Personal Safety in Dangerous Places

Abstract

Personal safety during fieldwork is seldom addressed directly in the literature. Drawing from many prior years of ethnographic research and from field experience while studying crack distributors in New York City, the authors provide a variety of strategies by which ethnographic research can be safely conducted in dangerous settings. By projecting an appropriate demeanor, ethnographers can seek others for protector and locator roles, routinely create a safety zone in the field, and establish compatible field roles with potential subjects. The article also provides strategies for avoiding or handling sexual approaches, common law crimes, fights, drive-by shootings, and contacts with the police. When integrated with other standard qualitative methods, ethnographic strategies help to ensure that no physical harm comes to the field-worker and other staff members. Moreover, the presence of researchers may actually reduce (and not increase) potential and actual violence among crack distributors/abusers or others present in the field setting.

A Serious Problem confronting many social scientists is assuring the physical safety of ethnographers and other staff conducting research among potentially violent persons who are active in dangerous settings. Of equal concern is attempting to assure the personal safety of potential research subjects. Even when extensive ethnographic experience shows that physical violence against ethnographers has rarely occurred, researchers may have considerable difficulty convincing others (including colleagues and family members) that they can safely conduct fieldwork.

Some ethnographic research may be a dangerous enterprise. Howell's (1990) discussion of safety offered an extensive discussion of common law crimes (robbery, theft, rape) in the field. Field-workers have encountered illness, injury, or death in the course of fieldwork due to natural and criminal causes. It is often unclear whether the field-workers were harmed by research subjects and other members of the social networks or whether they were merely victimized like any other citizen (Howell 1990).

The question of personal safety is rarely addressed as a methodological issue in its own right (Howell 1990; Sluka 1990), particularly in regard to the social milieu in which ethnographers carry out their work. There is relatively little discussion about how to minimize risks and dangers that ethnographers may face in the field, with suggestions to help ensure their personal safety. Some hints about safety may be gleaned from the extensive methodological literature in ethnography (Agar 1980; Fetterman 1989) that deals with such topics as gaining access and recruiting subjects (Johnson 1990), striking a research bargain (Carey 1972), entering the field, making observations (Broadhead and Fox 1990), selecting roles to pursue in the field (Adler and Adler 1987), building and maintaining rapport (Dunlap et al. 1990; Rose 1990), conducting interviews (McCracken 1988), and writing field notes (Fetterman 1989). In practice, paying attention to the personal safety of ethnographers goes hand in hand with learning and applying skills in these areas.

The lack of good guidelines and methodological strategies for conducting safe ethnographic fieldwork in potentially violent social settings is especially noteworthy. In one of the few articles addressing safety issues, Sluka (1990) provided a systematic discussion of the risks and dangers facing ethnographers in a politically charged, potentially violent setting by studying supporters of the Irish Republican Army in Belfast. His suggestions are strikingly relevant for ethnographers in the substance abuse field and for those who study street- and upper-level crack dealers. Sluka called for “foresight, planning, skillful maneuver, and a conscious effort at impression management” (p. 115) to minimize personal risk and danger in potentially violent settings. He further suggested that field-workers become well acquainted in the community, cultivate well-respected persons who vouch for them, avoid contacts with police, be truthful about the purpose of the research, identify potentially dangerous locales and topics, and be flexible concerning research objectives. He proposed that successful fieldwork in dangerous settings “can be done by recognizing how people are likely to define you, avoiding acting in ways that might reinforce these suspicions, and being as honest and straightforward as possible about who you really are and what you are really doing” (p. 121).

Sluka's advice about personal safety is among the best available at the present time. In this article we will conceptually extend and apply his and others' (e.g., Adler and Adler 1987; Denzin 1970; Douglas 1972) ideas to ethnographic research in inner-city settings. We focus on issues of personal safety while conducting fieldwork in potentially dangerous settings. Closely related ethnographic issues—rapport, recruiting subjects, ethnographer roles, reciprocity, personal experiences with contacts, and so on—are briefly included in the discussion. While recognizing that ours is but one approach to doing ethnographic research, we contribute to the literature on ethnographic methods by underscoring themes and practices for personal safety that may be of interest to all ethnographers and staff conducting research in dangerous settings—or even in “safe” settings.

Method

This article emerges from the authors' many years of experience in conducting both quantitative (Johnson 1973; Johnson, Elmoghazy, and Dunlap 1990; Johnson et al. 1985, 1988) and qualitative (Carpenter et al. 1988; Dunlap 1988; Dunlap et al. 1990; Hamid 1979, 1990, 1992; Johnson, Hamid, and Sanabria 1991; Williams 1978, 1989, 1991; Williams and Kornblum 1985) research among abusers and sellers of marijuana, heroin, cocaine, and crack. All of the authors have done much of their work among low-income and minority populations (Dunlap et al. 1990; Hamid 1990; Johnson et al. 1985; Johnson, Williams, et al. 1990; Williams 1989, 1991; Williams and Kornblum 1985). These professional ethnographers (Dunlap, Hamid, Sanabria, and Williams) have extensive experience in qualitative field research on drug-related issues and other topics in New York City, Latin America, and the Caribbean. Four staff members (Arnold, Beddoe, Randolf, and Miller) are ex-drug users and/or ex-dealers who developed wide networks among upper-level dealers. Collectively, the staff have many years of experience working in or researching various aspects of drug use and dealing. The authors are professional researchers and ethnographers whose primary careers are built around research funded by grants. We recognize that many academic ethnographers do not have resources for hiring staff, paying subjects, and other support available via grants.¹ Yet the strategies and approaches reported here are vital for and applicable to our academic colleagues in that they have been assembled from years of professional ethnographic experience in dangerous locations.

