An incursion into the “non-respectable” side of fieldwork

This article was originally published in 1981. It addresses ethical and methodological questions involved in fieldwork among people from the anthropologist’s own society, but from a subaltern social group. It asks: What does legitimate the scientific methods of investigating other people’s lives, making them appear as acceptable instead of as object of resistance? It uses Foucault’s analysis of social sciences’ regimes of power-truth to understand what frames different methodologies of research and the author’s practices. It considers that field research is structured in the context of a certain regime of production of scientific knowledge that legitimates relationships of power in which one asks to know everything and the others feel obliged to tell the truth that, however, only the questioner will be able to reveal. The article argues that what is presented as truth either in interviews or in the text of the analyst is the product of a certain relationship shaped by power imbalances and social inequality. Moreover, it suggests that the field relationship is productive: What is said did not exist before ready to be revealed, but was constructed in this uneven relationship. Therefore, the interpretation of field data must always consider the conditions of its production.


Resumo
Este artigo foi originalmente publicado em 1981.Ele discute questões éticas e metodológicas envolvidas no trabalho de campo com pessoas da mesma sociedade da antropóloga, mas de grupos subalternos, indagando: o que legitima os métodos científicos de investigação da vida dos outros, fazendo com que pareçam aceitáveis, ao invés de questionáveis?O estudo baseia-se na análise de Foucault sobre regimes de saber-poder nas ciências sociais, para entender o que molda diferentes metodologias de pesquisa e as práticas da autora.Considera que a pesquisa de campo é estruturada no contexto de um regime de produção de conhecimento científico que legitima relações de poder nas quais um pede para saber tudo e o outro se sente obrigado a dizer a verdade que, no entanto, só quem pergunta será capaz de revelar.O artigo argumenta que o que é apresentado como verdade ou em entrevistas ou no texto da análise é produto de uma certa relação, marcada por diferenças de poder e desigualdade social.Além disso, sugere que a relação de trabalho de campo é produtiva: o que é dito não existia antes para ser revelado, mas foi construído nessa relação desigual.Assim, a interpretação de dados de campo sempre tem que considerar as condições de sua produção.Palavras-chave: Trabalho de Campo; Conhecimento e Relações de Poder; Saber-Verdade; Entrevistas; Pesquisa Qualitativa.Brasília, p. 332-354, 1981.Author's note, June 2023.This article is being republished without any edits.The only change is James Agee's epigraph, which was previously published in French translation but appears here in English.Obviously, a lot has changed in these more than 40 years.Anthropology has changed, and alongside it, conceptions about fieldwork, authorship, and ethnographic writing have changed.More importantly, however, residents of the outskirts and their relationships with people from other social classes have changed significantly, exposing and challenging various aspects of power relations that previously went largely unquestioned.I analyze these changes in a recent text: Teresa P. R. Caldeira.Desigualdade e Legitimidade: Problematizando a Produção de Conhecimento Social. Tempo Social, São Paulo, v 33, n3, p 21-45, 2021.However, despite all the changes, the power devices underlying the production of qualitative research continue to operate and shape what we still call data.
An incursion into the "nonrespectable" side of fieldwork 1 "It seems to me curious, not to say obscene and thoroughly terrifying, that it could occur to an association of human beings drawn together through need and chance and for profit into a company, an organ of journalism, to pry intimately into the lives of an undefended and appallingly damaged group of human beings, an ignorant and helpless rural family, for the purpose of parading the nakedness, disadvantage and humiliation of these lives before another group of human beings, in the name of science, of "honest journalism" (whatever that paradox may mean), of humanity, of social fearlessness, for money, and for a reputation for crusading and for unbias which, when skillfully enough qualified, is exchangeable at any bank for money (and in politics, for votes, job patronage, abelincolnism, etc.); and that these people could be capable of meditating this prospect without the slightest doubt of their qualification to do an "honest" piece of work, and with a conscience better than clear, and in the virtual certitude of almost unanimous public approval.It seems curious, further, that the assignment of this work should have fallen to persons having so extremely different a form of respect for the subject, and responsibility toward it, that from the first and inevitably they counted their employers, and the Government likewise to which one of them was bonded, among their most dangerous enemies, acted as spies, guardians and cheats, and trusted no judgment, however authoritative it claimed to be, save their own: which in many aspects of the task before them was untrained and uninformed.It seems further curious that realizing the extreme corruptness and difficulty of the circumstances, and the unlikelihood of achieving in any untainted form what they wished to achieve, they accepted the work in the first place.And it seems curious still further that, with all their suspicion of and contempt for every person and thing to do with the situation, save only for the tenants and for themselves, and their own intentions, and with all their realization of the seriousness and mystery of the subject, and for the human responsibility they undertook, they so little questioned of doubted their own qualifications for this work.All of this, I repeat, seems to me curious, obscene, terrifying, and unfathomably mysterious."(Agee;Evans, 1969, p. 7) It may seem curious that it was a journalist who made these reflections and not a social scientist, especially one of those who dedicate themselves to "fieldwork," i.e., who supports a large part of their work in scrutinizing other people's lives.However, who knows, perhaps it is precisely this fact that gives us the first clue to understanding James Agee's astonishment and amazement: perhaps because he is not very committed to the rules of the "scientific method," he does not see this scrutinizing of other people's lives as very natural and legitimate, whereas, for social scientists, this practice would be so commonplace that it would not often occur to them to question it.
