(p.239) Appendix: Notes on Research Design and Method
(p.239) Appendix: Notes on Research Design and Method
Sociologists engaged in participant-observation research increasingly recognize the importance of providing readers with an explicit discussion of the research design and methodology of their studies. While I do not believe that participant-observation will ever be—nor should it strive to be—“scientific,” it can—and should—be rigorous and systematic. Information about a study's design and method add to a reader's ability to evaluate the merits of the research on its own terms. It is in that spirit that I offer the following brief remarks.
Design: The Birth and Development of a Research Project
One of the most useful features of the research reported in Bargaining for Brooklyn is that the project's design allowed me several comparative dimensions of analysis: across neighborhoods (two neighborhoods); across organizational types (four types); and across organizations (eight organizations). Few studies based on participant-observation data have this potential for comparison, as research sites are often singular: a street corner or block (e.g., E. Anderson 1978; Duneier 2000; Liebow 1964; Susser 1982), a housing project (e.g., Small 2004; Venkatesh 2000), a neighborhood (e.g., Bourgois 1996; Gans 1964; Hannerz 1969; Kornblum 1974), an organization (e.g., Halle 1984; Owen-Smith 2001; Vaughan 1996), and so on (for ethnographies with comparative designs, see, e.g., McRoberts 2003; Newman 1999; Sullivan 1989). While I have discussed some of the interesting differences between neighborhoods, organizational types, and organizations over the course of the preceding chapters, the strength of the comparative design turned out to be the view it afforded me onto a broader sociological process common to all: the structural position of community-based (p.240) organizations within fields of action, and the opportunities and constraints this position affords and imposes on CBOs as they attempt to strike better resource bargains for the residents of their neighborhoods. The fact that I was able to see this process in operation across multiple organizations, different organizational types, and distinct neighborhoods is a critical evidentiary basis for my conclusions. Since comparative claims are only as good as their basis for comparison, it seems appropriate to discuss how I selected my research sites.
The participant-observation method of research allows for the emergence and exploration of themes beyond those that the researcher originally intended to study. Indeed, participant-observation research is full of such accounts, where the original focus of the project faded into the background as other, more important issues were uncovered. This research was one of those instances. When I designed the initial phase of this project as my doctoral dissertation research, I envisioned a comparative study of poor urban neighborhoods' efforts to promote their own revitalization. At the time, both scholars and policy makers were becoming increasingly interested in the idea of “social capital” as a resource for improving the conditions of poor neighborhoods, and I was particularly interested in the role of resident participation in this process. The design called for one study neighborhood to be “successful” in these efforts, and the other to be “unsuccessful.” The object was to determine what contributed to “success” in neighborhood revitalization. Within each neighborhood, I planned to study four organizations, one of each of the following types: a community development corporation, a child care center, a church, and a family support center. On the most basic level, I selected these four organizational types in order to capture a range of neighborhood functions and residents, thereby gaining broad information about usage patterns, key residents, important gathering places, residents' opinions of the neighborhood, and other factors important to understanding revitalization efforts in the local environment. In addition, I had a rationale for choosing each of these specific organizational types, centered around my core interest in resident participation.
I included community development corporations (CDCs) in the study because it is their explicit mission to improve the conditions found in their neighborhoods. CDCs usually focus on the provision and maintenance of affordable housing, but they often pursue other activities as well, including the involvement of local residents in their work (e.g., Saint Nicholas Neighborhood Preservation Corporation 1997; Vidal 1992). I selected child care centers on the assumption that most parents are highly motivated to seek (p.241) a safe, quality environment for their children, and so would be likely to become involved in these organizations. I chose churches based on a then-growing interest in the importance of religious institutions as centers of neighborhood participation and action, particularly in low-income communities (e.g., Carle and DeCaro 1997; Rooney 1995; Scheie et al. 1994). Within the larger category of churches, I selected Catholic churches, for three reasons. First, as I will discuss shortly, I planned to focus my research on Latino neighborhoods, and most Latinos in the United States, as well as in their countries of origin, are Catholic (Espinosa, Elizondo, and Miranda 2005). Second, the Catholic church has a long history of activism on behalf of the poor, including recent church-based community organizing efforts (Rooney 1995; M. R. Warren 2001). Finally, the Catholic parish's organizational structure ties each church more closely to its geographic setting than is the case in other Christian denominations, and this has significant implications for neighborhood improvement work (McRoberts 2003). Finally, I included school-based family support centers in the study because of recent efforts to make school buildings “one-stop shopping centers” for a range of services to children and families. The logic behind these programs is that school buildings sit empty for much of each day, while youth and families desperately need facilities for activities that can constructively fill nonschool time and assist adults with efforts to upgrade their education and job skills. In addition, school-based service centers are thought to enhance parents' and children's relationships with the educational system. New York City has adopted these principles for its Beacon program, which has grown rapidly over the last ten years (C. Warren, Feist, and Nevarez 2002).
These decisions about the basic structure of the research design then led to the specific question of where—in which two neighborhoods and in which eight organizations—I would actually conduct my participant-observation. One of my dissertation advisers at the University of Chicago, Richard Taub, wisely pointed out that no matter how meticulous my research design, there were no guarantees that I would be able to find real-life research sites that matched the plan I had put down on paper. Of course, he was right. The neighborhoods and organizations where I conducted my research hewed close to my goal of establishing valid comparisons, but the complexity of social life meant making some compromises as well.
