Keywords

Research methods are ideally value-neutral—that is, each method is a potential means of gathering information about the world—but researchers develop “favorites” and academic disciplines value certain methods over others. The study of human sexuality often requires a researcher to work across disciplinary boundaries, however, and to deploy multiple methods. In my own research, I have used observation, participant observation , multiple in-depth interviews with each participant, one-time interviews, surveys, case studies, archival research, and so on. Each method, I have come to believe, has strengths and weaknesses. We should choose our methods not on tradition (“everyone in anthropology does ethnography” or “if you don’t collect quantitative data, no one will take you seriously in sociology”) but for their appropriateness for the questions being asked, the research setting, and each researcher’s traits, skills, and personality.

Many researchers have noted the difficulty in obtaining data on human sexual behavior (Berk 1995; Orbuch 1991; di Mauro 1995). Social stigmas and taboos influence how willing individuals are to talk about their practices, and sexual behaviors have a significant fantasy component as well as being shaped by personal histories, remembered or not. Specific challenges further arise when using observation in the study of sexuality. Much actual sexual behavior occurs in private; an observer would change the nature of the encounter. Thus, while social scientists can observe people’s self-presentation ( How do people signal erotic interest?), mate choices ( Do women tend to marry men who are taller than themselves?), or negotiations before sex ( How do potential customers approach sex workers on the street?), many aspects of sexual behavior cannot easily be observed. Sex can potentially be witnessed in sex clubs or bathhouses, although these venues present their own challenges and limitations.

Still, when carried out skillfully and under the right conditions, observation can generate extensive insight into human sexual behavior. Observational methods in sexuality research are discussed here from this perspective—as one tool in a toolkit for understanding sexual behavior. First, I differentiate between ethnography, participant observation, and observation, because these terms are sometimes conflated. Next, the core components of observation—perception and interpretation—are discussed in terms of research undertaken by human observers. Consideration of the researcher as part of the process of knowledge construction thus emerges as central to debates about the use of observational methods. The historical roots of the tension between insider and outsider perspectives are explored next, and I argue that all researchers should practice reflexivity , or a self-conscious awareness of how who we are affects what we see and believe about the world. The later sections of the chapter are concerned with the practical decisions facing researchers using observational methods, as well as a few of the ethical issues associated with research on human subjects: overt versus covert research and a researcher’s sexual participation in the field.Footnote 1

To discuss the interactive and participatory elements of interviewing would require another chapter; however, some of the same concerns about insider/outsider statuses, identity, objectivity, reflexivity, and so on remain salient in those discussions.

1 Ethnography, Observation, and Participant Observation—Clarifying the Terms

Unfortunately, the terms “ethnography,” “observation,” and “participant observation” are often used as if they are interchangeable with each other, or without enough specificity.

Ethnography is systematic and holistic research on a given society or in a specific locale, conducted by an individual or a team. Although ethnographic research is premised on the idea of “fieldwork”—the researcher gains first-hand knowledge by living, working, or studying in a particular place for a period of time, often more than a year—data are usually collected through multiple methods in such projects. (Some researchers use the term “fieldwork” to denote any research that takes place in a “field,” or somewhere other than a laboratory, although such studies are not always ethnographies). Observation and participant observation are methods that can be used in ethnographic research—and probably cannot be avoided, to some extent in field research—although they can be deployed in other types of studies as well. Ethnographers observe behavior as they interact with people at their field sites, but they may also collect data by drawing charts and maps, photographing or videotaping events, examining historical documents, recording physiological measurements, or conducting focus groups, surveys, or interviews. The term ethnography may sometimes also refer to the end product of such multi-method investigation—the analysis and representation of the data with the aim of offering holistic understanding of a setting. To simply claim that one is “doing an ethnography” is thus imprecise and vague .

Observation, in the broadest sense, concerns the collection of visual data, although the other senses also contribute information. Visual data becomes meaningful through processes involving both perception and interpretation—what a researcher sees, or thinks she sees, and the meanings and explanations used to make sense of it.

Observation is used in many types of studies, and can generate quantitative or qualitative data, descriptive narratives, and further research questions. Some information is difficult to attain except through observation. Researchers may want to understand how encounters or negotiations between individuals unfold; observation can provide greater perspective and context than asking questions of individuals. Observations of people’s nonverbal behavior in specific situations can also provide information that may be difficult or impossible for them to vocalize. People do not always know why they do the things they do; sometimes, they do not even know what they are doing. Further, as there is often a discrepancy between what people say they do and what they actually do, observation can shed light on these inconsistencies.

Experimental observation allows researchers to manipulate an environment and to record and analyze participant responses, as in a laboratory. Experimental observations can be conducted on animals or humans. In the 1800s, a researcher used a glass tube shaped like a penis to observe women masturbating to orgasm in a laboratory; his observations supported the belief that women’s orgasms resulted in physiological changes (Bullough 1995). Naturalistic observation refers to observing animals or people in everyday environments or without using experimental interventions. Naturalistic observation opens up many social situations to possible study, especially those occurring in public or semi-public places.

Observations can be structured, where the focus is on counting behaviors or assessing a particular variable, or unstructured, where as much as possible about the scene is recorded. Researchers have conducted structured observations when observing women’s “flirting behaviors” (Moore 1985), customer tipping behavior in strip club patrons (Brewster 2003), or negotiations between potential clients and street prostitutes, for example. Depending on the field site and the questions being asked, researchers may use sampling strategies or alternate the hours of the day or night at which they observe. Unstructured observations do not necessarily have such parameters, but can lead to important and even unanticipated insights. Teela Sanders (2004) conducted 10 months of research with female sex workers in the UK during 2000 and 2001 to study their perception and management of occupational hazards. In addition to interviews, she socialized with sex workers informally and kept field notes. Sanders (2004) recorded instances where she became the “butt of the joke” in a room full of workers; as time went on, she was included in the women’s humor. When she reflected on the pervasiveness of joking, sarcasm, and practical gags in this environment, she came to understand humor as a social and psychological distancing technique: a way for sex workers to manage emotions about clients, create supportive networks, and communicate important information to each other, such as how to defend themselves in an attack (Sanders 2004). Humor was not something she had set out to study, but her observations revealed its importance in this social setting.

Sometimes observation is preferable over other methods for ethical reasons. It would be extremely difficult, even unethical, to study some risky, dangerous, or aggressive behaviors in a laboratory or to provoke individuals in a naturalistic setting. Graham et al. (2014), for example, were interested in aggression during male-female interactions, especially during sexual advances, so they conducted observational research in nightclubs. They watched from different locations in the clubs, and recorded participants’ gender, intoxication level, the intent of interactions that occurred, and the responses of third parties. They observed 258 aggressive instances, 90 % of which involved male initiators and female targets.

The term participant observation is used in a variety of ways, from indicating that a researcher lived among the people being studied to suggesting that the researcher was an on-going participant in that group’s activities or way of life to varying extents. Unfortunately, despite this variation in choices that researchers make about how they conduct observations and interact with others at their field sites, “participant observation” is sometimes used loosely to describe any field study. Philosophically, of course, any observer of human behavior is also at some level a participant simply by being present, regardless of how unobtrusive he attempts to be. Even a sociologist sitting in a Starbucks and logging whether men or women are most likely to order fancy coffees, for example, is still a social actor. Other people may ignore him, engage him, wonder what he is writing, or frown at him for taking up space, but his presence could theoretically alter people’s behavior. He is also a participant in that his observations are filtered through who he is as a person—his research will reflect his perspective rather than an unadulterated “reality.” Methodologically, however, researchers must decide exactly how to engage with individuals at their field sites, and how to handle the specific practical and ethical issues arising out of that decision. Researcher intention does matter, and how much one is willing or able to participate in a social setting can potentially impact one’s data collection and analysis—or not. How much one is willing or able to participate may also have ethical implications. Rather than using “participant observation” as a vague catchall phrase, researchers would be better served by revealing the specifics of the extent and purpose of their participation and their observation.

2 Seeing Is Believing—Or Is It?

Have you noticed how nobody ever looks up? Nobody looks at chimneys, or trees against the sky, or the tops of buildings. Everybody just looks down at the pavement or their shoes. The whole world could pass them by and most people wouldn’t notice.—Julie Andrews Edwards, The Last of the Really Great Whangdoodles

As with all research methods, observational methods have limitations. Human observers are prone to numerous biases, some of which are particularly relevant to observation. Some biases are the result of human cognitive limitations, while others arise from “deep-seated personal, social, economic, or political interests and values” (Poland and Caplan 2004, p. 9). Some researchers are more skilled than others at observation; techniques for improving one’s observational skills exist but are rarely taught in graduate schools. Some critics argue that observational research is ultimately flawed because of the potential multiple sources of error. Other scientists believe that there is never an unbiased or truly objective position from which to conduct research, and instead emphasize that the researcher should be seen as a tool: When we understand how these limitations—cognitive and otherwise—affect our research, we can mitigate their effects and use them to further develop our understanding.

