wrote lurid articles about it. I answered quite truthfully that I would not write newspaper articles. "What are you going to write?" "A scientific book." (p. 40)
Being aware of the politics and diplomacy of building relationships kept Powdermaker in good stead in an environment in which a wrong step could have led to suspicion and exclusion. Her experiences, thus, highlight the importance of making choices while being aware of communal norms, especially during the critical early phases of a fieldwork project.
Each relationship, then, provides a piece of the puzzle, although it is a puzzle of mysterious design, only known when the researcher has decided that it is close enough to completion. The material from each individual must be interpreted in a different light, by means of the information provided by others. Just as hearing the chance remark allows us to reevaluate other information, so developing a contact allows us to understand better those other contacts that we have cultivated.
The third serendipitous feature of qualitative research involves the ability to establish connections between data and theory. The formal model of deductive research assumes that one knows what one is looking for before one has "found" it. Inductive research elevates "insight" or serendipity into the chosen stance of analysis, permitting numerous conclusions from the same data. We are each a Prince of Serendip.
While there are many social science examples (e.g. Johnson & Kaplan, 1987), Merton (1968, pp. 159-162) is the best known exponent of serendipity and thus deserves special attention. Studying a suburban development, Merton discovered that upon moving, families with young children increased their social and community participation, claiming that they could do so because of the abundance of available babysitters. Merton calculated, however, that the proportion of teenagers was actually lower in this community than in the communities from which these parents moved. The teenagers who resided in this new locale were defined as reliable and trustworthy because the parents were known. As Merton (1968) noted, in good interactionist fashion, relying on concepts of confidence and social cohesion, "It is not that there are objectively more adolescents in Craftown, but more who are intimately known and who, therefore, exist socially for parents seeking aid in child supervision" (p. 161). The development of this theoretical proposition was made possible by the happenstance of finding anomalous data, taken in context with known demographic information and his insight about the role of community membership.
By what processes does this insight occur? To be sure, much of this process cannot be easily determined. We establish linkages without being certain why they make sense. Despite this seeming unpredictability, several features help to explain the process of analytical serendipity.
First, the researcher has previously been exposed to the relevant literature. These previous studies provide a grid in which new data can be incorporated, creating a template for the development of new theory. Suddenly, one sees relevance where none was noticed before. Theory never develops out of thin air, but is responsive to those intellectual currents that are in circulation and to which the researcher has been exposed. These influences may be tied to disciplinary scholarship or even fiction or poetry. "Malinowski's Fraserian apprenticeship and perhaps also those tent-bound bouts of novel reading in the Trobriands" (Stocking, 1992, p. 53) is a good example of how literary and social domains of reality could be combined. Likewise, Powdermaker (1966) writes of her "omnivorous" readings of the fictional worlds created by Hawthorne, Dostoyevsky, and Thackeray, fictional worlds that provided models for insight outside her own immediate environment. Fine's (1983) study of fantasy role-playing gamers was enriched through his guzzling of science fiction texts, the reading of which immersed him in a distant culture. Moreover, those studies that often have the most immediate impact on our research are those that we happen to read or learn about during the research. The chance publication of an article or a hallway conversation in the midst of a research endeavor may have an enormous effect in altering one's theoretical analysis. The researcher who is open to those new sources has an advantage in creating new theoretical models.
Second, the data themselves speak to the researcher. On occasion portions of the data play off each other. The unexpected similarity or dissimilarity may provoke an "Ah-ha!" response. Suddenly, as in scientific revolutions, anomalous data fit, because of their relations to each other or to some aspect of the literature. One example shines a light upon another. For instance, in Fine's (1988) research on mushroom collectors, he was surprised by the extensive joking about dying from mushrooms and using mushrooms as murder weapons. He, thus, searched for the means to make sense of this, eventually focusing on the political economy of trust and finding comparable examples in other settings (for example, parachuting, mountaineering). Anomalous data were tied to the literature by being connected to theoretical constructs.