What, in our society, sustains and legitimizes this research practice2 to the point of making it appear natural?
Foucault has already shown us how the birth of social sciences is linked to the development of disciplinary practices and how power and knowledge are interconnected in our society.I believe that the concrete practice of field research, more specifically, the relationship established between the researcher and the informant, serves as an example of how this regime of power-knowledge and truth production is disseminated.
It does not seem too much to me to assume, for example, that if only the researcher thought it essential to inquire about the lives of others, they would have a great chance of receiving an astonishing number of refusals in response.However, as this is not what really happens, we are forced to think that we are facing a mechanism more widespread in society, which makes people see inquiry and subjection to it as natural, as long as, and above all if, in the name of Science.Otherwise, what would explain that some, because they speak and act in the name of Science, believe they have the right to ask about everything, to subject their "objects" to hours of boring questionnaires,3 to get irritated when an informant refuses to offer responses, to statistically predict the "expected and reasonable" rate of refusals, and so on?However, beyond that, what would explain why others, when placed in the position of "object," are willing to answer these extensive questionnaires to a stranger who knocks on their door, who were willing to report their lives, year by year, in all the details?I imagine that the answer to this is that we are in the presence of a device disseminated throughout society and with a similar functioning to that described by Foucault, which makes each person put into discourse the most intimate details about their own sexuality.
I believe that this device is what legitimizes, in the West, scientific knowledge as the only recognized form of knowledge.It gives the product of a reflection or research classified as scientific the status of unquestionable truth; it gives power to this knowledge.We would, therefore, be faced with a power-knowledge-truth device that would support and enable the practice of scientific research.It would be this device that, on the one hand, leads a researcher to be interested in knowing everything possible about a given object and makes them believe that the knowledge they can obtain through scientific methods resembles the truth (i.e., the truth itself); believe that they have the most effective instruments in their hands to obtain the information needed and that this must be the most comprehensive and detailed; consider that, by speaking and acting in the name of Science, the "others," those "researched" are obliged to provide them with the information they deem necessary.Furthermore, it would be this device that, on the opposite side, leads others to be willing to give statements or even to believe that it is their obligation to do so since it is a scientific research (even if they do not know what that is precisely); and that makes them strive to report the whole truth, since something tells them that what is looked for is precisely the truth.
I believe, in short, that it is the association between power and scientific knowledge that enables and sustains the relationship established between the researcher and the informant, which gives one the conditions to carry out their work (their power) and the other the idea of obligation and the need to submit.This implies seeing, therefore, that the relationship that is established in the field between the researcher and their informant is a relationship of power:4 a relationship in which one requires a statement and the other finds themselves in the contingency of responding, in which one asks that everything be said in the minor details, and the other strives to tell the truth, which, however, only the first can reveal.
Indeed, I do not imagine that such a mechanism works in these terms consciously, particularly since it is the fact that it is perceived as "natural" that allows it to function.I do not imagine, for example, that all social scientists want, consciously and deliberately, to know details to master, nor that the interviewee considers themselves so guilty and suspicious to the point of being forced to confess everything: certainly, at the level of conscious practice, the relationship established between the researcher and the researched takes on several other aspects.However, it does not seem possible to deny the action of this power-knowledge-truth device in the same way that one cannot hide the interference of the most varied range of phenomena, both objective and subjective (and here also from the researcher, of course).All these factors are present in the researcher-informant relationship, interfere in its progress, and leave their mark on the information obtained and with which the scientist will work.I believe that it is part of the responsibility of social scientists not only to think about their professional practice in terms of the repercussions of intellectual work on society as a whole but also to question their work, right where it is carried out, i.e., in the immediate relationship with their "object"-the people and social groups they study.I believe it will be of little use to denounce only the most general relations of power and exploitation that exist in society as a whole.Power practices exist diffusely throughout the social fabric, and the initial step for social scientists should perhaps be to denounce their own practice and take it as it really is, i.e., a power relationship and a relationship that is not neutral.

