Field Methods http://fmx.sagepub.com/
Audio Recording Everyday Talk Rosalyn Negrón Field Methods published online 18 January 2012 DOI: 10.1177/1525822X11432082 The online version of this article can be found at: http://fmx.sagepub.com/content/early/2012/01/06/1525822X11432082
Published by: http://www.sagepublications.com
Additional services and information for Field Methods can be found at: Email Alerts: http://fmx.sagepub.com/cgi/alerts Subscriptions: http://fmx.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav
>> OnlineFirst Version of Record - Jan 18, 2012 What is This?
Downloaded from fmx.sagepub.com at UNIV OF FLORIDA Smathers Libraries on February 4, 2012
Audio Recording Everyday Talk
Field Methods 000(00) 1-18 ª The Author(s) 2012 Reprints and permission: sagepub.com/journalsPermissions.nav DOI: 10.1177/1525822X11432082 http://fm.sagepub.com
Rosalyn Negro´n1
Abstract This article covers a number of strategies for audio recording everyday, naturally occurring interactions. Drawing on fieldwork where the author observed and recorded eleven key informants in New York City, the author provides a guide applicable to a number of research situations, but especially relevant to studies that require continuous observations—from unobtrusive to participant—across various social contexts. The author provides practical suggestions for choosing an audio recorder, recording in the field, managing informed consent issues, and addressing observer effects. Keywords audio recording, ethnography, field research, observation, naturally occurring interaction
Introduction Ethnographic and linguistic research on naturally occurring talk has always had to contend with important technical and methodological challenges: capturing the speech of speakers on the move and in various contexts; choosing among recorder options; ensuring audio quality; faithfully recording phonological and prosodic subtleties; addressing observer effects; and 1
Department of Anthropology, University of Massachusetts, Boston, MA, USA
Corresponding Author: Rosalyn Negro´n, Department of Anthropology, University of Massachusetts, 100 Morrissey Boulevard, Boston, MA 02125, USA Email:
[email protected]
Downloaded from fmx.sagepub.com at UNIV OF FLORIDA Smathers Libraries on February 4, 2012
2
Field Methods 000(00)
working with limited resources. There certainly are resources to help researchers decide what recording equipment to use (Kolovos 2007; Modaff and Modaff 2000; Plichta 2010; Stockdale 2002). But with a few exceptions—see, for example, Goodwin’s (1993) thorough but now technologically outdated guide and an appendix in Duranti’s (1997) linguistic anthropology textbook—sociolinguists and field ethnographers tend to omit meatier details about their audio recording tools and practices. Surely, many of the most trailblazing studies on human interaction in natural settings have been done using rather conventional audio recording techniques. Undoubtedly, other studies never see their full potential because they applied conventional techniques to research questions that called for more. Innovations in audio recording technology are expanding what researchers can imagine as possible, pushing methodological and theoretical boundaries. At the same time, there is greater awareness and acceptance of life exposure among would-be research informants. Reality television, the mainstream success of documentary films, and Web 2.0 technologies in which users display various aspects of their personal lives for public viewing have made it less threatening, and, indeed, desirable to expose mundane aspects of private lives. This article covers some technologies and techniques useful for studying mundane, naturally occurring interactions. I share my experiences doing fieldwork in New York City (NYC), where I continuously monitored and audio recorded 11 key informants as they went about their daily lives. This practical guide is applicable to a number of research situations, but is especially relevant to studies that require continuous observations—from unobtrusive to participant—across various social contexts. Whether you are studying people’s routine comings and goings in a particular community or focusing on specific behaviors, you will find that audio recording is a powerful means of documenting the ethnographic experience. Besides being a primary means of data collection for linguistic analysis, audio recordings can be used to complement field notes and as a durable and easily transferable record of fieldwork for multimedia use in presentations or the classroom. Audio recordings also offer alternatives to relying on people’s retrospective self-reports about their behavior in natural social settings. Significantly, audio recordings are a multipurpose data source with which to pursue multiple research questions. One day’s recording can yield rich information for a soundscape ethnography, a network analysis of a person’s typical interactions, user research on cell phone use, or a code-switching study. Such versatile, fertile data are possible with careful
Downloaded from fmx.sagepub.com at UNIV OF FLORIDA Smathers Libraries on February 4, 2012
Negro´n
3
consideration of audio recorder options, recording strategies, informed consent issues, and observer effects.