Building on this experience, we systematically trained staff members on issues of personal safety during an ongoing study called “Natural History of Crack Distribution.”² This was a qualitative study about the structure and functioning of crack distribution, including the careers of dealers in New York City (Johnson, Williams, et al. 1990; Johnson et al. 1991). During the fieldwork phase of this study, November 1989-March 1991, the ethnographic staff spent an average of 15-20 hours per week in several of New York City's most dangerous locales interacting with numerous street people. Staff members conducted intensive fieldwork in four New York City neighborhoods (Harlem, Washington Heights, Brownsville, and Williamsburg). They wrote field notes that contain observations and references to over 300 different crack distributors. They also conducted open-ended life history interviews (5-15 hours long) with 80 distributors. Fifteen of these were upper-level dealers buying and selling kilograms of cocaine; the remainder were independent sellers. To obtain this information, they conducted three or more sessions with most dealers. All interviews were recorded and transcribed. Our data and analyses rely on the strategies and experiences of ourselves and our ethnographic staff for maintaining their own personal safety as well as on specific experiences reported by other ethnographers in the drug abuse field.

Laypeople and ethnographers anticipate and are fearful about several potential sources of physical danger associated with the use and sale of crack (Brownstein and Goldstein 1990a, 1990b; Goldstein 1985; Goldstein et al. 1990, 1991a, 1991b; New York Newsday 1990; New York Times 1990a, 1990b; Washington Post 1990): Crack abusers may be paranoid and behave “irrationally”; dealers routinely use violence and may threaten subjects who talk to ethnographers; use of guns leads to “random” shootings; researchers may be robbed or have articles stolen. The mass media typically feature the most violent and extreme activities of crack distributors (Reinarman and Levine 1989), so laypersons are led to believe that severe violence occurs all the time in this business. Despite these fears, ethnographic research in dangerous settings has been safely conducted for years. Our staff and many other researchers (Adler 1985; Biernacki 1988; Feldman 1974; Goldstein et al. 1990; Hanson et al. 1985; Spunt 1988; Morales 1989; Smith and Kornblum 1991) have met, talked with, and interviewed many potentially violent persons over long periods and have never been physically assaulted.³

Styles of Safety

Researchers can create “safety zones” in which to conduct research in dangerous settings so as to protect themselves and the persons with whom they are interacting from physical harm or violence during the research endeavor. The following sections are organized according to conceptual themes regarding styles of safety that emphasize demeanor, protector roles, safety zones, neutrality, and common sense during fieldwork.

Style and Demeanor

Style and demeanor are central to safety. First impressions are very important. Wearing clothes appropriate to the setting prevents drawing undue attention and exhibits a sense of belonging in the setting. Researchers' attire can be viewed as an extension and manifestation of their personalities as well as a willingness to fit into the social setting. As ethnographers enter and attempt to establish a presence in the field, they explain the purpose of research, exhibit personal interest in others, and avoid drug use or sales (Adler and Adler 1987; Agar 1980; Horowitz 1986; Johnson 1990). Failure to establish this presence, and especially being perceived as a victim, by those in the drug business for instance, may greatly increase personal dangers of theft/robbery and difficulty in establishing rapport with potential subjects. Although various roles have been employed by ethnographers in a variety of settings (Becker 1960; Adler and Adler 1987), those conducting field research among drug abusers generally employ a variation of friendly stranger (Agar 1980) or friendly outsider. This role is partially mandated by a professional code of ethics forbidding illegal behavior⁴ and institutional requirements to obtain informed consent from research subjects.

Purpose and Access

Once accepted as an ordinary person in the area, initial conversations are the first step in seeking persons with whom to develop rapport. Williams has been conducting research among cocaine users since 1974. During a 17-year career, he has visited hundreds of after-hours clubs, base houses, crack houses (Williams 1978, 1989, 1991), number holes,⁵ and other settings where illegal and legal activities occur. Williams explained several strategies for gaining entry into such locales:⁶

Initially I prefer to be taken into a crack house or dealing location by someone who is known there. They vouch that I'm OK and no cop. When initially approaching a crack house without someone to introduce me, I'll claim to be sent by someone they may know, like Robby, KeeKay, or someone else with a common street name. When I get inside, I may explain that I'm writing a book on crack houses (or another topic). I usually have a copy of a book I've written to show people. This approach goes a long way toward convincing skeptical persons that I'm an author and serious about my intentions.

After gaining initial entry and some rapport with one or more persons in the setting, Dunlap found it necessary to arrange a meeting with one or more drug dealers to explain herself, to seek their permission and informed consent to conduct long life-history interviews, and to strike the research bargain (see Carey 1972). The dealers can also examine the project's Certificate of Confidentiality.⁷ Dunlap explained:

I begin by telling them about myself, my life, and why I'm interested in them. I spend much time explaining how their identity will be concealed and how our interviews will be protected and never be available to police or law enforcement agencies. I explain the risks and benefits of the research to them in terms of their participation and obtain their informed consent. Even after these lengthy explanations, most subjects tend to remain tense and somewhat terse in their answers. Only during and after the first session of the in-depth interview do they begin to relax and talk openly about themselves. Such conversations would not even begin, however, without the assurances of confidentiality and the promise of benefits.