Choosing the Study Neighborhoods
Since the year 2000, when Latinos officially became the largest minority in the U.S. population, scholars have produced a rapidly growing number of (p.242) ethnographic studies about this group.1 Bargaining for Brooklyn is part of a continuing effort to examine the wide range of Latina/o histories, achievements, exclusions, and challenges in the United States. When I began this project, despite the fact that poverty and its related social problems were widespread among U.S. Latinos, most research on these issues focused on African Americans. This was particularly true of urban ethnographies, where a rich tradition of using this method to study African Americans existed alongside a relative dearth of similar studies on Latinos.2 I had begun to develop an interest in New York City's Latino population at the end of the 1980s, when I worked for several years at a Latino arts organization while I was an undergraduate student. As John and Lyn Lofland remind us, researchers' professional interests often emerge from “where we are” in our personal lives (Lofland and Lofland 1995), and my work experience with Latinos in New York dovetailed with the state of the urban ethnography literature. Thus, I chose to conduct my research in two predominantly Latino neighborhoods.
The next step was to further delimit the characteristics of the study neighborhoods. I decided that they should be low-income—thereby assuring that some kind of neighborhood revitalization efforts would be operating—but not the most devastated areas in the city. Thus, I chose neighborhoods that met the social science standard for “moderate poverty”: between 20 percent and 40 percent of residents living below the federal poverty line (Sampson, Morenoff, and Earls 1999; Wilson 1987). Recognizing the effects that immigrant populations have been shown to have in revitalizing distressed areas (Winnick 1990), I sought areas with similar concentrations of immigrants. Finally, I wanted to avoid neighborhoods that were the recipients of unusually high levels of attention to and resources for neighborhood improvement work, such as being the targets of extensive private philanthropy or being designated as federal Empowerment Zones. It seemed unlikely that I would be able to match neighborhoods on more than these four basic characteristics, so they became my baseline for neighborhood selection.
(p.243) There are many ways to define a neighborhood (e.g., Hillery 1955; Sampson, Morenoff, and Gannon-Rowley 2002). I decided to use the administrative boundaries of New York City's fifty-nine community districts, which have cultivated a sense of common organization and identity among local residents since they were established in the early 1970s. It is hard to imagine today, but when I started this research in 1997, it was not possible for a researcher to immediately access the vast amounts of data now readily available online. Instead, I learned about the population characteristics of the community districts the old-fashioned way: from the occasional profiles published by the New York City Department of City Planning (New York City Department of City Planning 1992, 1993). A review of these sources, which used data from the 1980 and 1990 U.S. Censuses, revealed thirteen community districts where Latinos outnumbered other racial and ethnic groups.3 One additional neighborhood had nearly equal percentages of Latino and white residents, so I also considered it as a possible study site.4
I eliminated seven of these fourteen neighborhoods because they did not meet the other three basic criteria I had set up for the study (Marwell 2000). With seven neighborhoods remaining, I conducted interviews with fourteen informants knowledgeable about neighborhoods and neighborhood revitalization efforts in New York City including scholars, philanthropic foundation program officers, community activists, and employees of city and nonprofit agencies.5 The goal of these interviews was to develop (p.244) more in-depth profiles of the seven neighborhoods in order to choose two that showed noticeable differences in the success of their neighborhood revitalization efforts but were similar in terms of other descriptive characteristics.
In the interviews, I briefly described my research project, then asked respondents to name neighborhoods they felt were conducting successful revitalization efforts, as well as areas where work to improve the neighborhood had been unsuccessful.6 When asked where exciting neighborhood improvement work in Latino areas was taking place, many of my informants immediately named Williamsburg, in Brooklyn. They repeatedly described Williamsburg as an energetic neighborhood with vibrant community-based organizations accomplishing many important tasks. One informant remarked that Williamsburg was widely considered to be the site of the most successful grassroots-oriented neighborhood revitalization work in all of New York City.7 Another described in detail the housing development history of Williamsburg's two CDCs, which he said together owned and managed some twenty-four hundred units of housing, as many as the neighborhood's largest private landlord.8 A third informant noted that Williamsburg had recently been selected as one of only three New York City neighborhoods to take part in a new neighborhood preservation project being spearheaded by the New York Community Trust, the city's community foundation.9 Other expert informants made similar observations about Williamsburg, and no informant disagreed with the evaluation of Williamsburg as a place where highly successful neighborhood revitalization efforts were taking place.
(p.245) Williamsburg is one identifiable section of Brooklyn's Community District 1, which also contains the Greenpoint section. In 1997, the best available data showed that CD l's overall population was 43 percent Latino and 46 percent white, but that the two sections of the CD had extremely distinct population compositions (New York City Department of City Planning 1992). Williamsburg was majority Latino, though it also contained one of the city's highest concentrations of Hasidic Jews, who are classified as whites in the census. Greenpoint was majority white, home mostly to people of Italian and Polish ancestry (DeSena 1990). By focusing only on the Williamsburg section of CD 1, then, I would have an area that met my criterion of a majority Latino population and continued to meet my other three criteria as well. Furthermore, my expert informants indicated that Williamsburg was extremely active in neighborhood revitalization work. Williamsburg thus became an attractive option for the “successful” study neighborhood.