Studying Everyday Negotiations of Ethnicity During a six-month period in 2006, I conducted ethnographic research on the ways that NYC Latinos used linguistic resources to negotiate multiple ethnicities (Negro´n 2007, 2011). Digital audio recording was critical for capturing informants’ linguistic and discursive acts—including codeswitching, style shifting, accent changes, and declarations of ethnic disaffiliation/affiliation. The research was designed to cast a wide net to capture the many, sometimes subtle, ways that people made ethnicity relevant in their daily interactions and across social domains. To achieve this, I shadowed eleven Latino participants for one week as they went about their daily routines. While I observed each participant in their daily activities, they wore a small, portable digital voice recorder that stored their verbal interactions. Immediately after the first week of observations, each person agreed to do another week of independent recording. I selected small (4" 1½"), high end, lightweight recorders that fit into cell phone cases for clipping on to participants’ pant waists or pockets.1 Remote control clip on microphones were attached to the recorders and clipped on to shirt pockets, collars, or lapels. The remote control feature on the particular microphone model that I used2 was invaluable to the participants. All became adept at pausing or stopping recordings on the fly, for example, during trips to the bathroom, when needing privacy, or during silent moments of activity not discernable in the audio recording. However, a consequence of having these recorders on for hours at a time was that it remained on, unnoticed and forgotten, even during bathroom trips and idle points. I explained to each person that the purpose of the research was to document their everyday lives and to better understand how ethnicity figured into various aspects of their daily life. To minimize self-monitoring of speech, I did not bring special attention to the sociolinguistic dimensions of my research. In the first week, we scheduled 5–8-hour periods of daily intensive observations. These scheduled observations took place in every area of their life that they felt comfortable with me observing. Domains included work, school, at-home interactions, church, shopping, and recreational social gatherings. Each person was free to turn off the recorder whenever they wanted to. During these observations, I always carried a small notebook, as field notes were metadata for the audio recordings.
Downloaded from fmx.sagepub.com at UNIV OF FLORIDA Smathers Libraries on February 4, 2012
4
Field Methods 000(00)
The second week was designed to capture informants’ conversations without my presence. Typically, I met informants before they began their day to save the previous day’s recordings onto my laptop. It took about 5 minutes to transfer .WMA files to my computer where they would be stored according to the day and time that each file was first recorded, at the precise time that the informant pressed the record button. Now, because informants sometimes stopped the recorder at various times during the day, multiple audio files were created for most days of recording. These files could last from one minute to several hours, or even seconds if, for example, informants accidently pressed the recorder’s stop button. As necessary, I provided either another empty recorder with fresh AAA batteries or batteries for the recorder that they were already carrying. Since the recorder I used had a memory capacity of 8 hours 56 minutes for high-quality audio,3 it had sufficient space for a day of uninterrupted recording or two days of shorter recordings. Based on recording times, the 11 key informants collected a total of 288 hours of recordings. This was about the same amount of time I spent observing them in the first week (290 hours), for a total of 578 hours of quality audio. Because part of my research plan targeted specific instances of ethnic identity switching— which for most people and contexts does not happen frequently—I required a large data set from which to mine exemplary cases. Before embarking on the research, consider your data needs and plan accordingly. Your data needs should also determine what audio recorder you use.
Audio Recording Equipment Selecting Audio Recording Equipment Before selecting equipment, consider your priorities for the research. Whether you will focus on one focal informant or not, it is just as important to capture that person’s voice, as it is to capture the voices of others in an interaction. This means that recording equipment should be adequate for recording multiple voices at various distances from the recorder. Beyond that, the possibilities and choices will depend on your research needs. Variationist sociolinguistic or phonetic research on everyday talk would demand equipment sensitive enough to faithfully capture the presence (or absence) of phonological variables in a key informant’s speech. In that respect, there is more flexibility when studying language at the discourse level. Or consider if your research calls for collecting continuous interactions in a range of social domains. This would require highly portable,
Downloaded from fmx.sagepub.com at UNIV OF FLORIDA Smathers Libraries on February 4, 2012
Negro´n
5
sturdy equipment with great recording storage capacity. On the other hand, research on events with a more defined beginning and end in one set location—town hall meetings, boardroom discussions, tenant meetings—may call for less portability and more excellent multidirectional recording capabilities. Stockdale (2002) provides a useful guide with other factors to consider when selecting audio recorders. The following discussion of equipment is based on the options available for flash memory digital audio recording. Flash memory digital audio recorders have clear advantages over tape recorders (sound quality, durability of audio files, recording times) and some important advantages over direct-to-CD recorders (convenience, storage, and file organization). To help narrow down your equipment choices, consider portability, ease of use, sound quality, and versatility. Regarding portability, I suggest having informants carry the recorder around themselves. Besides the obvious benefit of having the recorder close to the main audio source at all times, informants will be able to take more control over the data collection process. In addition, a user-friendly, robust, and portable recorder enables ethnographers to focus observational, analytical, and interpretative efforts in the field and on the fly. A number of audio recorders (even the ubiquitous iPodÔ4) are compact enough to be easily fitted into purses and pockets or clipped on with the aid of a carry-on case. Matched with a high-quality lavalier microphone, these recorders can be comfortably carried around (and ignored) by informants as they go about their daily interactions.