The end result is that ethnographers have built substantial rapport with one or more persons, carefully explained the purpose of the research, provided assurances of protection and safety, and obtained informed consent from persons who will become potential research subjects. Of course, the ethnographer must continue to meet with and show a genuine personal interest and friendliness to such persons. Such further conversations and interactions help build strong rapport with subjects.

The “Victim” Role

As ethnographers, we need a “mind set” that assumes safety and does not lead to fearful behavior. Street people act on their intuitions and are experts in reading behavior. Dunlap expressed the critical importance of not being perceived as a victim (“vic”):

The ethnographer's state of mind on entering the field must not include fear about studying violent people; at least such fears must not be at the front of one's mind. Overconcern about violence may cause ethnographers to appear afraid or react inappropriately to common street situations and dangers that do not involve themselves. Fearful behavior is easily inferred by violent persons from the way one walks and the way one interacts with others. Fearful behavior may place an ethnographer in the “vic” category to be targeted by others as a true victim of crimes like robbery and assault.

Not exhibiting fearful behavior does not mean abandoning choices about a sensible course of action. Rather, the mindset we have found appropriate is cautious, friendly, understanding, and open. This mindset emphasizes a degree of determination and self-confidence that does not leave room for ethnographers to be labeled as “vics.” Likewise, not using or selling drugs is also important for avoiding the “vic” role. If potential subjects observe ethnographers buying or selling illicit drugs, they may be suspected of being undercover agents, or expect them to be potential customers, or people who will share or provide drugs. Further decisions about whether to enter specific locales or meet certain persons must be made deliberately and based on the other themes discussed below.

Locator and Protector Roles

Two roles are especially important in conducting research among upper-level and in many instances among lower-level drug sellers. The roles of locator and protector are vital to the safety of persons working within the illicit drug industry. Locating individuals who can perform these roles can be critical to ethnographers' safety with and access to upper-level dealers. The ethnographic literature (Agar 1980; Johnson 1990; Liebow 1967; Whyte 1955, 1984) provides advice about finding one or more key informants who can provide access to others in the setting and who give much information about the phenomena being studied. Classical ethnographers have assumed that such representatives can be found or that they will come forth voluntarily so that ethnographers need only cultivate and build rapport with them.

Crack sellers and upper-level dealers, however, have very good reasons to insulate their identities, locales, and illegal activities from everyone (excepting their trusted co-workers). They are concerned with avoiding detection and arrest and with preventing robbery or injury by other street persons. They systematically evade conversations that may build close relationships (Adler 1985). Yet to conduct their business safely, they must rely on others who perform a variety of roles such as steerers, touts, guards, lookouts, connections, runners, and muscle men (Johnson, Williams, et al. 1990; Johnson et al. 1991).⁸ Approaching a crack dealer directly (without an intermediary) threatens the dealer, as it proves his identity is known or suspected. Ethnographers will always be suspected initially of being a “cop” or an “informer,” thus elevating the probability of personal risk and possible harm from the dealer or his associates.

Ethnographers can seek access to drug dealers through someone performing a locator role and rely on others to play a protector role as access is gained. Experience has indicated that access to crack dealers was most successful when ethnographers worked with a highly trusted former associate of the dealer who performed both the locator and the protector roles. The same person, however, need not perform both roles.

Critical in studies with drug dealers is someone who will perform the locator role of introducing the ethnographers into a setting where dealers are present. Recovered substance abusers who have had management roles in drug-selling organizations or have been incarcerated for several years for drug distribution crimes are particularly valuable in such roles. These ex-dealers typically have a large network of current sellers and dealers, know how to negotiate with active dealers, and can be trained to assist with fieldwork. They can locate and introduce ethnographers to several dealers (the locator role), provide protection in dangerous settings (the protector role), become systematic observers and interviewers (field-worker or interviewer role), and explain many of the informally understood norms to a professional ethnographer (the “expert” role).⁹

Proper Introductions

At the early stages of fieldwork among crack sellers, ethnographers generally do not attempt to enter a setting alone. Someone familiar with the locale is recruited or hired to assist in arranging “proper introductions” of the ethnographers to dealers as well as to provide protection. As a paraprofessional staff member, Arnold contacted several dealers and helped arrange interviews with our ethnographers. He stated that “the contact person has a major affect for the ethnographer upon people in the setting.” From his network of acquaintances, he had initial contacts and helped persuade dealers to talk with the ethnographer.

Another paraprofessional, Beddoe, explained why and how proper introductions occur among street dealers:

They [good contact persons] tend to have contact across time in the given area. Most street dealers are middle men. They will continue to work together and routinely rely on each other. Introductions by one dealer who vouches to other dealers that someone is “right” and “not a cop” is a vital part of street life and everyday dealing hustles. If an ethnographer gets a positive reference from a dealer, another dealer will still be a little suspicious. They study how you handle yourself in the field and then decide whether to talk more.

Having the appropriate person provide an introduction to a dealer is vital. Group members respond according to the reputation of the individual who provides the introduction. If that person is not trusted, the ethnographer will not be trusted. Dunlap's field notes recorded why she was unable to gain access to several dealers in one Harlem block:

My early contact on this block was Chief, a female who worked for several dealers, mainly as a “fill-in seller” at the street level. Chief had committed some act which had deemed her untrustworthy to most of her suppliers. She was only trusted to sell small portions of drugs at a time, never large amounts. When she attempted to introduce me to one of her bosses, it was disastrous. The dealer refused to even meet me. Seeing this, other street sellers whom I had informally met at the same time ceased interacting with me. From this and other experiences, I learned that lower-level crack users/dealers can seldom provide good introductions to their bosses or suppliers.