Third, the researcher may discover a dramatic metaphor or narrative strategy that permits him or her to conceptualize and present the problem in a novel light. In practice, the social sciences are comprised of bundles of slogans, images and maxims. The catchy ones - perhaps uncomfortably similar to those advertising slogans that we profess to scorn - prove powerful. Consider the following: the fate of idealism, impression management, definition of the situation, awareness contexts, negotiated order, grounded theory, or the strategic perspective. These tags gain a life of their own. The ethnographer who creates such an image can then rely upon the power of that image to generate insight. One may begin with a single image, which then is elaborated by a motivated audience. The rich, unexhausted imagery of Mills's (1959) construct of "the sociological imagination" continues to attract fresh audiences in ways that the author may never have intended (Atkinson, 1990; Denzin, 1991).
As discourse, claims of serendipity connect to what Lofland and Lofland (1984) described as the "transcending perspective" that "things that are not (or are not only) what they seem to be" (p. 122). The world must be interpreted to be made interesting. We rely upon metaphorical and ironic rhetorical devices, gathered through our reading, thinking, and talking, to provide dramatic images to make field tales interesting and to share our consciousness of discovery. As Brown (1989) stated with regard to irony, certain "incongruities" are ironic because they pose "logical contradictions, rather than mere opposition or reversals of chance events" (p. 217). The pleasures of Erving Goffman's Asylums (1961; also, see Fine & Martin, 1990) consist in considerable measure in his ability to use irony, sarcasm, and satire to make powerfully unexpected connections. His serendipitous analysis of the depressingly mundane permanently transforms our understanding of institutions to images that we then use ourselves.
Fourth, a research may be influenced by being part of a scholarly world: a system of collective action (Becker, 1982) or invisible college (Crane 1972). The production of knowledge operates through this set of social ties. Through these ties, a shared intellectual community, fresh insights may arise. For example, in analysing art worlds, Becker (1982) noted that it is the network of "cooperative links," rather than the lone artist, that is central to the production of art.
We learn about collective action in drama by reading theater programs; acknowledgements are the programs of academics. For example, in Fast food, fast talk, Robin Leidner (1993, pp. vii-viii) thanks 5 organizations (Combined Insurance, McDonald's [her research sites], Northwestern University, the National Science Foundation, U.S. Department of Education [her funders]) and 30 individuals by name (and many others implicitly), including key players in her ethnographic sites, her dissertation committee ("all the assistance that one could expect of a committee"), her dissertation chair ("offering good food and friendship"), academic colleagues ("unfailingly sisterly and stimulating companions"), and her editor ("patient and supportive"). Leidner's depiction of her social world is duplicated in the early pages of most academic studies, making the case that the final product would be a very different piece of work without these contacts.
Other links are routinely left for the last paragraph in the acknowledgements section, a temperature gauge to one's most intimate relations, mediated by the need to make oneself seem happier than one often feels (for a discussion of the structure of acknowledgements, see Goffman, 1974). These figures suffer the pain of our frustrations, the pleasures of our satisfactions: our "dear and dependable friend[s]" (Leidner, 1993, p. viii). Reflecting on a lifetime of writing, Anthony Trollope, the English novelist, wondered if he owed more to the old groom who unfailingly brought him a cup of coffee at 5:30 a.m. every morning and helped him begin his daily writing routine than any one else (cited in Becker, 1982, p. 1). We make use of those who provide us with ideas, emotional support, and material pleasures, building a support system drawn from whom we chance to know, deploying them to our best advantage.
Keeping one's wits
Part of serendipity derives from those unplanned happenings that stem from one's own hands This involves the powerful role of mistakes leading to insight: a messiness that stems from the investigator (Fine, 1993). At one point there was something faintly embarrassing in talking about one's errors. Increasingly, however, in confessional ethnographies descriptions of learning from mistakes have been incorporated in qualitative research reports, perhaps glorifying the research role by demonstrating the power to persevere and overcome. Mistakes may be treated not only as unavoidable errors, but as events that uncover the preconceptions and choices of the researcher.