Human observers are necessarily imperfect. We are visually gifted compared to some animals—we laugh when our dog can’t find a treat on the floor in front of his nose—but the range of our vision and our attention to detail is relatively pathetic in comparison to other species. We cannot properly “see” certain things without technological interventions, whether a microscope, MRI machine, or computer simulations. Ducks initially appeared to be sexually monogamous to biologists, who observed the same male/female adult pairs each mating season, later followed by broods of wobbly ducklings. But when researchers began using DNA testing to determine the paternity of the chicks, they realized it was necessary to distinguish between social and sexual monogamy (Birkhead 2002). Because perception is necessarily coupled with the process of interpretation, observations can be tinged with ethnocentrism, racism, sexism, heterosexism, classism, cultural ideologies, and power relations. In the 1600s, early sex researchers reported seeing distinctly male or female sperm with a microscope, or that the sperm of a donkey looked like a miniature donkey (Bullough 1995). Early sex research on masturbation and homosexuality in humans was marked by a tendency to view these practices as unhealthy or even pathological; when the cultural climate changed, the science changed as well—one study at a time.

Selective perception, or the tendency to pay more attention to the things that one expects or wants to see, is an example of a cognitive limitation that can impact social science research. Starting in infancy, humans unconsciously filter out some stimuli while attending to other patterns and details. By allowing us to tune out seemingly extraneous sensory information, this tendency frees us up to perform complex mental tasks—but at the same time, it can make us “blind” to other aspects of our environments. According to experimental psychologist Daniel Simons (Simons and Chabris 1999), what we see when we look around us appears to be a stable and continuous world, but is actually an illusion, dependent on perspective and on interpretation. Sometimes, we see only what we expect to see in a given setting or interaction—and anyone who doubts this should try their hand at his famous “selective attention test:” https://www.youtube.com/watch?v=vJG698U2Mvo). In this experiment, subjects were asked to count how many times a basketball was passed between players. Subjects focused so intently on the task that most failed to notice an adult in a gorilla suit who ambles through the middle of the basketball court, beats his chest, and then leaves. For researchers, selective perception might result in focusing more attention on certain types of individuals than others when recruiting subjects or recording data—basketball players rather than gorillas, for example—or only “seeing” the behaviors that they are interested in.

Numerous techniques exist for mitigating the impact of selective perception on research. Perspective matters, and observers may focus on different things in the same setting. Multiple observers can thus be used, who then compare data. Dates and times of observations can be randomized, or at least spread across the known spectrum of possibilities, to obtain a fuller picture of an environment. A researcher would gain a better understanding of campus dormitory life, for example, if he observed in the hallways as well as the cafeteria, and at 3 a.m. in addition to the daylight hours. As information supporting our own theories, beliefs, or expectations is also more readily remembered, sometimes resulting in confirmation bias (an issue that is not unique to observation), researchers often document their observations and then develop coding systems to analyze their data.Footnote 2 Fieldnotes or observation logs can help with both the recording of detail and recall; audio and video recording can further preserve events and encounters for repeated viewing.

Sometimes, biases arise in observational research because behavior is defined or produced in such a way as to make it more likely to observe. When scientists believed that male rats controlled the mating process, for example, they defined sexual receptivity in female rats as adoption of the lordosis position—arched back, tilted hips—or allowing a male rat to mount. They also designed studies likely to produce this behavior by using small experimental cages where, as one researcher suggested, “a temporarily resistant female” was “deprived of corners in which she can crouch and prevent the male’s mounting response” (Beach 1938, p. 358). When rats were observed in these small, barren cages, researchers confirmed that female rats passively assumed the lordosis position when a male rat was introduced, doing little else to either thwart or court him. But when a pair of researchers observed rats in a more “spacious, semi natural setting,” they found female rats engaged in multiple solicitation behaviors—making the initial approach, then “grooming, crawling over the male’s head, or presenting her hindquarters (in the case of an unresponsive male)” or running away with a “dart-hop gait” or a “stiff-legged run” (Strum and Fedigan 2002, p. 282). The small cages used previously were not conducive to seduction, playing “hard to get,” or even to female avoidance.

When studying humans, researchers are often limited to convenience samples, for both practical and ethical reasons, and have fewer opportunities to manipulate the environment to produce certain behaviors, especially when it comes to sex. We cannot always see what we want to see in naturalistic environments either. People may change their behavior when they think they are being watched, especially if they are engaging in stigmatized practices. Much relevant behavior occurs relatively out-of-sight of researchers. Some enclaves where sexual activity takes place have few barriers to entry, while others erect multiple road-blocks—cover charges, membership requirements, participation rules, and so on. More research has been done on bathhouses and sex clubs than on sex parties in private homes, for example. Sociologist Charles Moser (1998) wrote retrospectively about his visits to BDSM parties, some of which were held in private homes and others in commercial spaces, to describe the rules and expectations of participants. His access to the private parties was premised on the fact that he had been interacting with the community for 25 years. Researchers must be cautious not to generalize if their access to “backstage” or diverse environments was limited.

Some populations are more visible than others, something that must be taken into account in study designs and in analyses of the data that is collected. Anti-prostitution activist and researcher Melissa Farley (Farley et al. 2004) conducted research on sex workers and found that 68 % suffered from Post Traumatic Stress Disorder. Farley’s work has been criticized, however, because her sample was composed of the most visible and accessible sex workers—street or brothel workers rather than those seeking customers online—or the most marginalized populations, such as those seeking community services. Her finding of high rates of PTSD, then, may have been due to poverty, stigmatization due to their visibility, vulnerability to client violence, and police harassment rather than to the act of exchanging sex for money. Farley et al. (2004) administered psychological instruments to participants and did not rely on observation alone to collect data, but her sample suffered from sampling bias, where subjects recruited for a study are not representative of the entire population in question.

Each of us is a particular race, class, gender, and sexuality; these social positions impact our perceptions and interpretations. Gender identification bias, which has been revealed in both animal and human studies, can lead to male and female researchers observing different behaviors or interpreting those behaviors differently. Biologist Sarah Hardy’s (2000) work on primates challenged prevailing beliefs that only male mammals were non-monogamous by arguing that promiscuity might have an adaptive advantage for females. Perhaps because male scientists focused more intently on male primate behaviors, and perhaps also because beliefs in female monogamy and male promiscuity were so culturally salient, male scientists had missed significant female primate mating behaviors and failed to interpret the behaviors they did observe as part of a unique evolutionary strategy.

Community members draw on a system of shared meanings that is not immediately apparent to researchers and can further affect observational data. While researching swinging in the United States, I noted that some academics and journalists reported witnessing “barebacking,” or penetration without a condom, when visiting sex clubs, something usually interpreted as “unsafe sex” and disturbing from a public health perspective. This observation did not fit with my experiences as a participant and observer in the lifestyle community, what I had learned conducting interviews with swingers, or what I had witnessed in sex clubs. Lifestylers, I had come to believe, were very cautious about avoiding STDs and barebackers were forcefully stigmatized. So what could account for the discrepancy? Upon reflection, I realized I was using the more nuanced definitions of sex that I had come to take for granted. Lifestylers referred to unprotected sex with outside partners as barebacking. But, similar to married or committed couples more generally in the US, they did not refer to condomless sex with a spouse (or primary partner) as barebacking, “unsafe,” or even “unprotected.” Further, even though lifestylers allow recreational sex with outside partners, emotional monogamy tends to be highly valued and presenting as a strong couple is important. Condomless sex between primary partners at clubs, events, and parties was often expected as one way to demonstrate commitment and the specialness of that relationship. Swingers’ clubs often post rules requiring condom use during intercourse, but condoms are not expected between committed partners who do not use them at home. Here, then, the same act—penetration without a condom—means completely different things depending on the relationships of the individuals involved. In fact, this distinction affected my interpretations of the behavior as I perceived it. I hadn’t witnessed barebacking, but I had witnessed penetration without a condom between committed partners. Does it matter?

It depends, of course, on the question.