Using Williamsburg as a likely research site, I explored with my expert informants the characteristics of the other six neighborhoods as possible comparison sites. I eliminated two Manhattan neighborhoods, Washington Heights and the Lower East Side, primarily because their population mixes were quite different from that of Williamsburg (Marwell 2000). The remaining four neighborhood possibilities included one in the Bronx, one in Manhattan, and two in Brooklyn. Given that the community districts are parts of various larger administrative areas in the five boroughs of New York City, it seemed appropriate, all other things being equal, to select my two research sites from the same borough. This would eliminate some of the political and administrative variation that would otherwise obtain: two neighborhoods in Brooklyn necessarily would have more similarities than would one neighborhood in Brooklyn and one in the Bronx. They would deal with the same borough president and might have the same U.S. representative, state assemblyman, state senator, or city council representatives. Also, certain government offices and personnel might be the same for two neighborhoods in the same borough but not for two in different boroughs. For this reason, I decided to eliminate both East Harlem (Manhattan) and Soundview/Parkchester (the Bronx).10
(p.246) This left me with two neighborhoods from which to choose: Sunset Park/ Windsor Terrace or Bushwick, both in Brooklyn. Both areas were good candidates. They had similar mixes of Puerto Rican and Dominican residents. Both incorporated industrial areas into their largely residential character, although Sunset Park/Windsor Terrace had a healthier industrial sector. There were some major differences in the economic status of the two neighborhoods, however. While Sunset Park/Windsor Terrace had a poverty rate of 23 percent, and a median household income of $25,875, Bushwick's poverty rate was 40 percent, and its median household income only $16,285. Another significant difference was the population breakdown by race. In addition to its 51 percent Latino population, Sunset Park/Windsor Terrace had a significant and growing Asian—largely Chinese—population (10.4 percent). In addition, Sunset Park/Windsor Terrace was 34 percent white and only 4 percent African American. In contrast, Bushwick's 65 percent Latino population was followed in size by the African American population (25 percent). Whites and Asians each comprised less than 5 percent of the Bushwick population.
Comparing the two potential study sites with Williamsburg, Bushwick appeared to be more demographically similar. Williamsburg/Greenpoint's poverty level was 35.7 percent, and its median household income was $18,905. My decision to focus only on the Williamsburg section of the area, removing the somewhat better-off Greenpoint section, pushed the poverty level slightly higher and the median household income slightly lower. Thus, Bushwick appeared more on a par with Williamsburg, in socioeconomic terms, than was Sunset Park/Windsor Terrace. Perhaps most important to my original research design, many of my expert informants indicated that Bushwick's neighborhood revitalization efforts were largely “unsuccessful.” Informants described Bushwick as being hopelessly factionalized, mired in political corruption, and generally the type of place where accomplishing anything at all was a monumental struggle. They made various negative statements about Bushwick, including claims that the local school district was by far the worst in the city in terms of both operations and performance; that social service programs sucked in money without producing any positive outcomes; and that very few competent community organizations existed.
(p.247) Based on all of the available evidence, then, I selected Williamsburg and Bushwick as my study neighborhoods. At the time, the two areas had the most similar demographic profiles of any of the potential study areas. Both had between 60 percent and 65 percent Latino populations, poverty rates of between 35 percent and 40 percent, and median household incomes between sixteen thousand and eighteen thousand dollars. The most important demographic difference between the neighborhoods was racial: the two areas have approximately inverse proportions (about 30 percent and 10 percent) of whites and African Americans. The two neighborhoods are contiguous, both lie within the borough of Brooklyn, and they have four district-based political representatives in common. And while Williamsburg was widely seen as successful in neighborhood revitalization, Bushwick was not.
Choosing the Study Organizations
I had already selected the four organizational types for the study, so my first step in choosing the specific organizations where I would conduct participant-observation was to enumerate the organizations of each type that existed in Williamsburg and Bushwick. I already had collected the names of several organizations in each area from my initial round of interviews. I contacted these organizations to set up interviews with their executive directors in order to get a more comprehensive, insider's view of the local organizations. Prior fieldwork experience had taught me that getting a response from the overworked staffs of community-based organizations generally is better achieved through a personal visit than a telephone call. This strategy was at least partially confirmed, as I was able to schedule immediate appointments in all of the organizations. Something about being seen as an actual body rather than simply a voice on the telephone seems to make the request for an interview more tangible. Although I had been worried that my identity as a student from a university outside New York might be a liability, in fact it turned out to be something of an advantage. People seemed flattered that I had come “all the way from Chicago” to learn about their organization and their neighborhood.
When I met the executive directors to conduct my interviews, I identified myself as a doctoral student from the University of Chicago and said I was doing a research project on community development and community building in Williamsburg and Bushwick. I asked them to describe the work of their organization, their targeted population, and anything else they felt was relevant about their neighborhoods. I also asked these contacts to (p.248) name other organizations in the area that were doing “good work in the community.” (Williamsburg organization representatives were asked to name other Williamsburg organizations, and Bushwick organization representatives to name other Bushwick organizations.) In addition to allowing interview respondents to generate their own lists of organizations, I specifically asked them to name community development corporations, child care centers, and Catholic churches. I already knew that each neighborhood had only one Beacon, and so did not specifically inquire about this fourth type of organization. I supplemented the information gathered in these interviews with newspaper articles, academic studies, government agency documents, and other archival material.
In Williamsburg, I identified two CDCs, four child care centers,11 four Catholic churches, and one Beacon. In Bushwick, there were two CDCs, five child care centers, three Catholic churches, and one Beacon. My research design called for selecting the most active organization of each type for my study. This decision was based on the overall direction of the project, to focus on the role of active resident and organizational participation in neighborhood revitalization efforts. Restricted time and resources, as well as limited past research, made it appropriate to seek out information on the most effective improvement work occurring in each study area. I made my final selection of the eight organizations where I would conduct my participant-observation after discussions with the growing number of neighborhood informants whom I was meeting with each successive day in the field. All eight organizations that I approached were generous enough to allow me to conduct my research there. The Williamsburg organizations were the Southside United Housing Development Fund Corporation (Los Sures), Nuestros Niños Child Development Center, Transfiguration Catholic Church, and the Williamsburg Beacon Center. The Bushwick organizations were the Ridgewood-Bushwick Senior Citizens Council Housing Office, New Life Child Development Center, Saint Barbara's Catholic Church, and the I.S. 111 Beacon.