When a respected and trusted former dealer provided the introduction to other dealers only a few blocks away, several meetings and interviews were the outcome. Arnold reported:

I contacted several dealers who trusted me because we had done prison time for drug sales. After explaining the study to them, they were willing to attend a meeting. I set up the meetings and got them there, so that she [Dunlap] could explain it in more detail and build some relationship with them. This resulted in several interviews.

Dunlap reflected on her experience with recruitment of several upper-level dealers:

Contact with middle- and upper-level dealers comes only through the dealer's trusted and limited circle who continuously work to assure his personal safety. Although the dealer may be known by several persons in his or her area, a “proper introduction” of the ethnographer to the dealer can only be provided by someone the dealer trusts.¹⁰ Only through that person will an initial audience be granted, so to speak. The trusted contact is someone who has worked with the dealer for many years (either working directly with the dealer or selling drugs at the same time). Prison is also an important place where dealers have met persons whom they can trust as “safe.” One dealer commented, “Serving time tells what a man is made of.” The norms among prisoners and ex-prisoners eliminate the chance that the individual is an informant or an undercover cop.

The Protector

Ethnographers usually assume that they do not need protection from persons in a social setting. In the context of the drug business, this usual assumption is false; everyone must arrange protection to assure their personal safety. Once ethnographers are properly introduced into a setting, finding someone to perform the protector role is usually not difficult. Everyone in a drug dealer's network is expected to “watch backs” (i.e., help each other avoid possible dangers). Even freelance sellers competing for customers on the same block quickly reach agreement to “divide up” the territory and to “watch backs” for each other in case of physical danger (Johnson et al. 1991). Williams reported:

In every field setting, some person always appears to perform a protector role and “watches the back” of the ethnographer; he discourages violence among others in the setting “because the Man [ethnographer] is here.” If I leave the street for a month, it feels like a year. I need to maintain regular visits. Because I rely on them for protection, my best protectors in the street are enemies of the police: drug dealers, con men, robbers.

After gaining experience in similar settings, ethnographers can enter another site and expect to rapidly encounter someone who will perform the protector role. Usually, the protector will be among the first to speak to the ethnographer. In the event that a protector does not emerge or cannot be found (see “safety zone” next), or if a feeling of safety is lacking (see “sixth sense” below), researchers are encouraged to leave that setting.

Field Roles

Ethnographers have an anomalous position that potential contacts may find unfamiliar or unclear. While conducting field research, they occupy roles that are “betwixt and between” (Jackson 1990) their own professional roles and the roles enacted by potential subjects in the field setting. The dual role of observer and participant (Adler and Adler 1987) played by the friendly outsider (Agar 1980) is unfamiliar to most subjects. Rather, subjects tend to project familiar roles onto ethnographers.

In fact, field roles are fluid and changing during a typical day and during the course of the field research (Denzin 1970; Spradley 1980). In conversation and interaction with individual subjects and with groups, ethnographers can listen closely for the roles that others assign to them. This is helpful in designing one or more field roles that are compatible with the research, yet understood by subjects and protectors. During this study, several subjects referred to Dunlap as “auntie,” “mom,” “sister,” or other fictive kin; Williams was perceived as a “book author” and a “sharp dude”; Hamid and Curtis (1990) were “voyeurs” when conducting research in a “freakhouse” (where crack use and sexual activity occur). These subject-assigned roles were effective because they permitted access to the setting, were used by the protector to briefly explain the ethnographer's presence, and permitted informal conversation, questioning, and direct observations to occur—without suspicion that the ethnographer was a cop or a police informer.

As a single female living in a crack dealing neighborhood, Dunlap did not want research subjects to know where she lived, but she had to return home during early morning hours when only drug dealers and street people were awake. Dunlap described how she created and maintained a “right citizen” role with five regular crack dealers who helped assure her safety when she returned home very late:

I first observed who had the most respect from others and who appeared to have control over various situations—this was usually crack dealers. Then I walked by and said, “How you doing?” and engaged in “nonsensical” conversations about such things as the music on the street, street language, the drunk leaning against a fire hydrant. We avoided conversations about what they were doing or about what I did. I also avoided talking to the drug users. By being friendly with the drug dealers, they quickly accepted me as someone who would do them no harm. In return, they protected me in little ways. For example, one night after speaking briefly to my local dealer, a crack user began to approach me for some money; the dealer told her to “move on” and not to bother me. If some threatening situation were to arise, I feel certain they would act to protect me or intervene if necessary.

Dunlap also practiced this role in other research settings when interacting with persons who were not to be approached as research subjects.

During the past 6 years, many ethnographers and paraprofessionals have assumed the role of health worker doing outreach on AIDS prevention projects (e.g., Broadhead and Fox 1990). The “AIDS outreach worker” is an effective street role for ethnographers; it has become well known and respected among street people in several communities. The AIDS outreach worker role clearly “sides” with subjects and potential subjects and provides a basis for interaction with a variety of persons. Such persons express concern about subjects' health, facilitate referrals to other health service agencies, and help ethnographers to avoid being seen in a law enforcement role.