3 Insiders and Outsiders: A Brief History

We don’t know who discovered water, but it wasn’t a fish. —Anonymous

Don’t judge a man until you have walked a mile in his shoes —Native American proverb

The idea that fish would never discover water exemplifies the value of an outsider’s perspective—a detached, or distanced, observer can notice things that are so customary or essential to a practice or way of life that they are taken for granted by participants. On the other hand, an insider’s perspective is also valuable—by “walking a mile in his shoes,” we believe we can at least begin to grasp another’s subjective reality. Some social scientists refer to these perspectives as “emic” and “etic,” contrasting them in a variety of ways: engaged/detached; local/global; particular/universal; insider/outsider, “ground up”/“top down” or “subjective”/“objective.”Footnote 3 Sometimes, emic and etic perspectives become associated with particular research methods , theories, or academic disciplines, with one view privileged as more accurate or essential than the other; other times, researchers attempt to strike a balance between them.

A brief discussion of the history of the tension between insider and outsider perspectives is necessary, as this history influences methodological decisions made today and the philosophical debates surrounding them.

Anthropology developed during an historical period marked by widespread colonialism, when distinctions between “civilized” and “primitive” became loaded with meaning. Supposedly “primitive” cultures were often idealized—as more natural or peaceful, for example (Shangri-La)—or denigrated—as childlike or inferior. Understanding native cultural groups was important to Western European colonizers for multiple reasons, including being able to better control the populations and support theories of racial hierarchy. Participant observation developed hand-in-hand with ethnography in this early anthropology. In the nineteenth and early twentieth centuries, anthropologists wanting to study native cultures would often live in those communities at length. In addition to conducting their “official” academic inquiries—which could include taking physical measurements, charting kin relations, or performing psychological experiments—early fieldworkers thus also necessarily became social actors, though to different extents. They learned native languages, developed relationships with key “informants,” ate the local food, and encountered what anthropologist Bronislaw Malinowski (1922) called “the imponderabilia of everyday life” (pp. 24–25). The process (ideally) of actually participating in daily life was believed to cultivate a richer understanding of people’s worlds than mere observation. “Do as the natives do,” he suggested, in order to understand their point of view. Insights into deep cultural meaning were expected through the process by which “the strange becomes familiar and the familiar becomes strange”; recreating this experience for readers was a goal of writing ethnography. Immersion became imbued with an aura of authenticity, although written reports required a delicate balance between subjective insight and scientific distance. Intellectually, such a move challenged “armchair anthropologists,” who studied other cultures from afar, as well as those who made their observations from “the colonial veranda,” or a safe position of privilege.

Actual levels of participation varied, of course. As Ralph Bolton (2002) asks, “How many anthropologists studying peasant working conditions have actually spent time plowing, sowing, or reaping? More than likely, they sat at the edge of the field and observed” (p. 148). Despite this variation, and despite the fact that turnof-the-century fieldworkers did not blend seamlessly into their new social environments, their experiences of near-total immersion in unfamiliar settings did indeed lead to valuable insights. In Coming of Age in Samoa (1928), for example, Margaret Mead shocked many of her contemporaries not just by suggesting that young Samoan women engaged in casual sex before marriage, but that such behavior could be considered “natural” in another cultural context. Her comparison of Arapesh, Mundugumor, and Tchambuli societies in Sex and Temperament in Three Primitive Societies (1935) again caused a stir by proposing that gender norms and dispositions were culturally constructed rather than inherent to biological sex. Mead’s research has been criticized over the years, but her influence on American society and social movements was immense.

For a researcher in the field, the flip side to obtaining a coveted insider’s perspective was the risk of “going native”—identifying so strongly as a member of the group that one lost objectivity, became unable to relay findings back to the home culture, and possibly even lost interest in returning. Proving that one had maintained proper boundaries took numerous forms in early ethnographies, from the language used in the texts to the patterned ways that research tales were told: first, the researcher appears on the scene as an isolated outsider, then passes a test or challenge to gain the trust and acceptance of the group; in the end, however, the researcher passes another test by abstaining from local practices that would be interpreted as indicative of losing perspective, such as sex, marriage, or religious conversion. In Malinowski’s The Sexual Life of Savages (1929), for example, he argued that sexuality permeated everyday life for the natives; the implication, of course, was that this was not the case for people in more civilized societies. His claims of both accessing a native’s point of view and remaining distanced enough to be objective, however, were later questioned when it was discovered that his personal desires had been relegated to his diary.

From the turn of the century until after World War II, sociologists and criminologists in the US were also becoming more interested in “fieldwork” and in enhancing cross-cultural—or “subcultural”—understanding through observation and participant observation . Sociologists at the University of Chicago began using anthropological methods to study urban environments and “closed” communities with barriers to entry, such as ethnic neighborhoods, gangs, or social clubs. The Chicago School especially privileged the idea of “naturalistic observation,” and viewed the city as a living laboratory in which to study social problems. Robert Park, a Chicago School sociologist, argued that,

the same patient methods which anthropologists… have expended… might be even more fruitfully employed in the investigation of the customs, beliefs, social practices, and general conceptions of life prevalent in Little Italy or on the lower North Side of Chicago, or in recording the more sophisticated folkways of the inhabitants of Greenwich Village.

(Bulmer 1986, p. 92)

Park told his students that although they had been taught that real research required “getting your hands dirty” by “grubbing in the library… accumulating a mass of notes and liberal coating of grime,” “first hand observation” was also needed:

Go and sit in the lounges of the luxury hotels and on the doorsteps of the flophouses; sit on the Gold Coast settees and on the slum shakedowns; sit in the Orchestra Hall and in the Star and Garter Burlesk. In short, gentlemen [sic], go get the seat of your pants dirty in real research. (McKinney 1966, p. 71)

Although Park instructed researchers to study across social classes, the fact is that some people are more visible than others, and some trigger more voyeuristic fantasies. Getting the seat of one’s pants dirty, it seems, was supposedly more likely in a back alley than at the Orchestra Hall. Similar to anthropologists who set off for remote jungle outposts, many sociologists turned their eyes toward the “exotic others” of their city milieus—hobos, criminals, juvenile delinquents, and the disenfranchised.

Although some researchers believed in becoming a “fly on the wall” in their chosen field sites—that is, to refrain from disrupting or intervening in social interactions in ways that might bias their analysis—others sought access to the inner lives of their informants and to understand the subjective meanings of their actions through varying degrees of participation. Symbolic interaction theory, which had been gaining favor in those years, proposes that human reality is constructed through our interactions with others. Researchers, then, could use their own experiences as data, as they learned to “think and feel” like the people they were studying—sometimes referred to as “sympathetic introspection” (Cooley 1909, p. 7). Once again, immersion experiences generated important insights. By respecting the meanings generated in various communities, researchers tried (and sometimes succeeded) at humanizing individuals who were often seen only as social problems. The focus on deviance as a social process involving labeling, stigma, and power relations unsettled the view that “deviants” were born psychologically damaged or inferior. On the topic of sexuality, sociologists produced work on “gay ghettos” and street prostitution , for example, exploring the ways that individuals acquired and managed stigmatized identities.

Today, researchers no longer usually face the possibility of complete isolation at a field site as early anthropologists did. Few tribal groups, if any, lack contact with outsiders. Even places that are geographically remote are economically, politically, and technologically linked to a global network. Many researchers now study close to “home” for practical, ethical, or other reasons. In such a world, it is more difficult to know what “total immersion” would look like and field sites are not necessarily distinguishable from one’s everyday social world. And although social scientists continue to study in social enclaves with barriers to entry, they can no longer claim to access a single privileged or authentic “insider perspective.” Community members may critique research findings or become researchers themselves. Still, questions that arose decades ago about the potential for, and usefulness of, developing insider versus outsider perspectives , and about whether researchers should engage with the individuals and social settings they are studying or remain distanced observers, continue to animate discussions of observational methods and to influence the decisions with which contemporary researchers are faced.

4 Reflexivity as Practice

The idea of the researcher as a tool has occupied a great deal of intellectual thought and debate. Humans are visual mammals, yet we are also prone to numerous biases, misperceptions, and misinterpretations. Do we trust our own eyes? Do we trust a particular researcher’s observations or interpretations? Why or why not? How do we know if one is gaining objective knowledge about the world or merely offering a subjective description? Although these questions are clearly relevant to all types of research, they are particularly salient when it comes to studying the world of human meaning.