Method: Doing Participant-Observation
The participant-observation data reported in this book come from field-work conducted over the course of three years, from May 1997 to September 2000. The first sixteen months of this period were the most intensive. (p.249) During this period I performed between six and nine months of weekly volunteer work at each of the eight study organizations. My stints at the various CBOs were staggered over the course of these months, but I maintained ongoing, regular contact with my informants at the organizations where I had worked early in this sixteen-month period as I completed my volunteering in the remaining organizations. In addition to my regular assignment at each organization, I took part in numerous other aspects of community life, attending community meetings, social gatherings, and political rallies, visiting people's homes, and participating in general street life.
I established my role as a volunteer in each organization through the same process. I first visited each group to request a meeting with the executive director.12 Whenever possible, I used a referral from someone I had already interviewed formally or met casually in the neighborhood. At the first meeting with the executive director, I conducted an open-ended interview about his or her organization's history, programs, and achievements. These interviews lasted from forty-five minutes to three hours. I concluded the interviews by offering my services as a volunteer for half a day each week, for a period of six months. I placed no restrictions on the type of work I was willing to do, except to ask that I be given a volunteer assignment that would put me in contact with neighborhood residents who were involved in the organization's work. Most of the organizations responded enthusiastically to my offer of volunteer work. Only one executive director expressed some reservations about my participation, mostly centered on concerns that with research as my main focus, I would not be a reliable volunteer.13 After assuring him that this would not be the case, he agreed to locate a volunteer assignment for me within the organization.
The underlying rationale for my volunteer strategy was the notion of reciprocal exchange. On the one hand, I hoped that by offering something of value—my commitment of labor—to the organizations, and particularly to the staff members with whom I worked, I might encourage a favorable perception of me. In turn, I hoped this would facilitate greater access to and candid discussion about the organization and the neighborhood. At the same time, I recognized the value of the organizations' allowing me to (p.250) conduct my research there and wanted to give something in return. I was aware of many cases in which CBOs had opened their doors to researchers, then received nothing for their cooperation—not even the courtesy of a report on the research results. I wanted to avoid this situation by contributing something to the organizations immediately. I also gave each organization a copy of my dissertation when it was completed, and I intend to do the same with this book when it is in print.
All of the organizations' executive directors except one referred me to another staff member to set up my volunteer assignment. This staff member usually became my closest informant within each organization.14 I engaged in a wide variety of tasks, ranging from laying the groundwork for the opening of a new child care facility, to documenting housing code violations in tenants' apartments, to tutoring reading and English, to writing funding proposals. I also made myself generally helpful whenever I was at the organizations, pitching in with ad hoc tasks like making copies, stuffing envelopes, finding materials, cleaning up, or entertaining children whose parents had business at the organization. As I conducted this work, I spent many hours discussing the organizations and the communities with my main informants. Hanging around with them in turn facilitated numerous conversations with other staff members, community residents, staff at other local organizations, and key political and community leaders. My main informants also steered me toward an array of community events, which provided further important observational opportunities.
At each organization, I maintained productive fieldwork relationships with at least four staff members. All of the staff members were aware of my status as a researcher, as well as of my volunteer work. Generally, for the first month or two of the fieldwork period these informants provided me with basic information on organizational and community issues. They hesitated to discuss the deeper meanings of these issues, however, particularly the sensitive ones. They often asked if they were being tape-recorded, if they were going to be quoted, or if what they said would “be in the book.” I assured them that they were not being recorded, and that, whenever they requested it, what they said would not be attributed to them. After I had spent some time at the organizations, informants appeared to gain some confidence in me and for the most part stopped asking these kinds of questions. They also became more willing to discuss the sensitive issues and to provide me with their own analyses of the factors underlying events that I observed.
(p.251) The importance of my being a reliable and accommodating volunteer during this process of confidence building cannot be overstated. The simple fact that I consistently showed up for work on time and was willing to do whatever tasks the staff requested of me went a long way toward ensuring that I was seen in a positive light. At numerous times during the course of the participant-observation, and especially in its latter stages, informants mentioned that they had been skeptical of me at the beginning but had come to see that I followed through on my commitments. Several informants explained their early skepticism by contrasting my behavior with past negative experiences they had had with researchers or volunteers. As Mitchell Duneier (2000) reminds us, however, informants' acknowledgment of my reliability as a volunteer should not be confused with some holistic notion of “trust.”
My initial fieldwork interactions were largely open-ended, unfocused conversations in which I learned a wide variety of things about each organization's work, as well as about the history, politics, organizations, leaders, and issues in each of the study neighborhoods. As the fieldwork progressed, and important themes relating to my research questions emerged, I tended to ask more focused questions of my informants, guiding the conversations more than I had previously (Duneier 2000). After sixteen months of volunteering and conducting intensive participant-observation, I moved into a second stage of data collection, in which I conducted eighty formal interviews with organization staff members and participants. During the four months that it took to complete these interviews, I maintained regular though unstructured contact with staff members at all of the organizations and continued to attend many organizational and neighborhood events.