Safety Zones

When conducting research in settings that may be dangerous, or among persons who may be suspicious or hostile toward researchers, a first order of business is to create and maintain a physical and social environment in which ethnographers and potential subjects accept each other's presence.

In settings where many persons are present, effort should be made to include several persons as protectors in a safety zone. This is conceptualized as a physical area extending a few feet around the researcher, in which researchers and other persons within this area feel comfortable. The safety zone has three major components. First, researchers must have a feeling of “psychological safety”; that is, they must not feel endangered, they must experience some degree of acceptance by others, and they must be willing to stay in the location (see “sixth sense”). Second, other persons in this zone should accept ethnographers' presence, trusting that they are “right” and “not a cop.” Third, the physical environment must not be hazardous (e.g., the floors should not be likely to collapse; the ceiling should not be likely to fall).¹¹

When entering a locale, ethnographers can quickly scan the physical environment for obvious signs of danger. They should test steps and flooring, especially in abandoned buildings, and be cognizant of all exits. By introducing ethnographers to others at the site, the protector can facilitate social acceptance. Ethnographers must then establish their own right to be present in the locale during subsequent conversation with others. Such interaction typically brings about an implicit (and sometimes explicit) agreement, thus creating a shared sense of psychological acceptance or a “safety zone.”

During the initial visit to a setting, ethnographers can state plans to return in the future and attempt to judge how others in the locale feel about this. If a good relationship has been developed with a key person at the site (apartment resident, owner/manager of crack house, street dealer), ethnographers can return to the location without the initial contact and rely on people in the setting to provide protection and to help maintain the safety zone. Beddoe suggested,

Look how they talk with each other, and how they deal with each other, and try to copy their style. This will help you get to other people in the social circle. Any conversation is generally better than none.

When entering a new setting, Dunlap generally located potential exits and figured out who was in charge:

This is accomplished in a subtle and gradual way in order not to cause suspicion or make anyone feel they are being watched. I call this getting the feeling of the place, people, and conditions. Try to fit in by taking a comfortable stance, giving the impression of familiarity with various situations or scenarios.

Williams usually created the safety zone by paying careful attention to the setting and people's activities:

Use your own style and smooth approach. Usually don't be aggressive. Try to figure who is available for a conversation and talk to them when [they're] ready; otherwise wait. They communicate with each other via certain physical gestures which can be learned, especially when “thirsty” for smoke [crack]. Let them know that you really want to talk to other people and meet others. You don't want to create enemies out in the field. You have to be constantly improvisational in the setting. Don't overstay your welcome. Three to 4 hours in one place is too long, so move on. You have to be aware of who you are [a researcher] and where you are at. This is not a recreational place; it is a place where you are conducting research but others are buying and selling illegal cocaine. One should follow the rules of the street—which is surviving.

As rapport with persons in such settings is increased, a safety zone is created among those present. Norms usually include strong expectations of reciprocity. Dunlap described how she responded to these expectations:

Be counted on to “do the right thing” for them personally, even though [you're] not taking part in what is happening (you do not sell or smoke crack). I was always prepared to participate in ethically appropriate exchanges. I would accept a cigarette but more frequently provided them to those present. I would provide food or coffee that was shared by all or help a person read something. On the other hand, I avoided sharing drugs and declined to chip in to buy drugs.

This safety zone is a short-term agreement among persons in a concrete locale about the right of other persons to be present. Such temporary agreements do not imply that the potential subjects present have provided ethnographers with informed consent, acceptance, rapport, or a willingness to be interviewed. The safety zone only provides a locale and time during which ethnographers can begin to obtain further cooperation from some of those present.

Humor and Neutrality

Even when ethnographers function within a safety zone, a variety of tense interactions and situations may arise in specific locales. The effective use of humor and neutrality in these settings by ethnographers may also have important benefits for persons in these settings.

Ethnographers' neutrality in tense social situations is well described (Fetterman 1989; Agar 1980; Adler and Adler 1987), but it is sometimes a source of tension between their subjects and themselves (Broadhead and Fox 1990). Humor can defuse such tense situations and build solidarity among group members (Seckman and Couch 1989). Less well documented is the way in which humor and neutrality may help in dangerous situations (Carpenter et al. 1988).

Crack houses and drug-dealing locales are characterized by high levels of mistrust, paranoia, and potential violence. At the same time, these locales are at least partially organized to reduce violence and informally control persons (New York Times 1991) who act aggressively. Hamid described the dangers:

In crack houses, users constantly argue and accuse others of using too much or hiding or stealing crack or money. Street sellers face frequent arguments about money, the quality of crack or other drugs, threats of robbery, and other topics. Usually these arguments are resolved by the disputants reaching some kind of agreement, but other persons (guards, boss, owner) may occasionally intervene if the argument begins to escalate to physical violence.

After establishing a safety zone and acquiring protectors, ethnographers in a crack house or crack sale location may introduce an element of stability and safety. Ethnographers are not under the influence of drugs or alcohol and can think swiftly and clearly. They do not want to buy crack or sell drugs or to be used for such purposes. They are neutral in the various disputes between persons and attempt to maintain communication with all. They have requested and generally been granted protection and safety while in the location. Moreover, ethnographers are sophisticated in interpersonal relationships and can deal with tense situations. Hamid described how he sometimes intervenes:

When two crackheads are arguing about who got the most crack or stole it [the truth is, they've both used it up], and the dispute is heading toward a physical fight, I begin telling an outrageously funny story that has nothing to do with the conflict. The disputants are distracted from their conflict, they laugh and separate; usually the dispute is forgotten. Humor is a major way that tense arguments between crackheads or distributors may be resolved without blows and without any loss of face by either party.