4.1 Degrees of Engagement

Because the tension between insider and outsider perspectives is so fundamental to observational and participatory methods, much intellectual effort has gone into trying to characterize researchers’ engagement with the communities they study. Gold’s (1958) typology of researcher roles included 4 modes of data gathering: complete participant, participant as observer, observer as participant, and complete observer. Sociologist James Spradley (1980) also developed a continuum based on researcher involvement, ranging from non-participatory (no contact), passive (bystander role), moderate (a balance between insider and outsider roles), active (“going native”), and complete (the researcher is already a member of the group). Adler and Adler (1987) focused on a researcher’s belongingness—not just participation—in groups being studied. They thus distinguished between researchers with peripheral membership (just observing), active membership (participation in at least some activities) and full membership (full participation). Sociologist Loic Wacquant (2011) used the term “observant participation” to suggest that a researcher can prepare to “go native” by equipping himself both with the tools and training of a social scientist and also by learning the bodily dispositions and practices of the community being studied. Wacquant trained as a boxer for his ethnography, Body and Soul: Notebooks of an Apprentice Boxer (2000).

None of these characterizations are appropriate for describing every type of group or fieldwork situation, however. Further, we are still left with the question of how precisely one is to distinguish between statuses when the distinction between insider and outsider is not straightforward. My own research on the male customers of strip clubs might be considered “active” or “full participation”—I was, after all, working as a stripper. I learned the bodily dispositions of a stripper (see Frank 2002b, 2005) and was often primarily viewed by others as a stripper, not a researcher (Frank 2002a). But as I was studying the customers instead of the dancers, wasn’t I in some respects also observing as a participant? What are the criteria that should be used in determining whether a researcher should be called a participant observer, an observant participant, a non-participatory observer, or something else? Is it how much time is spent in a particular field site? Whether there were return visits? Is it whether the researcher engaged in all of the activities as the other people present? And how does identity come into play? That is, if a researcher identifies as a BDSM practitioner, can she be termed a “complete” participant observer if she refrains from playing at her field sites for ethical reasons? What if she only takes part in a BDSM scene because she is conducting research and is curious about how it feels? How would we categorize a gay man who studies heterosexually identified men-who-have-sex-with-men at sex clubs?

Research is also a dynamic process—our methods, questions, and theories may change based on our experiences in the field. Joseph Styles (1979) set out to study men’s sexual encounters in bathhouses as “a nonparticipating insider”; however, after being groped in the crowded corridors to the point of losing his towel and realizing that his attempts at conversation were being interpreted as sexual interest, he thought, “to hell with it” and became a participant (as cited in Goode 1999, p. 305). In doing so, he gained a deeper understanding of the subtle sexual negotiations taking place inside the venues. Researchers need to negotiate complicated relationships with individuals at their field sites; participation in everyday activities may increase or decrease as friendships develop and deepen. As a researcher gains knowledge about a field site and comfort with previously foreign practices, his observations and interpretations may change. A researcher unfamiliar with group sex, for example, might at first be overwhelmed by the nudity in a sex club and focus his initial observations on the tangle of bodies on a mattress; later, though, his attention may turn to how space is demarcated for socializing or on the types of conversations unfolding along the sidelines.

4.2 Practicing Reflexivity

The focus on whether or how much a researcher participates, I believe, has overwhelmed consideration of other aspects of observational methods. More important than trying to characterize the researcher’s role through static terminology, I believe, is the practice of reflexivity.

Reflexivity “involves an awareness that the researcher and the object of study affect each other mutually and continually in the research process” (Alvesson and Skoldberg 2000 as cited in Haynes 2012, p. 73). More than “a simple reflection on the research process and its outcomes,” reflexivity is multi-layered contemplation that includes “considering the complex relationships between the production of knowledge (epistemology), the processes of knowledge production (methodology), and the involvement and impact of the knowledge producer or researcher (ontology)” (Haynes 2012, p. 73). Actual reflexive practices thus vary depending on a researcher’s assumptions about the nature of reality, or what there is to know about the world, and how we best learn about it. Some researchers, for example, place critical emphasis on reflection about the data gathering process, while others focus more on the acts of recording observations or writing up results. Ideally, however, reflexivity can be practiced at each stage of the research process and for every type of observational method. What C. Wright Mills (1959) called the “sociological imagination” is an ability to grasp the interplay between individual experience and social structures in one’s analysis. Reflexivity , then, requires social scientists to analyze themselves as well as others—asking, how are my questions, methods, values, and goals in this research influenced by the social structures around me? The answers to these questions are not always crystal clear, especially at the beginning of a project, which makes the development of reflexivity—or “reflexivities”—more like practicing piano than climbing a mountain.

Researchers need to develop an awareness of how broader power relations impact the very definition of “research,” as well as interactions with research subjects and their interpretations of those interactions. Social, economic, political, cultural and historical contexts influence research questions and processes. Consider what some researchers have called the “politics of visibility.” Much as colonial history and persistent global inequities situated some people and places as the likely objects of anthropological study, social marginalization makes some groups and individuals more visible and accessible to social scientists in the US. We have far more studies of sex workers who work the streets or in brothels than of those who use the Internet to find clients. More research is conducted on poor drug users in crack dens than on Fortune 500 CEOs who use cocaine.

Just as privilege allows certain groups to escape the prying eyes of social scientists, who we are influences what we choose to study, the questions we ask, and how we try to answer them. Our individual characteristics—race, class, sexuality, gender, and so on—affect how we perceive and interact with others in the field—and they with us. These characteristics can limit our access to certain spaces, which roles we can adopt, and how we interpret our experiences. Understanding our positionality, then, is an essential part of reflexivity.

Male researchers studying in strip clubs with female dancers, for example, have tended to adopt distanced roles such as “non-tipping customer” with little attention paid to positionality (Brewster 2003; Erickson and Tewksbury 2000). Female researchers, on the other hand, have often taken a more reflexive stance, perhaps because of their conspicuous presence or because of an assumed or actual involvement as dancers. Female researchers have also focused more on the complexity of gendered power relations in stripping (Barton 2001; Egan and Frank 2005; Frank 2007). But what does it mean when a male researcher positions himself in the audience to view stage performances but refrains from tipping or purchasing dances, as several have done? Is this approach more ethical, respectful, or likely to elicit a deeper understanding of the transactions occurring than interacting like a typical customer would be? Is either approach more risky, personally or professionally? Male researchers may be stigmatized even more than female researchers when they choose to study the sex industry, seen as lecherous by their peers (Barton 2001). Such fears of stigma influence methodological and analytical decisions, such as a desire to limit certain kinds of involvement or to focus one’s analysis on the “safer” aspects of the encounters. The experience of doing research can be fraught with a sense of danger, vulnerability, risk, and transgression; sometimes, explicitly acknowledging this complexity can lead to a more nuanced analysis.

As broader social contexts shape each of us and influence our interactions with others, researchers should thus be reflexive about how their own characteristics and privileges impact the research process, affecting everything from gaining access to a community to how much we participate to how we interpret our results. Some characteristics, such as our gender, race, ethnicity, age, height, and so on, directly affect how others respond to us and how we see the world. Other characteristics, such as social class or sexuality, may have a more indirect influence on our interactions through how we look or speak. Examining our own beliefs, upbringing, relationships, and personal histories in these shifting contextual fields presents additional opportunities for understanding the meanings we make in the field.

On-going reflexivity with regard to our emotional responses during our research can also guide theoretical, methodological and analytical decisions. When I began studying the male customers of strip clubs, for example, I focused strictly on the men’s relationships with the dancers. In writing about those interactions, I tried to maintain an awareness of how my observations and experiences were shaped by my positionality as a white American woman from a working class background, a graduate student well versed in feminist theory and an “out” exotic dancer. Consciously traversing a complex web of privileges and stigmas was part of my study design. I wanted to move beyond the deviance framework for understanding stripping and reverse the usual mode of inquiry—from “Why do some women do this?” to “Why do some men want to pay for this?” Yet my roles as “researcher” and “stripper” were not the only ones that shaped my experiences. As I continued both working in the clubs and conducting interviews off-site, I found that customers repeatedly inquired about my wedding ring, asking, “How does your husband feel about your dancing?” The question was uncomfortable for me because I had worried about his feelings prior to beginning the project. The query also made me wonder how my interactions with customers in the clubs affected the other women in their lives. How would these women feel about me? How would I feel about my marriage if I found out that my husband visited strip clubs? As George Devereux (1967) points out, anthropological data can arouse anxiety in researchers and this anxiety can influence our observations and analysis. Eventually, I reversed the question, asking, “How does your wife or partner feel about your visits to strip clubs?” I focused more on how these men’s outside relationships influenced the meanings of their visits to the clubs as my own problematic status and experience as a married person in this environment became evident to me. Reflecting on my emotional responses influenced my decision to explore customer beliefs about monogamy, and affected my interpretations of our exchanges in the clubs and during the interviews.