In Philip DeVita's aptly titled, edited collection, The naked anthropologist, authors wrestle with their search for truth and attempt to make sense of their surprises and gaffes. DeVita (1991) captures the essence of learning about ourselves from others in his preface to the collection of essays:
These were tales of human experience, where in most instances, the ethnographic stranger stumbled into a situation, where he or she learned something for which he or she had not been trained or prepared... In nearly all cases, there were significant contrastive lessons learned about issues of cross-cultural humanity and humanness, derived more often than not from serendipity than from the delicate practice of social science. (p. xvi)
Linda Kent's essay on "failed" research provides a potent example of fieldwork attempts that have "fizzled." Out of the ashes of two frustrating fieldwork experiences, one with the Gypsies in New Orleans, the other with the Irish Travellers in Mississippi, Kent learned that her errors of insight could contribute to future insight. She could not "find" a Gypsy to study in New Orleans, and her failure to recognize that the Gypsies and Travellers were enemies in her Mississippi study left her "naked" in the tongue-in-cheek way that DeVita (1991) uses this metaphor. Reflecting later on her loss of focus and spirit while in the field, Kent (1991) concluded, "It is through failures as well as successes, both our own and those of others that we learn about being human" (p. 23). This insight that remained latent in the earlier research emerges as a touchstone in her more successful dissertation fieldwork with Irish Travellers the following year.
As Hortense Powdermaker (1966) discovered, being" prepared" was not exclusively a matter of packing a folding army cot, portable typewriter, and mosquito boots. Neither was being prepared exclusively a matter of months of preparation or learning the vernacular. Powdermaker's classic Stranger and friend (1966) reveals that who we are is related to how we do in qualitative fieldwork. Serendipitous events are opportunities for staging interaction and the creation of the self. It matters little that Powdermaker, like us, stumbles. What matters is that she rallied as a result of her keen sense both of involvement and detachment in participant observation, turning adversity to account. Learning how to learn from mistakes is critical for using serendipity in qualitative research.
Rosalie Wax's (1971) writings on the use of one's "wits" in anthropological research complements Powdermaker's emphasis on learning as you stumble, although Wax noted that luck is indeed a "gift of the Powers" (p. 268) and cannot be acquired by determination alone. She ranked as the next most valuable element what the Scandinavians called "manvit:" "intelligence manifested in common sense, shrewdness, and flexibility - the property called having one's wits about one" (p. 268). Attributes such as humility, empathy, maturity, energy, determination, and creativity are not ends in themselves, but are means through which rapport can be established, data gathered, and theory generated. As Wax (1971) noted, "If [the fieldworker] does not have his wits about him while he is in the field, the chances are that he will not be given the opportunity to exercise any of his other attributes or virtues" (p. 268).
While we do not claim that ethnographic findings are random, we need to embrace the chance component of research as being central to the collection and interpretation of data. Chance has quite a different meaning in this conceptualization. Unplanned does not suggest that anything is possible, only that a range of things are. What is previously unexplained may be a result of contingencies that are not fully comprehended (Manis & Meltzer, 1994); exposing oneself to the unplanned may uncover these previously hidden contingencies. Examining the unexamined mix of contingencies produces the recognition of coincidences as meaningful (Becker, 1994).
Serendipity involves planned insight coupled with unplanned events, core to the philosophy of qualitative research. By recognizing the centrality of serendipitous findings and events in qualitative research, we come closer to understanding how research products are created and appreciated in practice. In a methodology that so values insight, the thin line between brilliance and nothingness is both a powerful image and vast chasm: our fear and our salvation. When a researcher prepares to enter a field setting, the worry exists that nothing interesting will be discovered. Such fear, however, fails to reckon with the intellectual preparedness necessary to make sense of the power of an ongoing social reality.
The display of serendipitous claims is, in part, a rhetorical strategy. Admitting learning from one's errors validates one's work, particularly in a culture in which transcending obstacles is valued. To be sure, these accounts of error and chance are tied to a sense of professional stability and likely would be downplayed in the writing of graduate students and others who consider themselves vulnerable to professional scorn. One can only denigrate oneself if one is confident that one's audience will discount that ploy, accepting the subsequent image of " transcendent genius."