Ease of Use Even with the user-friendliest models, uncomplicated use off the bat is not a guarantee. Consider that most professional voice recorders record in multiple audio quality modes, have several internal storage folders, and offer standard playback, rewind, and fast-forward features. Still, in most recording sessions, informants will only need to know how to use the most basic functions of the recorder—record, pause, stop. Take care to discern the accessibility of these three functions. Informants need to be able to quickly break and resume their recordings according to their needs and those of their interlocutors. The OlympusÔ audio recorder I used came with a clipon remote control microphone that made stopping and resuming recordings quick, with no need to fumble and rummage for recorders. Unfortunately, this microphone has been discontinued. As an alternative, consider a high-quality lavalier microphone along with a clip-on carrying case for the recorder so that it can be attached to the outside of a bag or clothing for easy access to
Downloaded from fmx.sagepub.com at UNIV OF FLORIDA Smathers Libraries on February 4, 2012
6
Field Methods 000(00)
operation buttons. Informants should be unencumbered by the recording equipment, enabling them to be more empowered participants in the research. Affording informants more freedom and control can help offset the costs of the often intrusive nature of studying everyday talk.
Sound Quality Considerations Sound quality is crucial, but versatility is an important consideration. Both are interrelated, so let us consider them concurrently. Versatility refers to a recorder’s ability to capture sound in many types of social situations with various sound ecologies and noise levels. This can also depend on the quality of the microphone. If your research focuses on specific behaviors that occur infrequently in ways that are difficult to predict in advance, you may have to cast a wide net and record continuously across various social domains. In such cases, high-end business voice recorders designed for both direct dictation and recording conferences and business meetings may fit the bill. Field recorders used by journalists to record interviews, informal conversations, and unplanned encounters are also quite versatile, if perhaps a bit less portable than many professional digital voice recorders. To give you some ideas, a recent comparison of digital recorders in a number settings found the Olympus LS-11 and TascamÔ DR-100 suitable for most recording situations (Dressing et al. 2010). Keep in mind that humans hear at a range of 20 Hz–20 kHz, with most speech frequencies occurring between 250 Hz and 8 kHz (Stockdale 2002). CD quality sound is sampled at 44.1 kHz, and most digital recorders have high-quality audio settings that will record at this level. With numerous recorders having highly comparable audio quality outputs, price, portability, and/or ease of use may be the ultimate deciding factors. Most audio voice recorders will have various options for audio output file format, including both uncompressed and compressed types. Uncompressed files, such as .WAV, a Microsoft and IBM audio format, store all the information of an audio track, which, depending on the bits per second sampled, will be as faithful to the original sound as possible. This maximum audio quality is standard for sound engineers and other audio professionals working with professional audio software used to edit, transform, and reproduce sound. Because they keep all information for an audio recording, .WAV and other uncompressed audio file types are large, making storing and sharing more difficult. Uncompressed audio files can be converted to smaller compressed file types, such as .WMA or .MP3, which take up less space but lose audio quality. For example, low and high frequencies of an original sound may be
Downloaded from fmx.sagepub.com at UNIV OF FLORIDA Smathers Libraries on February 4, 2012
Negro´n
7
deleted out in the compression. Uncompressed file types are typically beyond what is needed in research of daily interaction since compressed audio file types are of sufficient high quality for most research needs. Even subtle phonological variation is safe with compressed files, as compressed audio files reduce resolution of audio signals imperceptible to the human ear. Compressed file types are especially important in audio recording of long continuous interactions many hours long, which if uncompressed, would produce enormous files. For example, approximately 6 hours of quality uncompressed audio would take up 4 GB of hard disk space. Compare this to approximately 80 MB for a WMA of similar duration. Omnidirectional lavalier microphones are ideal for recording the voice of key informants and the speech of interlocutors because they capture sound at 360 of the microphone. Worn close to a person’s mouth, the microphone will best capture the voice of the person wearing it, but because of its omnidirectionality, it will also receive others’ voices as well as background noise. Omnidirectional microphones are advantageous when recording group interactions, with minimal background noise. Unidirectional microphones pick up sounds aimed directly into its center and are therefore best in research situations with one focal speaker at a time. Given my emphasis on focal informants’ speech, I used a unidirectional microphone placed closely to their mouths. Still, the microphone was powerful enough to capture the speech of informants’ interlocutors as well as background sounds. It was even able to record the speech of people who informants talked to over the phone. The microphone was suboptimal when loud background noises overwhelmed informants’ voices—even though the microphones were close to speakers’ mouths. Such noises included passing subway trains, music at a dance club, and ambulance sirens. The microphone was also overwhelmed by heavy, gusting wind. Background sounds are underappreciated. Think twice before selecting a microphone that privileges a speaker’s voice at the cost of muffling background sounds. Background sounds help contextualize the recorded conversations listened to months after fieldwork is finished. If your research took you to and through various places, surrounding noises will help you further identify and remember recording times and settings. Imagine, for example, how squalls captured by the recorder on an especially windy day could help you reenvision anywhere from the weather that day to where you were sitting when a particularly strong gust blew the pages on your notebook. Moreover, background sounds provide fertile ethnographic detail about the places that people inhabit and the lives that they lead in those places (see Feld and Brenneis 2004). My NYC recordings reveal a rich urban landscape
Downloaded from fmx.sagepub.com at UNIV OF FLORIDA Smathers Libraries on February 4, 2012
8
Field Methods 000(00)
full with the movements of people, car traffic, subway trains, and music blaring from stereos. The aural portrait constructed by these background sounds deepens the analyst’s empathetic engagement in the lives of informants, as she or he imagines what is familiar and mundane for informants. Indoor sounds, too—TVs, radios, telephones, water rushing from faucets, voices, steps—serve as tags suggesting where and how people routinely move and dwell and the ways daily life is spatially arranged.