In February 1999, I suspended data collection in Williamsburg and Bushwick for about eight months, returning the following September as part of a new research project on second-generation immigrants in New York City (Kasinitz, Mollenkopf, and Waters 2004; Marwell 2004a). I resumed participant-observation for the next eleven months, during which time I did not have any regular volunteer assignments but rather drew on my status as a known person in the two neighborhoods to circulate freely among organizational events, gatherings in people's homes, and the general flow of daily life inside the organizations and on the street. As I continued to pursue my earlier interest in the role of community-based organizations in neighborhood revitalization, I also paid particular attention to the ways in which the organizations were involving the second generation of Dominican immigrants in their work. In addition to my fieldwork in Williamsburg and Bushwick, I conducted participant-observation in three (p.252) community-based organizations in the Washington Heights neighborhood of Manhattan during this period.
Throughout the entire time I conducted participant-observation, I kept detailed narrative field notes on my experiences, documenting both events and conversations. I wrote these notes as soon as possible following fieldwork experiences, sometimes drawing on short notes I had been able to take down in the field. I also regularly made entries in a separate analytical file, writing about my own interpretations of the fieldwork, raising questions about the meaning of my experiences, and developing emerging themes of the research. I wrote these analytical notes more sporadically, as ideas and themes emerged over the course of the fieldwork. Throughout the preceding chapters, much of the evidence I present in support of my arguments comes from material recorded in my field notes. I present this material both in the general narrative of the book and in actual excerpts from my field notes. All such instances are cited in the footnotes, which indicate the date on which the events in question took place.
On My Use of Real Names
It has long been standard practice among ethnographers to use pseudonyms and otherwise mask the real names of research sites and informants, with the stated objective of protecting informants' confidentiality. In recent years, however, this practice has come under discussion, and some ethnographers have provided identifying information in their work (e.g., Duneier 2000; Kasinitz 1992). The most explicit rationale and strategy for identifying places and people by their real names comes from Mitchell Duneier. In his acclaimed book Sidewalk, Duneier argues that scholars cannot continue to appeal solely to personal and professional ethics when asking readers to trust their accounts and interpretations of events in the field. In light of recent scandals involving major newspapers' printing stories fabricated by journalists (e.g., Barry et al. 2003; Pogrebin 1998), Duneier asks why ethnographers should be viewed any less skeptically. His own answer to this question was to use both real names and photographs of many of the research subjects in Sidewalk, as well as to employ an innovative technique for obtaining the “informed consent” required by universities and the federal government when working with human subjects who might be harmed by research activities.
In this book, I have replicated components of Duneier's model, giving the real names of the neighborhoods, organizations, and people featured in its pages. Although in prior publications (Marwell 2004a, 2004b) covering (p.253) earlier phases in this research I followed the traditional practice of using pseudonyms, it seemed impossible to do so once the research was completed and written in its present form. The historical material on Williamsburg and Bushwick, reported mostly in chapters 2 and 3, constitutes a critical part of my argument. Key details that render the neighborhoods and organizations identifiable could not be left out without undermining the analysis. This state of affairs, however, did not absolve me of the necessity of obtaining informed consent from the people who appear in the book; I engaged in a strategy similar to Duneier's in order to do so.
During my writing process, I sat down with each person who is named in the book and portrayed as either doing something or saying something that I observed. I gave them a printed copy of the passages in which they appear, sat with them as they read through the pages, and gave them the choice of remaining a named actor or being made anonymous. Somewhat to my surprise, only a very few passages prompted actors to request anonymity; I complied with all of these requests. There were two exceptions to this practice: One was for people who spoke at public meetings, such as community board meetings or city council hearings, whose comments are thus on the public record. The other was for individuals considered to be “public figures,” who do not have an expectation of anonymity as regards their public duties; these were mainly elected officials. Although I did not review with the public figures the specific passages of the book that describe their actions and utterances, all of these individuals knew who I was and were aware that I was carrying out research. In addition to the protections of informed consent given by the strategy described here, this research was examined and approved by the institutional review boards of the University of Chicago and the City University of New York.15
Access and Subjectivity in the Field
Practitioners of participant-observation continually grapple with the fragility of access to their research settings. In part, access turns on a researcher's success at convincing potential research subjects to allow him or her to “hang out” in relevant settings, thereby enabling data collection. Another point of vulnerability lies within the researcher's own “subject position”—that (p.254) is, in his or her own identity, in the ways that identity informs how the researcher understands events in the field, and in the process by which the putative research subjects themselves construct the researcher as a subject.
Numerous discussions of how to “gain access” to fieldwork settings in the first of these two senses already exist in the literature (e.g., Lofland and Lofland 1995; Schatzman and Strauss 1973). One particular barrier, however, has not received much comment: conflict among research subjects. Gaining and maintaining a researcher role on both sides of a conflict requires fraught judgments about how each side will react to the researcher's spending time on the other side. Will doing so prompt accusations that the researcher is a spy? Will access to one side be cut off once the researcher is known to have access to the other side? In most instances of conflict that arose during this research, I was able to carry out participant-observation on both sides. These instances included the ongoing and flaring tensions between Los Sures and Assemblyman Vito Lopez; the disagreements between staff at Los Sures and the Nuestros Niños Child Development Center over tactics for creating social change; the longstanding antipathy between Saint Barbara's Catholic Church and the Ridgewood-Bushwick Senior Citizens Council; the internal dispute between parents and staff at the Williamsburg Beacon; and the disdain in which certain leaders at the New Life Child Development Center held Ridgewood-Bushwick and Assemblyman Lopez.