II
The question behind all scientific investigation is the question of truth.It can be said that the scientist, when starting research, is looking to discover a truth that is ignored or hidden (it is, therefore, up to them to reveal it).Also, it seems the interviewees, too-and I am thinking mostly of "qualitative" research-share this idea and are concerned with providing the truth and selecting what is and is not appropriate to report.Perhaps an example will help us to understand this aspect better.
An anthropologist interested in knowing what representations residents of the outskirts of a large city have of power and society after having decided that the best technique to begin obtaining their information was to carry out open interviews in which they would seek to obtain their life story, arrives at an informant's house and gives the following speech: The informant, a retired middle-aged man, after agreeing to talk, perhaps to alleviate the discomfort and awkwardness he was feeling, began by saying the following:

"Then I'm going to tell you the story of my farms up in the North."
He was then interrupted by his wife: "Look, he's going to lie; he's going to tell lies."Also to lighten the mood, the researcher said, in a joking tone: "You can't make up stories, huh."

And he heard the answer:
"I'm going to lie.If I tell the truth, I'll get in trouble, so I'll lie.I have to lie to defend myself." The researcher: "Why do you think you'll get in trouble if you tell the truth?""Well, there's nothing, there's nothing to complicate…" "Do you think that telling the truth makes things complicated?""No, I don't think so.It will get complicated if I lie." "Ah, good, but we're going to tell the truth.Where were you born?" Let us say that this is a somewhat exceptional dialogue in which, right from the start, the rules of the game were explained, and the power relationship was placed on the table.However, in many other ways, this topic always comes up again in interview relationships.In the most diverse ways, what always seems to be present in the informants' minds is that, because it is a scientific research situation, their obligation is to "tell the truth." However, it turns out that the truth is not something that exists readily and objectively and that, to be discovered, the interviewee only needs to put it into words and the researcher to record it.This truth is produced from a relationship by the very nature of an interview situation.As we have already seen, this is not a neutral relationship but rather one in which the exercise of power is present.Furthermore, it is not a relationship between things, between objectivities, but between two subjects, which can suffer various interferences.Let us try to identify which elements interfere in this relationship, from the most objective to the most subjective.
Let us return to the interview mentioned above.In a few sentences, the informant let the researcher know what probably distressed him throughout the interview: the feeling that he runs a risk both in the case of telling the truth and in the case of lying and that, therefore, the best one can do is try to defend oneself.In this defense, there is a margin for exercising the freedom of those who submit to the relationship, and there is space for exercising power on the interviewee's part.The most significant defense of the person being asked for an interview is undoubtedly to refuse the relationship.However, if one cannot or does not want to avoid it, they will find other defense mechanisms: they can select, hide, or deceive.
A game is then established, the outcome of which will depend on the interviewee's and the interviewer's actions.Anything can happen, from the production of an absolutely imaginative speech, but which is not that imaginative, insofar as what one thinks is appropriate to say can be being said, to the production of an absolutely "true" speech.
Honestly, what caught my attention the most in my experience as a researcher were not the imaginative speeches but the true ones.First of all, it was very rare that I felt faced with an interviewee who, instead of referring to concrete data about his life, was referring to a fantasy.Secondly, I believe that it is not very difficult to understand what makes a person try to hide their true story from a stranger.What always seemed incredible to me is that people made such an effort to present the data realistically.The impression was that they felt obliged to provide a true statement, marking the speech with an essential and confessional character and, often, transforming it into an outburst.However, can obligation to "tell the truth" alone, aroused by the fact that it is scientific research, be sufficient to explain that many interviewees find themselves forced to reveal extremely unpleasant or painful aspects of their lives, even without having been asked to do so?