How much of this process of critical reflection is shared with readers varies. Social scientists traditionally included discussions of how they gained acceptance to a community, disclosing mistakes and breakthroughs as a way to authenticate their observations and assure readers that objectivity was maintained—the researcher got close enough to understand people’s behavior, but not so close that perspective was lost. After these initial discussions, however, a researcher tended to maintain an authoritative presence, regardless of how much he had privately thought about his role in the production of knowledge. More recent reflexive writers tend to be forthcoming about their positionalities and their political and personal investments in their field sites, situate their research “story” within a wider social context, and consider the ethics of the research at each stage of the process, sometimes to a distracting extent. The detached observer is thus one extreme; Van Maanen’s (1988) “confessional ethnography,” where the researcher takes center stage in the narrative, is another. Critics view reflexive ethnography as “a largely egocentric, asocial activity” (Webster 2008, p. 67)—philosophical naval-gazing or confessional writing that tells us more about the author than the world. Many of us do not want ethnography to turn into autobiography, though precisely how much one should disclose about one’s identity, engagement in a field site, political ties, and so on, is still debated.

A key to resolving the debate lies in revisiting the reasons that we turned to reflexivity in the first place. The underlying issue is a philosophical dilemma about what it means to study others, especially when using observational methods. What is revealed to readers, then, should be information that helps them evaluate the researcher’s conclusions. We can thus consider the researcher’s positionality as a source of information rather than bias, and the process of reflexivity—regardless of how much is directly shared with readers—as a technique for mitigating some of the limitations inherent to observation.

In her research on women’s bathhouse events, sociologist Corie Hammers (2009) suggests that her identity as a lesbian/queer woman allowed for easier access and made her seem trustworthy to the organizers. Hammers decided against participating at the events, however, and thus thought carefully about how to remain unobtrusive. She carried a tape recorder and tablet, but kept them concealed except during interviews. She chose to wear a long sleeve shirt and jeans to indicate “unavailability” and “seriousness,” but also recognized that her attire set her apart from the crowd (Hammers 2009, p. 317). During the events, Hammers disclosed her research role to individuals who inquired or seemed interested in her erotically. Because several hundred women were in attendance, though, most participants were unaware of her objectives. Overall, she did not believe that her presence interfered with patron’s sexual activities, and her observations suggest interesting differences between men’s and women’s bathhouse cultures: men’s sex was often with strangers, while women’s bathhouse sex was more personalized; in contrast to men’s silence, women were often loud and celebratory; and women’s events had a strong emphasis on sociality. She raises questions, however, about whether her physical disengagement affected respondents’ honesty in reporting their motivations. Few women admitted visiting the bathhouse purely for sex in her interviews, for example. But were these lesbians/queer women potentially less willing to admit an interest in casual sex due to anxiety or vulnerability in the presence of a non-participant (Hammers 2009)? By analyzing how her personal characteristics and methodological decisions may have affected her interactions, Hammers (2009)? helps readers contextualize and evaluate her observations and interpretations.

5 Covert Versus Overt Research

Observation may be covert, where the researcher does not divulge her purpose and may even mask it, or overt, where people are aware that they are being studied and may or may not know why.

5.1 Covert Observation

Covert research is sometimes considered necessary in settings where people’s behavior would change significantly in the presence of a researcher. One of the most famous sociological studies of sexual behavior using covert observational methods is Tearoom Trade: Impersonal Sex in Public Places by Laud Humphreys (1975). When Humphreys wanted to study men who utilized “tearooms,” or public restrooms known for same-sex activity, he found that the layout of each facility and the reactions of participants to his presence affected his ability to only observe. He initially pretended to be a straight man entering the restroom, but because the men worried about being arrested or observed accidentally, a lookout, or “watchqueen,” would alert the others to cease their activity when someone was approaching. Eventually, Humphreys (1975) adopted this role of lookout, as it allowed him to observe without being expected to join in. Alternately keeping watch and retreating to his car to write notes, Humphreys recorded the types of sexual acts engaged in, and examined the strategies used and roles adopted by participants in sexual negotiations. Because tearoom participants were so concerned about not being observed, there was little chance of an outsider accidentally witnessing any sexual activity. Police stings, Humphreys argued, were thus not really combatting a significant public problem, but being used as a form of harassment and stigmatization.

In a more contentious part of the study, Humphreys (1975) added an element of deception. Because he believed that the men would decline an interview if he approached them in the restroom, he recorded the license plate numbers of some tearoom participants. He then tracked them down at their homes and interviewed them about their attitudes towards homosexuality under the pretenses of conducting a public health survey. He found that 54 % of the men were actually married and living relatively conventional lives; he also argued that many put on a “breastplate of righteousness” (p. 135), publicly condemning the same behavior they privately engaged in. Humphrey’s research was fascinating as it was uniquely able to address the disjunction between what people say and what they do. Still, his decision to conduct the interviews under false pretenses and to collect personal information that might have put his research subjects at risk sparked controversy (Humphreys 1975).

An account of how men negotiate anonymous sex in bathhouses or restrooms is concerned with patterns of behavior, does not require the identification of any particular individual, and should ideally pose no risk to the men who were observed. Places and people can be given pseudonyms and identifying details can be changed in researcher notes and publications. Richard Tewksbury (2002), also a sociologist, presented himself as a “potential participant” in covert research on two gay male bathhouses. Spending several hours at each location, he “circulated with and among patrons,” carefully observing “their activities, movements, interactions and the use of the physical features of the environment” (Tewksbury 2002, p. 85). Periodically, he retreated to private areas to write notes. But when Humphreys (1975) decided to examine the links between the men’s sexual practices, their social identities, and their political beliefs, however, he needed to collect more detailed information on each participant. Had the men’s activities somehow been inadvertently revealed to their families, neighbors, or employers, the results could have been devastating.

Some critics dislike all covert research because the individuals being observed have not had the opportunity to give informed consent. On the other hand, we are observed and “studied” in many everyday situations without having been asked for our consent. Even online, information is collected about us incessantly, whether we are posting on Facebook or shopping on Amazon.com. If the observations are conducted in a public place, individuals have no risk of being identified, and the possibility of causing harm is negligible, requiring researchers to obtain informed consent may be counterproductive and unnecessary.

The potential risk involved, rather than the simple fact of covert observation—or even deception—is what should weigh most heavily in ethical assessments. Potential harms vary, of course. Subjects are occasionally exposed to potential physical harm, as in some medical or pharmaceutical research, or psychological harm, as in Stanley Milgram’s (1963) famous experiments in obedience to authority, which caused distress in participants. Institutional Review Boards (IRBs) were created to mitigate these harms by carefully reviewing research involving human subjects. In IRB proposals, researchers detail the aims of their study, the methods used to recruit subjects, how data is collected and stored, potential risks to participants, and how findings will be used and disseminated.

Academic researchers are also bound by disciplinary ethical codes that apply even in covert situations. Social scientists undertaking covert research generally take care to maintain the anonymity of those involved—unlike when journalists undertake an exposé or spectators post videos of public behavior on YouTube. Sociologist Eric Anderson conducted covert research on the website AshleyMadison.com, a “married dating service for discreet encounters,” by monitoring conversations between potential partners (Luscombe 2014). Individuals were often unaware that their online conversations were being viewed and analyzed. Anderson was later accused in the press of spying on members, but the terms and conditions provided at sign up had indicated that user communications might be monitored—a good reason to read the fine print! Further, anonymity of members was maintained (Luscombe 2014).

Social scientists are further expected to take care not to cause physical or emotional harm and to adhere to scientific standards in study design, data collection, and publication. This is not the case, however, for all writers. In a problematic endeavor, for example, Charles Orlando (2014), a “relationship expert” and popular author, went “undercover” and dated women from Ashley Madison to “find out why women cheat.” With seemingly no ethical qualms except a worry about actually cheating on his wife, he set up three fake profiles “to see which would resonate fastest” with women on the site, then started chatting with and dating women who responded. He admits lying to the women about his intentions and personal circumstances; his published descriptions of his encounters include evidence of the women’s psychological distress at his eventual rejections (Orlando 2014). Orlando’s deceptive “experiment” would not have passed institutional review.

Unfortunately, more and more researchers are finding their studies unduly scrutinized when the research topic is sex. Many types of naturalistic observation are exempt from formal institutional review; studies involving sexual behavior, however, may be deemed “sensitive” and subjected to additional protocols. Some institutional review boards assume that asking any questions about a person’s sexuality can potentially cause psychological distress, although this concern may reflect the individuals reviewing the research more than the actual risks. I have interviewed people about their sex lives for years, providing consent forms warning of possible discomfort with my questions and offering therapists’ contact information. Overwhelmingly, though, interviewees describe the experience in positive terms, as more “like therapy” than interrogation.