Ultimately, we reject the perspective that it is the roll of the divine dice that determines if anything interesting is to be learned. It is through our intellectual readiness, coupled with exposure to a wide range of experience, that we create a sense of lived experience - ours and others. Each researcher must be ready to seize the clues on the road to discovery. We need to consider the fullness of the markings of serendipity, just as the ancient Princes of Serendip did.
A version of this article was presented to the qualitative Research Conference, Waterloo, Ontario, May, 1994 and, to a colloquium at the Department of Social and Behavioral Sciences at the University of California, San Francisco. The authors wish to thank Sharon Connolly for her assistance in the preparation of this article, and Martha Allexsaht-Snider, Adele Clarke, Penny Oldfather, Virginia Olesen, and Leonard Schatzman for comments on a previous draft. This paper was partially prepared while the first author was a Fellow at the Center for Advanced Study in the Behavioral Sciences. He is grateful for financial support provided by National Science Foundation No. SBR-9022192. The authors shared equal time and effort in the planning, preparation, and writing of this article.
1. According to Leo A. Goodman (1961, p. 456), Serendip is a corruption of the Hindustani name for Ceylon, Saradip, meaning "Golden Isle," where according to legend a city of gold was built.
2. The debate over the validity of Whyte's interpretation of the North End and his relationship with Doc (Boelen, 1992; Orlandella, 1992; Whyte, 1992) reminds us that our knowledge of the researcher's relationships is generally filtered through self-interested reportage.
Agar, M. H. (1982). Toward an ethnographic language. American Anthropologist, 84, 779-795.
Atkinson, p. (1990). The ethnographic imagination. London: Routledge.
Badash, L. (1965). Chance favors the prepared mind: Henri Becquerel and the discovery of radioactivity. Archives of the International History Society, 18, 22-26.
Becker, H. S. (1982). Art worlds, Berkeley, CA: University of California Press.
Becker, H. S. (1994). 'Foi por acaso': Conceptualizing coincidence. Sociological Quarterly, 35, 183-94.
Berardo, F.M., & Shehan, C. L. (1984). Family scholarship: A reflection of the changing family. Journal of Family Issues, 5, 577-98.
Beveridge, W. I. B. (1957). The art of scientific investigation. New York: W. W. Norton.
Boelen, WA. M. (1992). Street corner society: Cornerville revisited. Journal of Contemporary Ethnography, 21,
Brown, R. H. (1989). A poetic for sociology. Chicago: University of Chicago Press.
Cannon, W. B. (1945). The way of an investigator. New York: W. W. Norton.
Crane, D. (1972). Invisible colleges. Chicago: University of Chicago Press.
Davis, M. (1971). That's interesting! Toward a phenomenology of sociology and a sociology of phenomenology. Philosophy of Social Science, 1, pp. 309344.
Dean, C. (1977). Are serendipitous discoveries a part of normal science? The case of the pulsars. Sociological Review, 25, 73-86.
Deegan, J. G. (1993). Children's friendships in culturally diverse classrooms. Journal of Research in Childhood Education, 7, 91-101.
Denzin, N. K. (1991). Learning from Mills. In N. K. Denzin, (Ed.), Images of postmodern society: Social theory and contemporary cinema (pp. 53-64). Newbury Park, CA: Sage.
DeVita, p. R. (ed.). (1991). The naked anthropologist: Tales from around the world. Belmont, CA: Wadsworth.
Douglas, J. D. (1976). Investigative social research. Beverly Hills, CA: Sage.
Dowd, J.J. (1994). The tension of honest writing. Unpublished manuscript.
Emerson, R. M. (1988). Contemporary field research. Prospect Heights, IL: Waveland Press.
Fine, G. A. (1983). Shared fantasy: Role-playing games and social worlds. Chicago: University of Chicago Press.
Fine, G. A. (1987). With the boys: Little league baseball and preadolescent culture. Chicago: University of Chicago Press.
Fine, G. A. (1988). Dying for a laugh: Negotiating risk and creating personas in the humor of mushroom collectors. Western Folklore, 47, 177-194.
Fine, G. A. (1993). Ten lies of ethnography: Moral dilemmas of field research. Journal of Contemporary Ethnography, 22, 267-294.