Recording Capacity You will find that the latest digital recorders on the market tout long recording times—ranging from 20 to 600 continuous hours. However, the recording time actually available to you will depend on the level of quality that you select to record in, whether the audio file is uncompressed or compressed, and whether you use a removable memory card. The larger a card’s capacity, the more recording time you will get from your recorder. If you require high-quality audio and long recording times, you may need to make the additional investment on a flash memory card. Some recorders, which have from 1 to 4 GB of storage space, may not require this additional purchase. As a general rule, uncompressed high-quality audio will take up considerable memory and will reduce the number of recording hours available. Lower quality audio—LP or SP, for example—will take up less space and will make more recording time available to you. Also keep in mind that battery life is usually shorter than recording times. The battery life on most digital voice recorders lasts from 10 to 30 hours. When recording in the field for extended periods of time, have backup batteries available.
Before You Begin Recording Whatever recording equipment you select, test it before you begin research. Be sure to set the date and time as this will ensure that time stamps on the recording will accurately reflect when the recordings were made. Time stamps are incredibly useful in conjunction with detailed field notes. If you include detail on the precise time that an event of interest took place, you can later cross reference this event with the recordings using its time stamp. Wear the recorder, and any microphone you get for it, for a few hours and in different settings. This will help you prepare for the inevitable logistical hang-ups that happen in the field. Get a sense of how easy it is to start and stop recording and prepare instructions and suggestions to guide research informants. Once you begin the research, you will find that recording
Downloaded from fmx.sagepub.com at UNIV OF FLORIDA Smathers Libraries on February 4, 2012
Negro´n
9
problems will arise because of confusion about how to operate the recorder. For example, accidentally pressing ‘‘Menu’’ or a similar button on some recorders may cause confusion if the user does not know how to exit the menu and return to normal recording mode. If applicable, be sure to provide informants with straightforward instructions on how to exit menus and folders. You may also need to consider a backup recorder in the event a recorder is lost or damaged. In addition to thinking through contingency plans for the recording sessions, try various audio quality settings until you find the settings that meet your quality and recording time needs. Finally, understand where and how audio files are stored and practice copying these files onto your computer for storage.
More on Portability and the Challenges of Informed Consent Ethical and Legal Considerations Portable recorders that are small and sleek—some even complement what a person is wearing—may lead the person carrying it to forget that he or she has it on. Like wearing a clip-on cell phone, as some of my informants described it, portable recorders do not raise much attention, enabling informants to go about their daily activities more comfortably. Beware, however, because a comfortable, low-key recorder can often go unnoticed by those who your informants encounter. Ethical research practice—and institutional review boards—demand that people be informed about the presence of a recorder, particularly for long encounters where identifying information is revealed. However, let’s be clear that when recording everyday life, the recorder will capture the voices of many anonymous people. Realizing this, I researched New York State laws that applied to recording brief periods of conversation without consent. If persons engaged in a conversation reasonably expect that their conversation is private, recording without consent is illegal. Conversations in public spaces within earshot of others—such as at a restaurant or street corner—are generally not considered private, so recording in such situations is protected by law. Furthermore, U.S. federal law allows for ‘‘one-party consent’’ statutes. This permits a person to record her or his conversations without the consent of other parties. Be sure to consult the applicable laws for your field site. Strive to obtain consent from all participants in a conversation, but accept the reality that you will not always be able to announce the presence of a recorder. Certain encounters, because
Downloaded from fmx.sagepub.com at UNIV OF FLORIDA Smathers Libraries on February 4, 2012
10
Field Methods 000(00)
of their often impersonal, speedy, and routine nature, challenge the balancing act between keeping informants’ right to avoid embarrassment and their interlocutors’ privacy. Take steps to ensure as much as possible the privacy of those who do not volunteer for the research but nonetheless are captured in the recordings. In most cases, you will find that in brief encounters, nonconsenting speakers do not provide identifying information, so their anonymity and confidentiality is protected.