Regardless of whether one’s institution requires official review for a particular study, researchers should think carefully about the impact of the research on participants. When people engage in stigmatized or illegal behaviors, there is a risk of exposure to peers or authorities. Exposure could result in legal or social penalties, depending on the context, or cause personal distress. Even when anonymity is maintained for individuals, researchers may worry about disseminating findings that reinforce stereotypes or are damaging to a group as a whole. Researchers working on highly politicized issues such as teen sexuality or “gay parenting,” for example, may find that their work attracts more negative publicity to already stigmatized groups or that their findings are coopted by the media or special interest groups.

5.2 Overt Observation

In overt studies, researchers are generally forthcoming about their purposes and participants may be asked to give informed consent, although how this is done may vary. Self-identified lesbian researchers Catherine Nash and Allison Bain (2006) observed women’s behavior at bathhouse events. On the evenings of the events, the researchers presented themselves as both voyeurs and potential participants to the other patrons, although they avoided sexual activity. Patrons were not asked to sign consent forms; some knew the women were researchers and others did not. The organizers of the events, however, were aware of the women’s ongoing research, and Bain and Nash (2006) had also been conducting interviews with patrons and gathering survey data, practices for which they sought informed consent.

Sometimes, the nature of one’s research or the particulars of a field site require mixed strategies. In my research on strip club customers, my initial interactions with potential interviewees necessarily took place in my role as a stripper. I was an employee of the clubs, and thus club managers were invested in my performances as a dancer, not as an anthropologist. There were far too many men in the clubs each night for me to approach every one, much less describe my study to him. Further, my focus was on the regular customers, the men who used the clubs relatively frequently and considered it an important part of their erotic repertoire. Working as a dancer enabled me to identify the regulars. If conversation allowed, I discussed my project with them and asked for an interview off-site at another time. The constraints of the environment, then, meant that I observed hundreds of men each week, some who knew I was an anthropologist and some who did not even care to know my stage name, much less my real name. This design made sense to me practically and ethically. Informed consent was relatively meaningless in the public workplace setting of the strip club because I was not collecting data that could identify customers and was using pseudonyms for the clubs. Informed consent became very important during in the in-depth interviews, however. The interviewees provided substantial amounts of personal information, unlike other customers of the clubs, and participated in an additional transaction—the interview. Thus, interviewees were provided with consent forms stating the purpose of the research, and knew that they could refuse to answer questions or stop the interview at any time.

Technological innovations pose new questions about the meaning of observation and participation, and raise new ethical concerns (Binik et al. 1999). What exactly does informed consent mean, or require, in a virtual world? Are online forums “public” or “private”? How can researchers respect and protect those whom they virtually “observe” in online interactions? Researchers may participate to varying extents in online worlds. Researchers working in virtual communities should learn enough about computer security to be able to protect the identities of individuals they interact with online; care should be taken using screen names and direct quotes that can be found easily through search engines, as personal information could be revealed inadvertently.

Each discipline in the social sciences has a code of ethics that should be respected whether research is covert, overt, or a mix of the two.

6 Researcher Sexuality and Sexual Interactions

Some researchers believe that sexual involvement with research subjects—and sometimes even eroticized interactions—should always be avoided for both practical and ethical reasons. Still, it happens. Anthropologists return from the field with a spouse or children. Researchers have flings and affairs. Researchers who admit crossing sexual boundaries, though, can face censure or stigmatization, and transgressions are often relegated to informal conversation or gossip. Malinowski (1967) wrote about his sexual desires in his diary, never expecting it to be published posthumously. Some ethnographers allude to sexual experiences in their writing, although fewer are completely open about them. Erich Goode (2002) argues that gay men and women have been more likely to write about sexual encounters in the field than heterosexuals, and that anthropologists have been more likely to disclose than sociologists.Footnote 4 Women risk sexual harassment and violence during fieldwork, which might be one reason why female researchers have addressed sex more frequently and directly in their academic work. Further, involvements between female researchers and male informants or between gay men may be “less likely to conjure up an image of traditional exploitation,” which may make it easier for those individuals to write about their experiences without professional repercussions (Goode 2002, p. 502). That does not mean that male researchers have not had sexual experiences with female informants, of course—just that such interactions are potentially judged more harshly and thus do not become a part of the published record. Maintaining an artificially distanced academic persona, however, does not rectify existing power differentials; in fact, it may reproduce them.

One justification for abstinence has been to prevent exploitation and ensure that subjects are not coerced into either sexual activity or participation in a project. But while the potential for an abuse of power should always be considered, research carried out in naturalistic settings involves complex social relations. As researchers can be exploitative or dishonest in any relationship with informants, and vice versa, why should sexual relationships be singled out as somehow more problematic? Researchers may also be more or less privileged than their informants, or members of the communities they are studying. Unlike in the early days of field research, people who are written about can now often comment on or publicly reject a scholar’s results. In 1969, Humphreys could claim that an observer in a tearoom is not yet “suspected as being a social scientist,” but this is not necessarily the case nowadays. BDSM communities, for example, have been extensively studied in the past few decades and members now often engage proactively with researchers or even conduct research themselves. Organizations like the Community-Academic Consortium for Research on Alternative Sexualities (CARAS) promote communication between activists, participants, and researchers in alternative sexual communities, recognizing that these roles are not mutually exclusive. Whether or not researchers identify as members of the communities that they study, they must always think critically about how they will balance their various roles and interests, handle issues of consent and disclosure, and manage close relationships at their field sites.

As with other forms of researcher participation, concerns have also been raised that sexual involvement will distort a researcher’s judgment and ability to present her findings objectively. Goode (1999) suggests that researchers who are intimately involved during fieldwork are not inclined to romanticize the people they are studying, as being acquainted with the mundane details of their lives actually prevents unabashed advocacy. Still, he argues, sex with informants does affect what a researcher can write about—to disclose some details would be inappropriate, harmful, or even just embarrassing to the individuals involved (Goode 1999). Whether or not—and which—details need to be shared with readers in the first place, of course, is a question that should be carefully considered. Anthropologist Kate Altork (1995), who writes eloquently about her erotic experiences while researching firefighters, warns that the point of reflecting on such experiences is not “to encourage sensationalistic, National Enquirer-type confessionals” about one’s sex life. Instead, researchers should simply remain open to discussing the possible ways that sex “changed, enhanced, or detracted from what we felt, witnessed, and interpreted in the field” (Altork 1995, p. 121).

For researchers studying in explicitly sexualized environments, and especially those using observational methods with any element of participation, the ethical issues at stake can be intensified and potential stigmas multiply. Sexual interaction can occasionally enhance rapport or speed acceptance into a community. Some critics view this claim as a self-serving attempt at justification after the fact (Bryant 1999), but a more tempered view would be that it depends on local social norms, what and where one is studying, and how one’s sexual relationships are handled. In his research on gay leathermen in the Netherlands during the 1990s, Maurice van Lieshout (1997) used an “opportunistic research strategy,” suggesting that sociologists might take advantage of familiar social situations. As he had already participated in the Dutch gay leather scene, he gained rapid entry into the setting he wished to study and easily developed rapport with participants (van Lieshout 1997). In a study of sex and romance among members of the National Association to Aid Fat Americans (NAAFA), Goode (2002) recalls his legitimacy being called into question by a male NAAFA member. “I really don’t think you belong here if you are not attracted to fat women” (p. 508). Answering the man’s questions about his erotics became a test of his loyalty and good will towards the community; his affirmations that he really had desire for fat women were a prerequisite for gaining access. But sexual relationships can also negatively impact access if researchers make mistakes. Goode also admits, for example, that he dated too many of the women in NAAFA too quickly, causing irreparable damage to his reputation in the community. Still, he points out that such issues are not unique to sexual relationships, but to any relationships generating strong emotion.

Anthropologist Ralph Bolton (2002) suggests that participant observation —in the fullest sense of the term—has a place in sex research because it allows access to private space and encourages the development of intuitive understanding. “Unless the observer has had wide-ranging sexual experience,” he writes, “it is unlikely that he or she can even know what questions to ask or imagine all of the permutations and complexities of sexual events” (Bolton 2002, p. 148). Anthropologists Charles and Rebecca Palson (1972), a married couple, were involved in swinging before they decided to formally study it. The Palsons (1972) claimed that, “most of our important insights into the nature of swinging could only have been found by actually experiencing some of the same things that our informants did” (p. 29). English professor Tim Dean (2009) admits to participating in unprotected sex in his book on barebacking. Dean does not consider his work to be ethnographic; he is not a social scientist and didn’t conduct formal interviews. He did, however, listen and observe. He also had sex. Barebacking, he claims, is an “underground sexual subculture” that “tends to resist conventional research methods .” After “uninhibited, multipartner sex,” he writes, “men tend to speak more freely.” Being in an “overtly sexual space” such as the back room of a gay bar helped “dissolve some of the barriers and pretensions that constrain verbal exchanges elsewhere” (Dean 2009, pp. 29–34). Sociologist Russell Westhaver (2003), who writes on gay male circuit parties, was a participant at events and worked for a company involved in their production. He situates himself as an insider who engaged in “sensuous scholarship,” which he explains as ethnography “grounded in a commitment to seeing, hearing, feeling, smelling, and tasting the body through poetic processes of transcribing, revisiting, and elaborating bodily experiences and memories as fieldnotes” (Westhaver 2003, p. 21). Participation was a crucial element in his understanding of the emotional power of the events.