Williams also noted that

a humorous remark, well-timed comment, or casual-appearing interruption by the ethnographer may distract persons who are headed toward violent confrontations. Bring humor into a very tense situation; get them laughing with the ethnographer and with each other. The gift of gab will get you [and them] out of all kinds of difficult situations.

Dunlap explained how she has deliberate conversations with crack users while they are “straight” to reduce the potential for subsequent violence:

Many crack users try to convince themselves and others that the drug does not affect them. They claim their behavior remains the same after they smoke crack as it is before they smoke crack. I always bring this discussion up when the individual is sober, before he of she has ingested any drug. When their behavior begins to change after smoking, I can usually bring them back to normal behavior by remarking that they are acting differently by smoking crack. Persons will try to prove that the drug does not affect them in various ways, and that they can handle the drug. While restating these claims, they generally abandon various kinds of behavior associated with crack intoxication. Also, I never take sides in any disagreement. Let the situation work itself out. If I feel the situation is becoming too dangerous, I leave.

Williams also noted,

In crack houses or similar settings, the ethnographers' presence can help reduce the risk of violence among people who argue and fight over drugs. People respect ethnographers and choose not to express overt violence in their presence. Local norms that are tolerant of threats and violence are set aside, and temporary norms prohibiting such threats are accepted.

By remaining neutral but interested parties, ethnographers gain respect from people in these settings. In many potentially violent situations, ethnographers may be the only “neutral” person who is not high and may become a mediator between individuals and groups. Such neutrality involves not engaging in personal (especially sexual) relationships with subjects during the study as well as avoiding alignment with only one group. Ethnographers' personal “safety zone” is frequently extended to protect subjects and potential subjects from the possible dangers that their own behavior and willingness to use violence may bring about. Thus the presence of ethnographers probably reduces the risk of violence among crack users and sellers in crack houses and dealing locations rather than increasing the potential for violence.

“Sixth Sense” and Common Sense

Not all conflicts and issues in dangerous settings can be resolved by neutrality and humor. Ethnographers need to be prepared to respond effectively in a variety of potentially dangerous circumstances (paranoia, sexual approaches, robbery, theft, shootings, police raids, and arrests) that actually occur infrequently but are a major fear among nonethnographers. Reliance on prudence, common sense, and a “sixth sense” can help reduce physical violence to a minimum. Different kinds of potentially dangerous situations can be handled by evasion and movement away from the danger, controlled confrontation, or rapid departure from the setting. The ability to handle a variety of situations requires both a “sixth sense” for danger and skill in moving away from and evading physical harm. Dunlap provided an illustration:

Acting from the “sixth sense” is relatively easy. We use it all the time in everyday life when we walk into new situations. There is an uneasiness, an inability to verbalize what is wrong. You may be able to explain everything that is taking place, do not see anything out of the ordinary, but still feel uncomfortable. This is a warning that something may go wrong. When such discomfort occurs, leave as soon as possible. For example, I had planned to hang around Ross and his family on a particular weekend. Each time I made preparations to leave, this uneasy feeling arose—I did not want to go and could not explain it, so I did not see him until the following week. Upon arrival, Ross reported that one of his partners had been shot and killed. If I had gone that weekend, I might have been next to Ross, who was sitting beside his drug-selling partner when the latter was shot by the father of a crack customer.

When and if ethnographers get a feeling of discomfort without reason, they will be safer by leaving the setting and returning another time—even though they may fail to gather some data and violence may not actually occur. But if their “sixth sense” has extracted them from the locale, they will not be harmed during those rare occasions when serious violence does occur.

Crack-Related Paranoia

Cocaine and crack induce a short-term paranoia in which users are very suspicious of others around them. They may believe others are enemies out to arrest or harm them. If challenged, pushed, or threatened, they may become unreasonably aggressive or violent. Yet crack users opt for avoidance and nonconfrontation to handle such short-term paranoia exhibited by other crack users. Williams has dealt with crack-induced paranoia in many settings:

When people are smoking crack, they go through different stages, one of which is paranoia. The crackhead may comment, “I don't like to be around people who don't get high” or “Why are you watching me?” This person may even be your sponsor or protector but is no longer the rational person you came with. The easy solution is to move away and not watch. Above all else, don't confront or challenge them. Usually you can find another person who is in a talkative stage where they want to talk. After a while, the first person's paranoia will subside and the person is open to conversation again—with no or little recollection of his comments or implicit threats while high.

Sexual Approaches

While using crack, a person may express a desire for physical closeness or sexual intimacy and approach others (including the ethnographer) for satisfaction. Williams explained how he responded to various levels of physical closeness:

There are touchers; persons who seek affection while they ingest drugs and smoke crack. They seek such affection and closeness when they get high, just before the effect wears off. Usually, I just move away or shift to conversation with someone else. What do you do when sexually approached? Be forceful and let them know that your aren't available for sex play; they usually will not pursue it further.