In one instance of conflict discussed in this book, I was unable to conduct research on both sides: the battle between Williamsburg's Latinos and Hasidic Jews over affordable housing resources, recounted in chapter 2. When I began fieldwork at the four Williamsburg organizations, all of them (by design) substantially representing the interests of the neighborhood's Latinos, I was unaware of the Latino-Hasidic conflict. As I learned about its content, longevity, and depth from the Latino organizations, I felt it would be impossible to conduct fieldwork with the Hasidim, for two distinct reasons. First, I was pessimistic that the leaders of the Hasidic organizations, all Hasidic men, would allow a female fieldworker into their midst. Prohibitions on cross-gender social interactions are extensive among the Hasidim, and it seemed very unlikely that I would be able to surmount them. Although female staff members of government officials have been able to work with male Hasidic organizational leaders, these women have had the advantage of controlling access to public resources that the Hasidic organizations are seeking. I had no such standing. A second reason for not attempting fieldwork among the Hasidim was rooted in my already existing relationships with the Latino organizations. The tensions between (p.255) the Latino and Hasidic organizations—especially at Los Sures—were so high that I was convinced I would jeopardize my continuing access to the Latino organizations if I attempted fieldwork in the Hasidic organizations. All of the information on Hasidic organizations and interests that I present in this book, then, comes from others' published accounts or my own observation of Hasidic individuals and organizational leaders in public settings.
In all of the other instances of conflict, I was forthcoming with all parties about the fact that my role was that of a researcher, and that I was spending time “hanging out” on both sides of what I had learned was a conflict. I was extremely vigilant about not repeating to people on one side of a conflict what someone on the other side had told me. Most everyone seemed to respect my position, as evidenced by the fact that almost nobody asked me for information about what was going on on the other side of a conflict. On the very few occasions when I did receive such requests, I reiterated my understanding of my role as a researcher and explained that I did not feel comfortable providing such information. This was usually sufficient to resolve the immediate situation and to preempt future requests of this sort.
In addition to navigating these and other potential barriers to basic access to research settings, my own subject position fundamentally influenced the scope of my observations and the findings to which they led. As George Herbert Mead and Erving Goffman pointed out many years ago, our individual identities are always a social encounter—a “presentation of self” to others (Goffman 1959; Mead 1956). This presentation, or performance, of identity contains both conscious and unconscious dimensions. In addition, our own identity performances elicit responses in others, who simultaneously construct identities for us on the basis of their own characteristics and experiences. Certainly, substantial overlap exists between our own sense of our identity and others' sense of our identity, but there is divergence as well.
While doing fieldwork, I was conscious of performing certain of my personal characteristics differently than I did in my “regular” (that is, non-fieldwork) life. This is not to say that I lied about those characteristics—only that they became more salient to my identity when I was positioned as a researcher in the field. At the time of my research, I had the following key personal characteristics: fourth-generation U.S. American, upper middle class, white, Jewish (nonobservant), Spanish-speaking, student, young, heterosexual, and female. Some of these were easily visible, others discernible during conversation, and others largely hidden from view. Each one affected my research, more so even than I could possibly know. I think that my age, student status, gender, and ability to speak Spanish fluently (p.256) and without an “American” accent proved most important to the process of my fieldwork.
My relative youth (I was twenty-nine when I began the research) and my identity as a student helped legitimate my usual approach to the organizations where I hoped to conduct fieldwork. When I first introduced myself to organizational leaders and participants, I told them (truthfully) that I was a student interested in finding out as much as possible about the people, organizations, and neighborhoods of Williamsburg and Bushwick. The “student” role proved to be one that most everybody understood, and people often said they were happy to be helping me out with my “school project.” My age reinforced my student status, and both characteristics seemed to contribute to a sense that since I had so much to learn about the neighborhoods and organizations, it was appropriate for me to be hanging around a lot. I am acutely interested in how I will establish a role in a new participant-observation project when I no longer have the young student role as an option.
Being female created both opportunities and constraints during the research. As Nancy Naples has argued, female ethnographers can gain a kind of access to the experiences of women that male ethnographers cannot (Naples 1998). Since many of the Williamsburg and Bushwick organizations I studied had largely female staffs, my gender afforded me a certain ease throughout much of my fieldwork. It was not seen as peculiar or threatening for me to spend time alone with other women; to engage in cooking, child care, or other tasks associated with nurturing; or to be interested in helping others through community work. Being female was the usual condition in the settings where I conducted most of my research, although a number of the highest-ranking staff members at the organizations were male. My gender did not pose a barrier to frequent interactions with these men, but our interactions did tend to be of shorter duration and more focused on a particular topic—in contrast to the often unstructured, undirected time I spent in the company of women.
While the gender-based advantages I obtained in my fieldwork in fact seem to me more difficult to identify and enumerate, the gender-based constraints I faced are (and were) very clear. Nothing delimits the possibilities of one's own identity and experience like running up against a barrier. As I have already noted, I presumed that my gender would have proven an insurmountable barrier to an attempt to conduct fieldwork with the Hasidim. One gender-based barrier that was apparent stemmed from my own and others' perceptions about my personal safety while in the field. When I was a beginning graduate student contemplating the undertaking of a (p.257) participant-observation project in a poor neighborhood, a very senior sociologist expressed to me his concerns about whether I, as a young woman, would be safe in such an environment. He suggested that rather than exposing myself too often to the dangers of the streets, I might be better off doing something that he described as akin to Method acting. From the safety of my own room, with little else but a mirror and my imagination, I could enact some of the experience of doing fieldwork an approximation of the “feel” of the poor community I hoped to study.