Protecting Participants’ Right to Privacy I took a tiered approach to protecting nonkey informants’ privacy. Close family and friends who informants interacted with regularly for extended periods of time were fully aware of my research before observations and recordings started. Actually, for some informants their participation in the research depended on the consent and support of these close relations. Relatives, friends, coworkers, classmates, and others who informants regularly talked to were told about the recorder at the first encounter once research had commenced. Typically, informants did this themselves, but when they forgot, I would point out the recorder to their contacts. Finally, in encounters with distant acquaintances or strangers, I used a 1-minute rule. If the encounter went on longer than a minute I would announce the presence of the recorder or signal for informants to do so themselves. For brief interactions lasting less than a minute, I deferred to informants’ discretion about whether to bring attention to the recorder or not. Informants also had the choice to pause the recorder if they could not comfortably gain consent, and they sometimes did. Beyond these privacy safeguards, the recorder was never hidden from view. The recorder’s lavalier microphone, blinking lights and all, were always visible. Therefore, all interlocutors had a reasonable chance to become aware of the recording in progress. Even so, one informant, a street salesman named Abel, had many brief encounters with strangers in which personal information was exchanged. Understandably, he was reluctant to interrupt these sales encounters to announce the recorder. As an added measure, I provided Abel with a nametag reading: ‘‘Research participant. Recording in progress.’’ And, of course, my continual presence with a field notebook in hand provided further evidence that research was, in fact, in progress.
Managing Participants’ Reaction to Recording However interlocutors become aware of the recorder, expect to receive mixed responses. Reactions will range from the outright enthusiastic to the
Downloaded from fmx.sagepub.com at UNIV OF FLORIDA Smathers Libraries on February 4, 2012
Negro´n
11
downright defensive. Take, for example, the case of an acquaintance of my informant Alba. The acquaintance, a seamstress Alba visited for a business referral, discovered the blinking lights of the recorder before it was announced. She immediately became suspicious. She looked at me squarely, considering me as she ascertained in seconds whether I was trustworthy or not and asked what was going on. Before Alba, who stood blushingly by, could explain, I quickly clarified why the recording was being used and apologetically suggested that the recorder be turned off. Even though the acquaintance was still slightly apprehensive, she went along with the interaction and consented to the recorders’ use. Her relationship with Alba was the major reason that she allowed the recorder to remain on. Remember, informants’ contacts are more likely to trust the researcher and the recording if they trust the informant who has introduced them into the research situation. Whether a conversation will be allowed to be recorded or not can also depend on the subject matter under discussion. Now and then, my NYC informants voluntarily paused recording when their interlocutors discussed health problems, relationship troubles, and other highly personal topics. To preserve the trust that an informant and his or her interlocutors accord you as the researcher, swiftly and discreetly accede to pause recording at informants’ request. But do not assume that in all situations where sensitive information is being shared that informants and their cospeakers will want to stop the recording. The day after Alba and I saw the seamstress, we met with the owner of a hairstyling salon who was looking for potential buyers for her business. The businesswoman invited us into a tiny, cramped backroom used for facials and waxing to discuss the details of the sale. Noting the extra steps she had taken to meet with us in private, after she closed the door I asked if she wanted me to turn the recorder off. While she showed some discomfort with its presence, she did not actually ask Alba to turn the recorder off. With utmost respect and discretion, err on the side of fearlessly capturing all the ethnographic gold that you can. On this final note, it is important to be aware of different reasons why people refuse to be recorded because they may require different responses from you or the key informant. Besides issues of interpersonal trust or conversation topic, another important reason why some people ask not to be recorded is that they have developed a general distrust of, or philosophical aversion to, being recorded. Even in ordinary public interactions, I have encountered some (but few) people loath to relinquishing control of any likeness of them—an image in a photograph or their voice in a recording. They may highly value privacy or tend toward self-preserving or distancing behaviors. For people with such concerns, affirm that the recording is for
Downloaded from fmx.sagepub.com at UNIV OF FLORIDA Smathers Libraries on February 4, 2012
12
Field Methods 000(00)
research purposes only, will be kept anonymous, and that only you, the researcher, will have access to the audio. Finally, some people feel embarrassed by the sound of their voice in a recording or that they will say or do something humiliating during the recording. Help them feel at ease by empathetically sharing your personal experience—humorous, embarrassing, and frank—with being recorded.
More on Reactivity Researcher Inconspicuousness While recorder portability poses clear informed consent challenges, it has the distinct advantage of freeing the researcher to take notes and attend to observation. Just as the informant is free to go about their daily life as comfortably as possible, without a recorder to carry around the researcher is less conspicuous as an observer. This is one of many ways to reduce observer effects that can accompany behavioral research. Given the continuous monitoring conditions of my research, much of the time I played a handsoff role in the various activities in which informants were involved. Rather, I was often in the background to informants’ encounters with others, documenting the interactional contexts and communicative behaviors. My steady presence and low level of intervention in informants’ activities meant that, at least initially, they were highly aware that they were being observed. But spend enough time with your informants and such reactivity eventually tapers off (Bernard 2006:425), as it did in my research. The same applies in situations where the recorder is the ‘‘observer’’ and no researcher is present. Mehl et al. (2001) tested people’s comfort with wearing an electronically activated recorder (EAR) used to regularly (but not continuously) sample daily conversations. They found that in the second day of recording, the research participants were less aware of the EAR. Participants also reported that by the second day, other people were less influenced by the presence of the recorder.