Several female ethnographers have had their fieldwork severely constrained or have had to terminate it completely (Horowitz 1986; Adler and Adler 1987; Howell 1990; Warren 1989) due to the sexual expectations and demands of subjects or other males in the research setting. The threat of sexual assault or rape is a real concern for most female ethnographers and staff members. As a woman, Dunlap followed several strategies to reduce vulnerability to sexual approaches:

Smoking crack causes many individuals to be stimulated sexually. Yet when first developing rapport, potential subjects frequently assign a fictive-kin role.¹² I may seem like a sister, cousin, mother, or aunt to them. Assuming such roles leads individuals to become “close friends” and share many behaviors they would not otherwise exhibit. When projecting such roles to me, they place me “off limits” for sexual approaches and affairs. Enough crack-using women are available for sexual affairs; neither male nor female subjects need me for sex and usually agree to protect me from advances by others. The crack-sex link focuses on the sexual act, not personal relationships. Even women who routinely exchange sex for crack or money will refuse sexual foreplay and intimacy for short periods during their crack consumption cycles; both men and women leave them alone at those times.

The value of the protector role was evident one evening while Dunlap was observing several prostitutes with whom she had established good rapport:

I was standing on the sidewalk talking with Lisa (a prostitute who used crack), when a John (customer) drove up and starting talking dirty to us. Lisa talked back to him while I listened. When Lisa said she wants $20 for a blowjob, the man replied, “I don't want you. How much is she [referring to me]?” Lisa exploded: “She ain't one of us. You leave her alone and keep your fuck'n hands off her.” She started kicking and pounding the car. The man looked surprised and drove off quickly.

Abandoned Buildings and Other Dangerous Locales

Most ethnographic research is conducted where the physical environment is structurally safe. Assuming such safety can be dangerous when researchers are studying crack dealers. Crack dealers may set booby traps to slow police or potential robbers. The sale and use of crack often occurs in abandoned buildings, rundown tenements, and hidden locations (e.g., under bridges or tunnels). Such locations are best approached only with a protector who knows it well. Even then, visits should occur only when the ethnographer feels comfortable. Dunlap recalled her trepidations:

A street contact said, “Let's go to a place where a friend lives. I'm doing this as a favor to you.” She took me into an abandoned building where her friend gave us a back room. I could have been robbed. But nothing happened. The interview went well but the place was unheated and filthy. On other occasions, I have rented apartments or hotel rooms for interviews because I didn't want to go into particularly bad abandoned buildings where subjects lived. If I enter an abandoned building (most have serious structural defects like broken steps or holes in floor), I do so only with people who know their way around defects that could cause serious harm.

The presence of a protector who can vouch for the safety of the premises and serves as a guide around several obstacles is critical in deciding whether to go into abandoned buildings or outdoor locales that researchers perceive as dangerous. Typically, ethnographers and subjects prefer more neutral settings like a coffee shop, restaurant, storefront, or apartment of a friend (which usually have comfortable chairs, heat, and some privacy).

Crimes and Threats Involving Money

Robbery, burglary, and theft from field staff are uncommon but do occur (Spunt 1988). In fact, many crack distributors are frequent and proficient robbers, burglars, and thieves (Johnson, Elmoghazy, and Dunlap 1990). Furthermore, crack users are constantly broke and in need of money. Thus we have developed strategies to minimize criminal victimization and monetary losses. Ethnographers and field staff can expect to be constantly approached for money, “loans,” and “advances” (Johnson et al. 1985, 205-6). When these are not provided, implicit threats may be made. Dunlap defused threats by trying to provide balanced reciprocation:

While declining to provide cash to the “kitty” towards the next purchase of drugs, providing cigarettes, candy, food, drinks, and refreshments will usually satisfy one's social obligation to contribute to shared group activities in a crack house or among drug sellers.

Usually, persons in protector roles will prevent threats from becoming robbery attempts. Johnson described one simple precaution that may reduce the magnitude of monetary losses if a robbery or theft occurs:

While in the field, wear clothing with a lot of pockets. Distribute the money into different pockets and keep $10 in a shoe for emergencies. While in the field, only take money from one pocket—conveying the impression that all my money is in that pocket. If someone observes and actually attempts a hold-up [which has not happened yet], give the contents from only that one pocket. When money in that pocket gets low, go to a private place (e.g., a bathroom) and transfer money to the spending pocket.

In prior or concurrent research projects (Johnson 1990; Johnson et al. 1985; Goldstein et al. 1990), some staff members have been robbed, and in one case, a physical assault without serious injury (Spunt 1988) occurred. When crimes occur, staff members usually report them to police to indicate that such violations will have consequences. Several thefts of tape recorders and minor personal possessions were not reported, due mainly to lack of police interest.

Fights

A physical fight or show of weapons may break out without warning so that ethnographers have little chance to use humor and neutrality to prevent it. Almost always, such weapons and fights have nothing to do with the ethnographers' presence. Rather, they are linked to disputes with other crack abusers in the locale. Williams followed several strategies for dealing with such occurrences:

Sometimes knives or guns appear, more frequently as a display of possessions (like gold chains or sneakers) than as a means of threatening persons. If they seek approval for their new possession, I may comment about how nice it is, but add “Guns aren't my favorite thing. Could you put it away?” I've been in hundreds of crack houses and dealing locations where weapons were widely evident, but I've never been present when guns were used in a threatening manner. If such an event were to occur, I'd leave as soon as possible, and not get involved as an intermediary.

Stickup Men and Drive-by Shooting

Perhaps the most dangerous situation is a “rip off.” This occurs when robbers surprise the occupants at a crack-dealing locale with the clear intention of taking all cash and drugs present. Likewise, when two or more drug dealer groups are competing for a good selling location, they may try to “warn” others by street shootouts. These are not situations for mediation or humor, only for getting out of the way or following orders. Beddoe noted:

Stickup men have usually cased the location and are quite certain who is present before coming in. They want money and drugs. Keep quiet and provide what they want.