I was very much confused by this odd, if well-intentioned, suggestion. I went ahead and did fieldwork the old-fashioned way despite it. But the encounter itself is food for thought with regard to my gender and its role in my urban ethnographic project. Most of the “classic” works in the still-solidifying urban ethnography canon aimed to investigate the deviance and social problems often associated with city life in the sociological imagination: slums, crime, poverty, unemployment, gangs, “oppositional culture,” and so on.16 Most (though not all) of the authors of these works have been men, a fact that might prompt us to ask why maleness has been so strongly associated with these kinds of questions. Perhaps male researchers, as part of their own subject position, consistently identify these issues as the preferred, somehow most legitimate, province of urban ethnographic inquiry. And perhaps this equation underlay the erstwhile advice that senior sociologist gave me those many years ago.
Of course, there are also works of urban ethnography, some of them “classics,” that do not focus explicitly on deviance and social problems. Some have been “community studies” in the Chicago School sense, attempting to gain an understanding of the entire social milieu of a particular urban neighborhood (Drake and Cayton 1945; Kornblum 1974; Molotch 1972; Pattillo 1999; Susser 1982). Others—including my own—have been animated by questions about poverty and poor neighborhoods but have explored different dimensions of the experience of the urban poor, such as family life, community formation, religion, the economic demands of motherhood, and activism (e.g., McRoberts 2003; Naples 1998; Newman 1999; Small 2004; Stack 1974). Urban ethnographies by women have been more likely to examine these latter kinds of issues, rather than choosing deviance as their starting point. My own research interest in neighborhood revitalization thus may have been connected to my gender from the start, (p.258) while the setting of my fieldwork—inside formal organizations—may have offered some relief for gendered anxieties over how to gain entry into an unfamiliar place where I felt my safety might indeed be an issue.
Ethnography and folk wisdom alike point to the risks of unfamiliarity in settings where social order is shaky (as indicated, for example, by higher-than-average crime rates) or governed by highly parochial rules (as in Gerald Suttles's  concept of the “defended neighborhood”). To be unknown in such settings is to be a potential threat, a status that can easily make one a target.17 A common strategy among urban ethnographers is to acquire a “sponsor” who can provide an introduction into the field setting: someone who has an accepted place in the social world of interest. At worst, this helps to reduce the ethnographer's unfamiliarity; at best, it conveys legitimacy to him or her. My own fieldwork took place largely (though not entirely) inside formal organizations. This decision derived primarily from my overarching interest in neighborhood revitalization, a process in which community-based organizations are central actors, but it also offered me a shortcut to reducing my unfamiliarity. The organizations I studied provided controlled settings with regular participants and (largely) benign, predictable rules of social order. Ethnographers who choose to conduct fieldwork in more informal settings often need to negotiate a much wider and less predictable range of social situations. While women have done and are continuing to do fieldwork in such settings (e.g., Jones 2004; Maher 1997; Moore 1978), I suspect that gendered forms of risk play an important role in the distribution of men and women over ethnographic research questions, and thus over fieldwork settings.
Finally, the importance of language during my research cannot be understated. The first language of many residents of Williamsburg and Bushwick, including many staff members and participants at the organizations I studied, was Spanish. Some spoke little or no English, while others spoke English well. I spoke exclusively in Spanish with some of the people I met in the field, in a combination of Spanish and English with others, and exclusively in English with others. My Spanish has a not-quite-identifiable Caribbean accent, a sort of mishmash of the clearly distinctive Puerto Rican, Dominican, and Cuban pronunciations. Being able to speak Spanish, and this particular kind of Spanish, not only facilitated communication with organizational staff and neighborhood residents, but also complicated their (p.259) assumptions and perceptions regarding my subjectivity in ways that perhaps made me seem less of an outsider. Upon hearing me speak Spanish, many people assumed that I had Latina, probably Caribbean, heritage. Numerous times, some months after an initial casual meeting, a person would be surprised to learn that I did not in fact come from this background.
My language skills thus often took precedence over the visual aspects of my privileged location in the complex arrangements of racial and ethnic identity in the United States. Given the racial diversity and ambiguity of those classified as “Latino” in this country, it was perfectly conceivable that a person of my appearance—light skin, green eyes, reddish hair—could be Cuban, Puerto Rican, Dominican, or of some other Latina ethnicity. The fact that I spoke fluent, Caribbean-accented Spanish enhanced this possibility. This haziness about my ethnicity often remained even for people with whom I spoke mostly or entirely in English, since they had usually heard me speaking to others in Spanish. Language thus offered me a dimension of plasticity in my subject position, which, when combined with my other characteristics, allowed me, in certain ways, to “pass” as an insider. At times, a person in the field would ask me directly about my ethnic roots. In the early months of my research, this question left me rather flummoxed, as my ethnic roots were not a salient part of my own identity formation, and I did not have a comfortable response. I usually said something about being “from here”—that is, born in the United States—and that I did not have any Latina heritage. I quickly learned that this was not a satisfactory answer for most of my questioners, and began to offer the only other category I could think of: that my family was of Jewish origin, but not religious. I do not know what response this answer provoked in my questioners, but it usually proved sufficient to bring the conversation on this topic to a close.
These comments only begin to scratch the surface of the complex and thorny issues surrounding subject position, its effects on fieldwork practice, and the ways it conditions the findings of participant-observation research. I offer them here largely to give the reader additional insight into my research process. Participant-observers whose practice includes reflection on their own subjectivity confront a series of important choices at all stages of their work: data collection, data analysis, and data reporting. At one extreme lies the conviction that subjectivity is so decisive that we cannot make claims about anything in a fieldwork project besides ourselves and our own experiences. At the other extreme lurks a vision of the researcher's detached, technical relation to the object of analysis. Bargaining for Brooklyn is ultimately my own bargain with my research subjects, my readers, and myself. (p.260)
(1.) See, e.g., Arlene Dávila (2004), De Genova (2005), Dohan (2003), Freidenberg (2000), Hondagneu-Sotelo (2001), Mele (2001), Menjíivar (2000), Pérez (2004), Ramos-Zayas (2003), and R. C. Smith (2006).