Erotic entanglements may lead to a deeper understanding of social networks in some communities. Bolton found that the line between his personal and professional lives blurred while he was studying gay bathhouses in Brussels. “In gay culture ,” he writes, “sex is where the action is” (Bolton 1995, p. 142). His relationships with friends and lovers provided him with access to social events and experiences that would have been unlikely had he remained distant:

I became a player in the scene, reciprocating by introducing my tricks, friends, and lovers to others in my network…. By experiencing them, I came to learn of blow jobs from bartenders when the door was locked at closing time, of jacking off in cruising spots in a park near the Grand Place in partially public view, of sexual encounters in alleyways between someone headed home from the bars and someone on his way to work at dawn, of sexual action in the dunes along the coasts and on the piers in Ostende and in the backrooms of discos and in the bathrooms of ordinary bars. (Bolton 1995, p. 148)

Participation also informed his research in bathhouses and saunas. Although some sites where sex took place were relatively public, such as the steam room and the orgy room, he found that nonparticipants altered the flow of interaction and that the dim lighting presented difficulties with observation. And while interviewing could have been done in nonsexual areas of the sauna such as the bar area or television lounge, most conversation took the form of “post-coital sharing” (Bolton 1995, p. 150). These conversations provided valuable information. He did not ask sexual partners to sign consent forms; some did not know he was conducting research on sex and AIDS. Still, Bolton “never engaged in sex for the purpose of collecting data,” never coerced anyone into having sex with him, and protected people’s confidentiality. He also stresses that his partners did not suffer physical or psychological harm from the encounters (beyond the emotional pain of relationships ending on their own). His research objectives were “subordinate” to his participation as a member: “I never engaged in any behavior that I would not have engaged in had my research objectives been different” (Bolton 1995, p. 151). His identity as a gay man did not stop him from reporting negative findings about men’s risky sexual encounters.

Sometimes, abstaining from participation can disrupt one’s investigation and relationships. During fieldwork in Mozambique among marginalized young men known as moluwenes, anthropologist Christian Groes-Green (2010) found that because of differences in gender, race, and status, his informants perceived him as “morally righteous” and were wary of discussing their sexual practices with him. Groes-Green (2010) slowly earned their trust by drinking with them, partying, “being wild,” and “celebrating spontaneity, naughtiness, and excess.” But when he turned down a local woman’s offer to participate in an orgy one evening, he suddenly reverted back to being an outsider, even a “traitor,” and realized his access to the community was at stake in such decisions. His awareness of his privileged position in relation to the community he was studying often led him to withdraw from lust-provoking situations and “create social boundaries and physical distance.” Yet, the social milieu also required managing his ambivalence. He continued to experience anxiety and guilt when confronted with scenes of unsafe sex, feeling “complicit” in their risky activity because he was unable to intervene without losing his ability to observe. Yet, Groes-Green (2012) grasped that “delimited involvement”—by which he meant being in close proximity without including “direct sexual or carnal merging”—was critical both to his access to the community and to his aim of understanding why moluwenes made the choices they did with regard to sexual behavior.

Researchers Bain and Nash (2006) defended their decision not to participate at the women’s bathhouse events they studied on the grounds that one researcher was monogamous and that their “feminist ethics” prohibited them from doing so. Not surprisingly though, their decision to wear street clothes and position themselves on the outskirts of the activity meant they felt “awkward” when play began. They worried about being perceived as inappropriately voyeuristic, inhibited, or judgmental by other attendees. Observers, after all, can themselves be observed. The organizers of the events, whom Bain and Nash (2006) interviewed prior to attending the bathhouse events, made the researchers feel they were not being “honest” in their research if they did not participate. This was not just because their decision was made ahead of time, but because they also were not “using the space in the ways [the organizers] had envisioned” (Bain and Nash 2006, pp. 99–106). When Nash and Bain broke etiquette in such a relatively small and tight-knit community, their fantasy of maintaining a “fly on the wall” researcher position was smashed by the “elephant in the room.”

If anthropological and feminist ethics suggest attention to power differentials, what are the ethics of academic voyeurism, especially if it causes discomfort or confusion for others? When researchers decide ahead of time what they are willing to experience, might they become like tourists, disrespectful of local customs and oblivious to their own social impact? Do prior intentions not to engage sexually in particular settings protect researchers against the vulnerability that participants expect and experience, and thus inhibit a researcher’s ability to understand a field site? Bolton (1995) suggests that the social and emotional risks to the researcher, at times, can be equivalent to or greater than the risks to any of the participants—the researcher, after all, may be far from her comfort zone. The researcher role, Bain and Nash (2006) admit, served as a “cover,” providing psychological safety by offering little opportunity to “dwell on, or even discuss” insecurities about their attractiveness to other women (p. 103). Hammers (2009) wonders:

Had I participated more directly, would I have had a deeper understanding when it comes to sexual negotiations and the exposure, vulnerability, and empowering appeal that spaces such as these induce? Having set myself apart from the scene, did I, like Styles, miss out on the subtleties, cues, and complexities when it came to body language and negotiations of sex?(p. 317)

Likewise, Groes-Green (2012) acknowledges that despite his ambivalences about risky sex, his understanding of his informants grew when he personally experienced the “bodily momentary intensities that drive youngsters to play with death and danger, ecstasy and annihilation, orgies and frenzy” (p. 56).

To their credit, these researchers raise these questions themselves in their published work. Researchers should never be required to participate in activities that violate their personal, ethical, or emotional commitments in the name of science. The point, however, is that neither participation nor abstention from sexual activity is inherently unethical or problematic. Rather, decisions about sexual participation are made by particular individuals in specific contexts and should be evaluated as such. Every research method has strengths and limitations. Survey research may suffer from low response rates or from a community’s dislike of being studied by outsiders. When limiting themselves to observation, researchers may not have access to back rooms, semiprivate exchanges, or less visible individuals. Participant-observers enjoy greater access but may feel conflicted over disseminating findings that portray a community negatively or find themselves stigmatized in the academic community. All researchers should reflect on the appropriateness of their methods to their questions and on power dynamics in the field, not just when contemplating sexual involvement with a particular informant but at each stage of the process, from the choice of where to study to deciding what questions should be asked and of whom. Researchers should also follow the code of ethics for their discipline. Some researchers suggest that anthropologists have somewhat more flexibility with regard to sexual encounters in the field than psychologists (Montes Penha et al. 2010), although anthropologists would also be expected to respect local norms and practice a high degree of reflexivity.

Goode (2002) asks several provocative questions that can guide ethical reflection about sexual involvement:

One: Can sex with informants harm them—that is, over and above what ordinary, nonsexual interaction does? Two: Does sex with informants alter what the researcher writes about? Three: Is sex with informants categorically unethical? Four: Does sex with informants gain access to information and insight that is otherwise inaccessible? (p. 527)

And to these questions I would add a fifth: What are my personal erotic investments in my interactions with informants, and how do they affect my research? As in all research, the process of reflexivity is key to developing a deep understanding of what we see and how we interpret it; the answers to why we found ourselves there in the first place may not be immediately clear. Outsiders sometimes want to study strange or “deviant” sexual practices with little reflection on their own sexuality; researchers may be seeking many things in addition to knowledge that affect their choice of questions and field sites—healing, affirmation, excitement, adventure, and so on.

Some insights generated by emotional or physical intimacy can help a researcher better understand his or her questions or the population being studied. Sharing some of those insights may help readers understand how questions developed or research progressed. Other times, however, details about a researcher’s intimate encounters are irrelevant.