Williams noted that ethnographers who have good rapport with dealers may be relatively safe:

How do you know a territorial dispute is going on? Generally, someone will let you know so you can stay out of the way ‘til some order has been reestablished. You have more warning of trouble than ordinary citizens in the contested area. Your contacts can provide information later—without your being present.

Violence and shooting in the drug culture/business is unpredictable and without warning because surprise is frequently a major element in its use. But most violence by crack dealers is intentionally directed at specific persons and occurs in a concealed setting (so no witnesses are present). Drive-by shooting/machine-gunning of people on street corners and “stray bullets” that kill children remain the exception in drug-related violence, even though they are a major feature of sensationalizing mass media coverage (Daily News 1990; New York Newsday 1990; New York Times 1990a, 1990b).

Contacts with the Police

Since 1983, police task forces directed against dealers have frequently engaged in surprise raids against dealers and crack houses. Despite concerns about police action and fear of arrest, ethnographers who avoid using and selling drugs themselves are rarely involved with the police. The police are authorized to use force only when a person resists physically, so ethnographers contacted by police are rarely arrested (Bourgois 1990). Particularly in street settings, ethnographers must be careful in dealings with police. Informal conversations with police should be avoided so that subjects and potential subjects do not have a basis for believing that the ethnographer is talking to or “informing” the police. When the police behave unprofessionally toward subjects, ethnographers who are observed to “stand up” to police gain respect in the eyes of potential subjects. Williams reported one such incident:

One night I was on the streets with a white ethnographer in a copping area. One police officer came up and asked, “Say, white boy, why don't you buy drugs in your own neighborhood?” and pushed him against the squad car to search him. After producing identification showing that he lived within a couple of blocks, he got off with no further hassle. People in the community saw this as harassment by police and concluded that the ethnographer was not a police officer.

Dunlap's field notes recorded the following incident:

One afternoon I was with two female subjects who were going to cop some drugs inside an abandoned building. While they copped, I went to the store to buy sodas. When I came out, two male cops had the two subjects against the wall and were patting them down. One of the cops was verbally degrading a subject. I inquired what happened and observed the police action. Not finding any drugs (only a crack stem), they let the subjects go. The women believed that my inquiry had saved them from arrest and that the cops were trying to “shake them down” for money. They felt that my presence had deterred the police from carrying out any adverse actions.

Despite many hours and days spent with crack dealers in crack sales locations where police were observed several times a day, our research staff have never been present when “busts” occurred. During a parallel study of sex-for-crack in Miami in 1989, however, Inciardi (personal communication) walked through the back door of a crack house as a police raid came through the front door. He and others present were taken to detention where he was held for 5 hours; he was released at booking without formal arrest following the procedures outline below. He did not, however, return to that crack house or others in its general vicinity.

On a parallel research project evaluating the impact of Tactical Narcotics Teams (TNT), ethnographers at Vera Institute (in cooperation with the New York City police) have been instructed about appropriate procedures to follow in the event of being caught in a police sweep or raid. The ethnographers are not to resist arrest nor attempt to talk to police officers at the arrest location. Rather, they are to follow instructions, let police gain control, and allow themselves to be handcuffed and taken to the station house. At the point of booking, ethnographers should present identification as a researcher working for a nonprofit organization and ask the booking sergeant to call the principal investigator or let the researcher make such a call. If possible, staff members try to arrange the researchers' release at booking, without formal arrest charges. Otherwise, senior staff or a lawyer will be present at arraignment and will attempt to persuade the judge to drop charges or provide bail money. Subsequent efforts will be made to have charges dismissed or the conviction overturned. To date, researchers at Vera Institute or Narcotic and Drug Research, Inc. have not been arrested while conducting research during the 1980s. Narcotic and Drug Research, Inc. now retains a lawyer to act quickly to represent staff, both ethnographers and paraprofessionals, arrested during field-work or AIDS outreach activities.

Conclusion

In this article, we have drawn on concrete experiences with a wide range of dangerous situations and subjects we encountered in conducting field research. The approaches described here have evolved over more than 25 years of ethnographic research successfully conducted by ourselves and others among users and sellers of heroin and crack in some of America's most dangerous social settings (see Broadhead and Fox 1990; Feldman 1974; Goldstein et al. 1990; Johnson et al. 1985; Preble and Casey 1969). Yet after spending 2 years involved in direct research among crack distributors and many other years of research with robbers, burglars, murderers, and heroin sellers, none of our professional ethnographers or paraprofessional staff (ex-dealers, ex-drug users) has ever been physically injured; few, in fact, have been robbed or burglarized while performing their research roles. If violence in the drug culture truly occurred on a random basis, then several ethnographers should have been harmed. This is not to suggest that the participation and activities of ethnographers is inherently less dangerous than those employing survey or archival methods. Yet recognizing the potential dangers, it is still possible for ethnographers to choose their field of study with a clear awareness, preparation for, and avoidance of the risks involved.

Biography

Terry Williams is an African-American sociologist who specializes in ethnographic research. His recent books document changes in the cocaine culture (Cocaine Kids 1989; Crack House 1991) and low-income teenagers (Growing Up Poor 1985). Recent research focuses on con games, pornography, and drug use in public housing. He is Associate Professor of Sociology at the New School for Social Research.