(2.) Some classic examples of urban ethnographies focusing on African Americans include Drake and Cayton (1945), Liebow (1964), Hannerz (1969), Rainwater (1970), Stack (1974), E. Anderson (1978), and Duneier (1992). Prior to 2000, only a few ethnographic studies examining (partially or exclusively) the experience of Latinos in U.S. urban settings had been published: Bourgois (1996), Moore (1978), P. C. Sexton (1966), Susser (1982), Suttles (1968), von Hassell (1996).
(3.) Mott Haven (Bronx CD 1, 66.9 percent Latino); Hunts Point (Bronx CD 2, 78.9 percent Latino); Morris Heights/University Heights (Bronx CD 5, 56.8 percent Latino); EastTremont/ West Farms (Bronx CD 6, 58.8 percent Latino); Kingsbridge/Bedford Park (Bronx CD 7, 50.6 percent Latino); Soundview/Unionport (Bronx CD 9, 53.5 percent Latino); the Lower East Side (Manhattan CD 3, 32.3 percent Latino); East Harlem (Manhattan CD 11, 51.9 percent Latino); Washington Heights/Inwood (Manhattan CD 12, 67 percent Latino); lackson Heights/North Corona (Queens CD 3, 43.7 percent Latino); Elmhurst/South Corona (Queens CD 4, 41.8 percent Latino); Bushwick (Brooklyn CD 4, 65 percent Latino); and Sunset Park/Windsor Terrace (Brooklyn CD 7, 51.4 percent Latino).
(4.) Williamsburg-Greenpoint (Brooklyn CD 1, 43.6 percent Latino).
(5.) I conducted interviews with Eric Brettschneider (Agenda for Children Tomorrow, New York City Administration for Children's Services), Dena Davis (community development consultant), Mariana Gaston (New York director, Resolving Conflict Creatively Program), Marilyn Gittell (professor of political science, City University of New York Graduate Center), Marc Jahr (New York City director, Local Initiatives Support Corporation), Patricia Jenny (director, Neighborhood Strategies Project, New York Community Trust), Phillip Kasinitz (professor of sociology, Hunter College and the City University of New York Graduate Center), Melvin Oliver (program officer, Asset-Building and Community Development Program, Ford Foundation), Aida Rodriguez (program officer, Equal Opportunity Program, Rockefeller Foundation), Robert Sherman (program officer, Surdna Foundation), Susan Saegert (professor of environmental psychology, Housing Environments Research Group, City University of New York Graduate Center), April Tyler (community activist and former member of New York City In-Rem Task Force), Avis Vidal (director, Community Development Research Center, New School for Social Research), and Louis Winnick (Fund for the City of New York and professor emeritus, New School for Social Research).
(6.) Although I did not realize it at the time, my “expert” informants' own definitions of what constitutes “success” in community revitalization proved to be extremely important to their characterization of the potential study neighborhoods. As a result, although I began the research with the belief that I was contrasting a “successful” and an “unsuccessful” neighborhood, I abandoned this notion about midway through the project. A reader will note that the “successful” versus “unsuccessful” dichotomy does not appear in this book. For a discussion of this aspect of the research, see Marwell (2000).
(7.) Field notes on interview with Professor Susan Saegert, Department of Environmental Psychology, City University of New York, February 3, 1997.
(8.) Field notes on interview with Marc Jahr, New York City Director, Local Initiatives Support Corporation, February 10, 1997.
(9.) Field notes on interview with Patricia Jenny, Director, Neighborhood Strategies Project, New York Community Trust, February 28, 1997.
(10.) Coincidentally, 1 myself lived in the Soundview/Parkchester section of the Bronx during the time I was conducting the participant-observation for this book. I had moved to that area (corner of Castle Hill Avenue and Chatterton Avenue) for several reasons, among them the idea that my research sites might turn out to be in the Bronx, the city's only majority Latino borough. Although I ultimately conducted the research in Brooklyn, the fact that I returned from the field each night to a neighborhood quite similar in many respects to the study neighborhoods lessened the sense of fieldworker dislocation that might have occurred had I been living in, for example, an upper middle-class, largely white neighborhood like the one where I live now.
(11.) In both Williamsburg and Bushwick, I inquired only about child care centers that served low-income families and had significant or majority populations of Latino children.
(12.) While “executive director” was not the title of all of the organization heads with whom I met—for example, at the two Catholic churches, this initial meeting was with the pastor—for ease of expression, I apply it to all the organizations.
(13.) Field notes on interview with Michael Rochford, executive director, Saint Nicholas Neighborhood Preservation Corporation (administering agency of the Williamsburg Beacon), June 13, 1997.
(14.) One executive director oversaw my volunteer work himself, and he became my main informant at that organization.
(15.) The University of Chicago Institutional Review Board approved the first phase of my fieldwork (May 1997 to February 1999), on which my dissertation (Marwell 2000) was based. The fieldwork conducted between September 1999 and August 2000, under the auspices of the Second Generation in Metropolitan New York project, was authorized under that project's examination by the Institutional Review Board of the City University of New York.
(16.) See E. Anderson (1978, 1990)), Bourgois (1996), Duneier ( 1992, 2000), Cans (1964), Hannerz (1969), Liebow (1964), Moore (1978), Sanchez-Jankowski (1991), Stack (1974), Suttles (1968), Venkatesh (2000), and Whyte (1943).