Well-trained researchers can conduct careful, thorough studies regardless of the methods they choose, how much they participate, and their personal identities. In 2010, anthropologist Margot Weiss and sociologist Stacey Newmahr each published books on BDSM in the United States, based on research conducted during roughly the same time period. Weiss (2010) observed in a BDSM community without participating, while Newmahr (2010) became a BDSM player during her fieldwork. Their resulting ethnographies take different theoretical approaches: Weiss focuses intently on BDSM as part of capitalist consumer culture while Newmahr spends more time exploring the creation of authentic “scenes.” What each researcher observed, experienced, and concluded about BDSM was related to who she was and how she interacted with others at her field sites. Still, their descriptions of BDSM are factually similar, and both discerned the importance of authenticity for many contemporary BDSM practitioners. Weiss doesn’t seem to have “missed” significant aspects of BDSM because of her nonparticipant status, although she contextualizes the scene more broadly in US culture than Newmahr does. Newmahr doesn’t appear to have become too “close” to the community to analyze it effectively, although she homes in on the nuances of interaction and the phenomenology of BDSM play more than Weiss.

7 Some Practical Suggestions

The world is still a weird place, despite my efforts to make clear and perfect sense of it.

—Hunter S. Thompson

Observation is a skill that requires practice and ongoing reflection, but there are some practical ways to make the most of one’s abilities.

Reflect on the politics of visibility and mobility as you are designing your study and during fieldwork

Unfortunately, some students are sent out to practice conducting observations without any discussion of power dynamics or having done even a cursory literature review. Street sex workers, homeless people, strippers, and other visible, but marginalized, groups are frequently chosen as research subjects because they seem intriguing and edgy—the exotic others. When I was working in strip clubs, I sometimes found myself more offended by the intrusive questions and disrespectful behavior of unprepared researchers than by any of the regular customers. Reflecting on your motivations for selecting a field site and your assumptions about the individuals you interact with can make the difference between voyeurism and observation.

Some researchers interested in alternative sexual practices tend to study only within identity communities, recruiting from BDSM groups, polyamorous groups, and so on. Although identity-based communities are easier to observe—they may hold meetings, host events, and congregate in known places—such a strategy can miss individuals who do not embrace existing labels. The organized polyamory “community,” for example, tends to be privileged in terms of race, class, and education; individuals in other social strata, however, still purposefully engage in multiple sexual/emotional relationships.

Determine which environments are conducive to answering your research questions and then strategically gain access

Too often, researchers use the most accessible potential field site or do not adequately match their questions to their sites—many questions, for example, cannot be answered by observing in a single sex club or through observation alone. While it is obviously important to select a site where you will eventually gain entry, encountering barriers may be a sign that more preparation is necessary.

If one wishes to study a group that has been historically targeted by social scientists, sensitivity to community concerns about previous research will pay off. Groups like CARAS (https://carasresearch.org) can help with linking researchers to potential participants and designing studies that have the potential to “give back” to the community (or are at least respectful). Some groups may prefer that researchers follow established pathways when announcing studies or attending events.

Design your study to allow for maximum coverage

Observing a given environment at different times of the day, on different days of the week, and in different seasons of the year can generate essential information.

Reflect on the specifics of your participation

Participation may not be possible or desirable for illegal behaviors, such as drug use or prostitution. In other cases, deciding how you will interact ahead of time is beneficial even if you change your mind later. Will the boundaries you set for yourself be perceived as respectful or offensive by the other people around you?

Reflect critically on your positioning, appearance, and behavior

How you literally position yourself during your observations will impact how insiders interact with you. Understanding as much as possible about the norms of the environment can aid immensely. If you are attending an erotic event, for example, try to ascertain how you will indicate your intentions nonverbally. What are participants expected to wear? Are you comfortable in such attire? If so, would wearing it mislead other participants about your intentions? And if not, how might you present yourself so as not to offend participants or cross your own boundaries? Showing up in black leather might work well, but donning a hood so that no one recognizes you could be misunderstood.

Relax If you do not intend to participate, do not also assume that every individual who approaches you is interested in sex or attempt to prevent advances by adopting an avoidant body posture.

Be realistic about your attention span. Shift perspectives, and look away when necessary

Merriam (1998) suggests alternating between narrow and wide perspectives when observing a scene. One can also focus for a few minutes, look away, and then focus again. This advice can be taken literally, as when one is watching a performance or ritual unfold, or figuratively, as when one takes a short break from fieldwork.

Document in a way that makes sense for YOU

Sure, traditional fieldnotes are the gold standard of fieldwork. But depending on your research site, it may not make sense to record as events unfold. When working in strip clubs, I kept a notebook in my locker so that I could write some things down immediately. But I was far too tired after an 8-h shift to type fieldnotes at my computer after work, so I found it useful to audio record during the drive home. Later, I could either transcribe the recording or listen to it as many times as necessary.

Although some researchers traditionally suggested avoiding computers in the field, today’s technology is more omnipresent. Typing notes on a smart phone may work in some field sites, although care must be taken to maintain confidentiality in case the phone is lost, or stolen. Photographs and video-recordings can be legally and ethically problematic in some environments, but acceptable in others. Even if photographs or video-recordings are not prohibited, ask participants before using this technology—individuals engaging in stigmatized activity may have concerns about privacy and anonymity that outweigh the usefulness of preserving the scene and that should be respected. Written permission may be necessary to publish photographs or to use visual materials that are collected for non-research purposes.

Follow the confidentiality guidelines

Follow the confidentiality guidelines provided by your Human Subjects Institutional Review Board, or if more stringent, those of your discipline or that you have set for yourself. Some observational projects will be exempt from review, but if multiple methods of data collection are involved, informed consent will eventually become necessary. Confidentiality—or preferably anonymity—is important to maintain during covert and overt research, except under special circumstances. Some types of participant observation can jeopardize the confidentiality of others indirectly. Be careful when writing, storing, and publishing descriptions that could inadvertently be used to identify a place or person. Many researchers are critical of human subjects review boards for making decisions about research based on social acceptability rather than actual risk to participants, and the process of seeking human subjects approval can seem like a bureaucratic nightmare. You may be required to store consent forms separately from interview materials, delete identifying information from field notes, or keep your materials in locked cabinets. Although it is easy to let those steps slide, doing so is a mistake. If you are going to be collecting sensitive data, following the guidelines precisely can also protect YOU from lawsuits, being forced to identify your informants, or causing unintentional harm. One project that I designed, for example, called for multiple, in-depth interviews about cheating and monogamy with spouses, interviewed separately. The human subjects review board strongly suggested one-time interviews rather than repeated interviews, based on the potential for the interview process itself to cause issues in a marriage and to avoid any chance that I might accidentally divulge confidential information from one spouse to another. I accepted the suggestion, which did not substantively change the study. When I learned how little some spouses share with each other, I was glad to have done so.

If you do need IRB approval, start the process early to avoid costly delays if revisions are necessary.

Reflect on the dynamics of your relationships

Some theorists term those with whom researchers develop the most important relationships “key informants.” Others dislike the term. Either way, a particularly opinionated individual can shape a research project, especially if the researcher is inexperienced or extremely unfamiliar with the setting. On the other hand, gaining the trust and interest of an influential person is often crucial to interacting with a community. Individuals who embrace your research may be different in some fundamental way than those who avoid you—or not.

Consider ways to substantiate or invalidate your observations

Creating a system of checks and balances on your observations is specific to each field site. One possible strategy would be to ask for participant comments on your observations; another strategy would be to use multiple methods of inquiry, balancing observations with interviews, for example.

Don’t shy away from documenting the mundane

Knowing what is “regular” can help illuminate extraordinary occurrences. Seemingly unimportant events or details may take on new meaning as a project develops.

Don’t rely only on your eyes

Observation ideally engages all of the senses, not just sight. In your field notes or diaries, try to cover each of the senses when describing a scene or interaction.

Don’t interrupt

Although interviewing skills deserve a chapter of their own, one of the most useful field techniques is learning how to be quiet. Observational methods often involve informally conversing with people at a field site, and if so, try to allow for digressions and long silences. People often use talk to develop rapport, which means that they quickly fill in silences and sometimes do not even respond directly to what was said, as when stories are shared rapidly among a group. As a researcher, you will want to alternate strategies. Sometimes, we also unconsciously shift conversations in directions that we want or expect them to go. Using talk to develop rapport is fine, but practice shifting your focus to asking questions that you need answered—and giving people the time to answer them.

Pay attention to your mistakes, misunderstandings, and discomfort in the field

Part of the reflexive process is to challenge some of your most basic or cherished beliefs—something that is likely to cause discomfort. Certainly, foibles can be a source of embarrassment for researchers, sometimes not even recorded in field notes. But anthropology is rife with examples of misunderstandings that eventually led to a deeper appreciation of the culture being studied (Lee 1969; Bohannan 1966). Irritation with one’s informants or the others one encounters in the field may later prove to be illustrative of one’s own anxieties; overcoming emotional discomfort might later guide your analysis in new directions.