Factors Affecting Observer Effects I concur with Hammersley (2003:343–44), who notes that the ‘‘character and seriousness’’ of observer effects ‘‘varies depending on what is being investigated, the context in which it is being investigated, and the resources available. It does not involve a fundamental, irresolvable dilemma or paradox.’’ Observer effects are important to keep in mind, but need not be obsessed over. Rather, faithfully and reflexively document what effects your presence has in a particular situation (Hammersley 2003; Monahan
Downloaded from fmx.sagepub.com at UNIV OF FLORIDA Smathers Libraries on February 4, 2012
Negro´n
13
and Fisher 2010). Build into the research schedule a period of independent recording (see below), which can be used for further reflection, analysis, and comparison with the recordings done in your presence. I found that the effects of my presence and the recorder tended to differ depending on informant characteristics, like age, but also, on relational factors. The youngest participants—less than 30 years old—tended to have interactions with younger people who were, in turn, less averse to the idea of being followed around and recorded. In general, relational factors trumped the benefits of youth such that the weaker their relationship to an interlocutor, the more uncomfortable they felt with the recording and observation. For example, the less familiar informants were with someone, the less they knew about how she or he would react to the presence of the recorder. So, it is important to note that the comfort of nonfocal speakers— like an informant’s acquaintance—is an important factor affecting observer effects. In talking with the NYC informants about their experience, I learned that if they sensed discomfort in their interlocutors because of the research situation, they because more self-conscious and uncomfortable. Therefore, to the extent possible, build relationships with others in your key informant’s life who you will encounter often during recording sessions.
Minimizing Observer Effects Beyond reflexive awareness, there are number of things you can do during the observations to minimize observer effects. Gumperz (1964) and Labov (1984; see also Labov et al. 1968), two pioneers in the study of naturally occurring speech, reduced reactivity through participant observation and group sessions. Both Gumperz and Labov had tape-recorded informal conversations with groups of people that already knew each other. In this way, speakers’ awareness of being recorded was counteracted by the preexisting group dynamics of the speakers. Interestingly, though, I found that in the presence of friends during early recording sessions, some informants overperformed certain behaviors. Rather than becoming more formal or measured—as can commonly happen with reactivity—some informants overplayed a certain outgoing or hyperactive persona to be more interesting for me to observe. One such informant confessed his concern that I would find his daily routine and behavior not interesting or useful enough for the research.5 To help him adjust to the research conditions, I affirmed that even the most mundane aspects of his life were fascinating and relevant from an anthropological viewpoint. Besides, as Monahan and Fisher (2010:363) argue, ‘‘staged performances’’ are important and analytical useful
Downloaded from fmx.sagepub.com at UNIV OF FLORIDA Smathers Libraries on February 4, 2012
14
Field Methods 000(00)
‘‘because they are deeply revealing of how individuals perceive themselves and would like to be perceived.’’ Another approach to help lessen reactivity is to accompany informants during special activities that help diminish the research/participant boundary—like parties and family gatherings. One strategy that is not often discussed as a way to reduce observer effects is to screen potential research participants. This may be especially important if you have limited research time and resources. Think about selecting informants according to personal attributes that could make recording a positive experience for both the informant and the researcher. Ask specific questions about daily routines, recording comfort level and previous experience with research, extent, and variety of daily activities, the typical interactions they have in a regular day, and so on. In this way, you can gauge whether a prospective informant could face difficulties overcoming observer effects. As one informant described it, having someone observe and note down your behavior is a bit strange. Even those who are enthusiastic about the process will find themselves in situations with people who they do not feel at ease with. Add being audio recorded to this, and the experience can be downright intimidating. Being recorded can induce anxiety in some (Blaxter et al. 1996:154). People may feel embarrassed by the sound of their recorded voice or nervous about what slips of the tongue or embarrassing conversations will be captured for perpetuity. Mehl et al. (2001) allayed privacy concerns by giving research participants the option to review their recordings and erase what they did not like. Despite natural exposure fears, I found that informants enjoyed listening to their recordings. They had the opportunity to view seemingly humdrum activities in new ways, thus building explicit awareness of things they otherwise took for granted. Esperanza played back the recording of a presentation she did and discovered that she tended to pepper her talks with far too many ‘‘umms’’ for her taste. Anthony listened to a recording of himself during a date to evaluate how it had gone. Julia had great laughs with her friends and they listened to their recorded group conversations. Accentuate these potential benefits of the recording experience to alleviate fears, minimize informants’ self-consciousness, and build enthusiasm for your research.