Fine, G. A., & Martin, D. D. (1990). A partisan view: Sarcasm, satire, and irony as voices in Erving Goffman's Asylums. Journal of Contemporary Ethnography, 19, 89-115.
Fleming, A. (ed.). (1946). Penicillin: Its practical application. Philadelphia: Blakiston.
Geertz, C. (1973). The interpretation of cultures. New York: Basic books.
Goffman, E. (1961). Asylums. New York: Anchor.
Goffman, E. (1974). Frame analysis. Cambridge: Harvard University Press.
Goodman, L. A. (1961). Notes on the etymology of serendipity and some related philological observations. Modern Language Notes, 76, 454-457.
Hammond, p. E. (Ed.). (1967). Sociologists at work. Garden city, NY: Doubleday.
Johnson, J. M. (1975). Doing field research. New York: Free Press.
Johnson, R.J., & Kaplan, H.B. (1987). Corrigendum: Methodology, technology and serendipity. Social Psychology Quarterly, 50, 352-354.
Kent, L. L. (1991). Fieldwork that failed. In p. DeVita (ed.), The naked anthropologist: Tales from around the world (pp. 17-25) Belmont, CA: Wadsworth.
Kirk, J., & Miller, M. L. (1986). Reliability and validity in qualitative research. Beverly Hills, CA: Sage.
Klockars, C. B. (1974). The professional fence. New York: Free Press.
Kuhn, T. 5. (1962). The structure of scientific revolutions. Chicago: University of Chicago Press.
Leidner, R. (1 993). Fast food, fast talk: Service work and the routinization of everyday life. Berkeley: University of
Lofland, J., & Lofland, L. (1984). Analysing social settings. Belmont, CA: Wadsworth.
Malinowski, B. (1950). Argonauts of the Western Pacific. New York: E. p. Dutton.
Manis, J. G., & Meltzer, B. N. (1994). Chance in human affairs. Sociological Theory, 12, 45-56.
Melbin, M. (1987). Night as frontier. New York: Free Press.
Merton, R. K. (1962). Science and the social order. In B. Barber & W. Hirsch (Eds.), The sociology of science (pp. l~28). Glencoe, IL: Free Press.
Merton, R. K. (1968). Social theory and social structure. New York: Free Press.
Mills, C. W. (1959). The sociological imagination. London: Oxford University Press.
Mitchell, Jr., R. (1993). Secrecy and fieldwork. Newbury Park, CA: Sage.
Orlandella, A. R. (1992). Boelen may know Holland, Boelen may know Barzini, but Boelen doesn't know diddle about the North End. Journal of Contemporary Ethnography, 21, 69-79.
Penzias, A. A., & Wilson, R. W. (1965). A measurement of excess antenna temperature at 4080 Mc/s. Astrophysical Journal, 142, 4 l9-421.
Peshkin, A. (1986). God's choice: The total world of a fundamentalist Christian school. Chicago: University of Chicago Press.
Powdermaker, H. (1966). Stranger and friend: The way of an anthropologist. New York: Norton.
Remer, T. G. (1965). Serendipity and the three princes. Norman, OK: University of Oklahoma Press.
Salamon, E. (1989). The homosexual escort agency. British Journal of Sociology, 40,1-21.
Snow, D., & Anderson, L. (1993). Down on their luck: A study of homeless men. Berkeley: University of California Press.
Stocking, G.W. (1992). The ethnographer's magic and other essays in the history of anthropology. London: University of Wisconsin Press.
Tombaugh, C. W., & Moore, p. (1980). Out of darkness: The planet Pluto. London: Lutterworth Press.
Van Maanen, J. (1988). Tales of the field. Chicago: University of Chicago Press.
Wax, R. (1971). Doing fieldwork. Chicago: University of Chicago Press.
Whyte, W. F. (1943/1955). Street corner society. Chicago: University of Chicago Press.
Whyte, W. F. (1984). Learning from the field. London: Sage.
Whyte, W. F. (1992). In defense of Street corner society. Journal of Contemporary Ethnography 21: 52-68.