Independent Recording To help control for observer effects, have informants audio record themselves independently. Although this has an obvious advantage in that it helps you assess the representativeness of talk recorded under observation, independent recording does have its drawbacks. With independent recordings, it will be more difficult to determine what is going in certain
Downloaded from fmx.sagepub.com at UNIV OF FLORIDA Smathers Libraries on February 4, 2012
Negro´n
15
situations. Consequently, it is better to do independent recording after a period of observation in which you are able to familiarize yourself with typical places, people, and routines in your informant’s life. Starting with observations also has the benefit of preparing informants for recording on their own. After a period of recording in your presence, informants will be better able to address technical, informed consent, and logistical issues during the independent recording. Another limitation of independent recording is that without field notes to refer you to key behaviors and events, you will have to spend more time listening to recordings to sift for exemplary cases. There are several things you can do to address this. At the end of each day of independent recording, ask for basic details about the informant’s schedule. Note who they spent time with, where they went and what their routine was, any special activities they experienced, problems encountered, and their thoughts about the recording session. Keep that information in a spreadsheet or other archive along with the name of the corresponding digital audio file. Given the recording capacities of most digital recorders—but depending on how many of hours you ask informants to record—it may be necessary to meet on a daily basis to save files and debrief. Finally, consider having informants keep a personal journal of their audio-recorded activities. It may be useful to provide informants with a set of questions to answer in their journal. If you must reduce the number of hours of independent recording to review, you could provide informants with specific instructions about what activities or events to record. After a period of observation (ideally no less than a week, including weekend days), you will have a better sense of where to focus recordings. Devise a plan with your informant for which situations to record and which to skip.
Some Final Thoughts While it is beyond the scope of this article to cover postdata collection in-depth, it must be mentioned that reviewing, transcribing (especially for linguistic analysis), and analyzing your audio-recorded data will be a herculean task. It is important to know this before you even get started with collecting data. Consider your timeline for research and publication. Depending on your resources and time constraints, you may need to limit the number of hours of talk you record and how many informants you work with. Still, even 2 hours of conversation will yield enough data for an article or chapter. Keep in mind that transcribing naturally occurring interactions—especially group conversations—will be more challenging and time consuming than transcribing interviews between two speakers following
Downloaded from fmx.sagepub.com at UNIV OF FLORIDA Smathers Libraries on February 4, 2012
16
Field Methods 000(00)
a predictable Q & A protocol. Depending on your research questions, it may be sufficient to just review and take notes on key interactions that you record, rather than transcribing. I found it quite useful to review recordings using specialized digital audio editing software that displayed sound waves. Since my informants recorded for hours at a time, there were many total hours of minimal or zero talk. I used the wave editor in Sony Sound ForgeÔ to look for near flat lines that indicated silent moments, and clicked on the peaks and waves that indicated talk. This saved me the time needed to listen to every second of the recording and saved me from the added annoyance of fastforwarding until I heard talk again. In this article, I have discussed strategies for audio recording people doing things and being in natural social settings. Drawing on examples from my research, I focused on the process of continuously recording key informants’ talk. A key benefit of doing continuous recordings of naturally occurring interactions is context. With continuous recordings (as with continuous ethnographic observations), you will more broadly capture information about the rhythms of daily life: the web of interactions, environments spanned, actions’ antecedents, and the ways that talk affects the course of a day’s activities. Many of the techniques and technologies discussed also apply for studies of more discrete events, in a set location (e.g., a home), with a known group of people who are all preinformed participants in the research. Some of the best such studies use video rather than audio-only recording (Goodwin 2006; Ochs et al. 2006). I chose audio recording because the ‘‘on-the-move’’ nature of recording sessions would have made it challenging to use a video camera. Still, for many research questions, particularly those related to talk-in-interaction, video is ideal because it captures both auditory and visual communication channels. Whatever the focus or approach, documenting everyday talk is a rewarding process for both researcher and informant. It attunes the researcher to the stuff of thick descriptions (Geertz 1973). Those daily acts—flirtations, asking for the time, quick calls to say hello, humming a tune, laughing at casual humor, agreeing, complaining, comparing shoes sizes, discussing plans for a trip, asking for an opinion, reporting to a supervisor, being still—can go unperceived or disregarded for their very commonness. For informants, the opportunity to be recorded is an opportunity to see one’s life as an exemplary case of the human experience. Declaration of Conflicting Interests The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Downloaded from fmx.sagepub.com at UNIV OF FLORIDA Smathers Libraries on February 4, 2012
Negro´n
17
Funding The author disclosed receipt of the following financial support for the research, authorship and/or publication of this article: This research was funded by a National Science Foundation Dissertation Improvement Grant.
Notes 1. Olympus Digital Voice Recorder DS-2200. This model has been discontinued. For reviews of other digital audio recording options visit: http://labrosa.ee.columbia.edu/recorders.html, http://bartus.org/akustyk/, http://www.audiotranskription.de/english/comparison-of-digital-recorders, or http://transom.org/? cat¼6. 2. Olympus MR-12 Remote Control Microphone. 3. The Olympus DS-2200’s quality settings were stereo Super High Quality (SHQ), High Quality (HQ), Standard Play (SP), and Long Play (LP). 4. Visit http://transom.org/? p¼7626 for guidance on how to use an iPodÔ or iPhoneÔ as a field recorder. 5. Focal informants were compensated $200 for her or his 2-week participation. Despite clear indication to the contrary in the informed consent forms, a couple of the informants viewed the research arrangement as something akin to a contractual agreement. In this view, compensation connoted an expectation of appropriate returns. Thus, partly underlying this informant’s confession was his concern that his participation would not warrant compensation. In Russell et al. (2000), a study of unpaid volunteers’ view of research compensation, participants opined that research compensation might bias results or induce unethical behaviors. Consider that observer effects can be further influenced by the amount and nature of research compensation.
References Bernard, H. R. 2006. Research methods in anthropology: Qualitative and quantitative approaches. Walnut Creek, CA: AltaMira. Blaxter, L., C. Hughes, and M. Tight. 1996. How to research. Buckingham, England: Open University Press. Dressing, T., T. Pehl, and D. Georgi. 2010. First comparison of audio digital recorders, http://www.audiotranskription.de (accessed August 10, 2010). Duranti, A. 1997. Linguistic anthropology. Cambridge: Cambridge University Press. Feld, S., and D. Brenneis. 2004. Doing anthropology in sound. American Ethnologist 31:461–74. Geertz, C. 1973. Thick description: Toward an interpretive theory of culture. In The interpretation of cultures, ed. C. Geertz, 3–30. New York: Basic Books.
Downloaded from fmx.sagepub.com at UNIV OF FLORIDA Smathers Libraries on February 4, 2012
18
Field Methods 000(00)
Goodwin, C. 1993. Recording human interaction in natural settings. Pragmatics 3:181–209. Goodwin, M. H. 2006. The hidden life of girls: Games of stance, status, and exclusion. Oxford, England: Blackwell. Gumperz, J. 1964. Linguistic and social interaction in two communities. In The ethnography of communication, eds. J. Gumperz and D. Hymes. American Anthropologist 66:137–53. Hammersley, M. 2003. ‘‘Analytics’’ are no substitute for methodology: A response to Speer and Hutchby. Sociology 37:339–51. Kolovos, A. 2007. Audio field recording equipment guide, http://www.vermontfolk lifecenter.org/archive/res_audioequip.htm (accessed August 10, 2010). Labov, W. 1984. Field methods of the project on linguistic change and variation. In Language in use, eds. J. Baugh and J. Sherzer, 28–53. Englewood Cliffs, NJ: Prentice Hall. Labov, W., P. Cohen, C. Robbins, and J. Lewis. 1968. A study of the non-standard English of Negro and Puerto Rican Speakers in New York City. Philadelphia: U.S. Regional Survey. Mehl, M. R., J. W. Pennebaker, M. D. Crow, J. Dabbs, and J. H. Price. 2001. The electronically activated recorder (EAR): A device for sampling naturalistic daily activities and conversations. Behavior Research Methods, Instruments, & Computers 33:517–23. Modaff, J. V., and D. P. Modaff. 2000. Technical notes on audio recording. Research on Language and Social Interaction 33:101–18. Monahan, T., and J. A. Fisher. 2010. Benefits of ‘‘observer effects’’: Lessons from the field. Qualitative Research 10:357–76. Negro´n, R. 2007. Switching of ethnic identification among New York City Latinos. Unpublished PhD diss., University of Florida. Negro´n, R. 2011. Ethnic identification among urban Latinos: Language and flexibility. El Paso, TX: LFB Scholarly Publishing. Ochs, E., A. Graesch, A. Mittman, T. Bradbury, and R. Repetti. 2006. Video ethnography and ethnoarcheological tracking. In The handbook of work and family: Multidisciplinary perspectives and approaches, eds. M. Pitt-Catsouphes, E. E. Kossek, and S. Sweet, 387–410. Mahwah, NJ: Lawrence Erlbaum Associates. Plichta, B. 2010. Field recording, http://bartus.org/akustyk/signal.php (accessed August 10, 2010). Russell, M. L., D. G. Moralejo, and E. D. Burgess. 2000. Paying research subjects: Participants’ perspectives. Journal of Medical Ethics 26:126–30. Stockdale, A. 2002. Tools for digital audio recording in qualitative research. Social Research Update 38, http://sru.soc.surrey.ac.uk/SRU38.html (accessed December 12, 2011).
Downloaded from fmx.sagepub.com at UNIV OF FLORIDA Smathers Libraries on February 4, 2012