It is common, for example, the following situation: after the interview has ended, and the researcher is about to leave, satisfied with the information obtained, the interviewee calls them with phrases such as: "You know, there is something I need to say, I'm not really married to him, and he's not that girl's father."Then, from there emerges a whole story of marriage, marginalization, second marriage, and the concealment of the fact.As is common, at a certain point in the story, the husband turns to the woman watching the interview and asks: "We have to tell the truth, right?So we'll have to tell everything as it was, straight out." Situations like this left me quite embarrassed.On the one hand, I did not know what to make of the report, which was generally unpleasant and very emotionally charged (I will return to this aspect later when discussing the researcher's emotions).On the other hand, I still did not understand the reason for the report since it was far from me to request it directly (although nothing guarantees me that the simple fact of sitting with a recorder on and the university's presentation was not already making the request).Even without understanding it very well and with a growing feeling of uneasiness, I continued to collect life stories.The discomfort came, on the one hand, because I began to wonder about the pertinence of asking for life stories, which is when there is more clearly an invasion of other people's lives (much more than in the case where I asked for objective opinions, for example, about the city)-is this technique used so much just because there is an obsession with scrutinizing everything?The discomfort, however, also came from elsewhere.I started by supposing that life stories predisposed people to embarrassing situations because they involved taking stock of one's life and reliving it.However, the situations persisted in interviews on objective subjects and during the application of a quick survey.Sometimes, the questions in the questionnaire (whose application took 15 minutes and generally asked on the street, outside the house), which, in my opinion, were absolutely innocuous, such as the number of residents in the house, number of rooms, origin, religion, education and work of the residents, were enough to trigger long reports, lamentations, and even tears.It became clear that the answer to embarrassing and detailed situations should be sought in other aspects.The interviews with women, in which this type of situation was more common (probably because I was a woman), began to provide me with some indications to try to explain what was going on.
Several times, when interviewing women, the conversation quickly turned to discussing the couple's complicated-and often dramaticproblems without me even mentioning the subject.The fact is that when I asked a woman to tell me about her life, she often dismissed all other aspects as secondary and went on to talk at length about marital problems, which, without a doubt, are the ones that give the most pretext for anguish.They were sad, distressed reports, almost invariably accompanied by crying spells (there were very few interviewees who did not cry when telling about their lives) and often interspersed with questions about very concrete subjects (for example, the use of contraceptives), in search of information.However, the question remains: Why talk about such intimate subjects that are practically untouchable in everyday life with a stranger?I think this is precisely why: talking to your neighbors is risky, as one can see their lives discussed throughout the neighborhood overnight.Firstly, my distance ensured the possibility of getting closer, letting off steam, and asking for advice.Furthermore, because I am an "anthropologist" (which should mean practically nothing to the residents of the outskirts but merely indicates importance, distance, and a superior position for those who know), I could provide more precise information than those of a neighbor who would be in the same conditions as them.Often, the answers to questions I had asked during an interview were interrupted.
The interviewees asked me to explain something to them or inform them about the most varied subjects: from why there are elections, which party Adhemar de Barros belonged to, where they can obtain retirement papers from the National Social Security Institute (INSS), even whether or not it is advisable to inform teenage daughters about the existence of menstruation or which contraceptive method is most appropriate (the most frequently asked questions were about children's education and female sexuality); reaching situations like that of a woman who asked me what "cardiac" meant, saying that she wanted to know because her father, whose death occurred 19 years ago, causing a significant impact on her life, had died because he was "cardiac." I believe there is yet another reason why the interview turns into an intimate and dense conversation: it is an opportunity to talk, to stop and reflect a little on something that goes beyond the washing machine and the stove, to give a little bit of order to the world.It is, in short, an exceptional occasion that, by distancing oneself from everyday life, allows one to organize little bits of experiences that have been accumulated over time in the form of scattered fragments, which are left without connections or explanations.Therefore, it is common for the interview to turn into a cathartic relationship that often ends in a nostalgic and sad tone-of someone who, for example, has faced their past, thought about the present, and imagined the future-in thanks and relief, if not crying.An interviewee told me this: We keep putting so much rubbish in our heads that when we stop to think, something comes along, it distracts us, and we start living again.Now, as I'm talking to you, we seem to unburden ourselves a little, feel more relieved, I don't know.There are times when we can't talk; there are many neighbors here, and we can't talk to them about our lives; some laugh in our faces, take it as a joke, and even ignore it.So, there are times when we close ourselves off […].So many things, sometimes so many people that we could feel lighter, and our heads are full of confusion.
Although I believe that the situation of lack of information and interlocutors is present for both men and women, I believe that it is more serious in the case of women, who, for the most part, are restricted to the domestic universe.The problem that appears in interviews with women is one of the reflections of the daily experience of a specific situation: that of the condition of a woman and, more than that, the condition of a poor woman, for whom social conditioning certainly weighs more than for a middle-class woman.Many of them come to intuit their displacement in the world and relate it to the fact that they are women.The same interviewee mentioned above stated:

Women don't need to learn anything; only men do. The woman, I don't know why the woman always has to stay behind. I also don't know what this is; what defect does a woman have that a woman can never do anything about?
There is no explanation for this.
In short, the fact that women have fewer instruments to think about the world (outside the domestic universe) and their position in it are elements that, in my opinion, allow one to understand the more significant load of anguish contained in their interviews and also the more intimate and "true" story. 5owever, other aspects are also involved in an interview and telling a life story that can help to understand both the true aspect of the report and the agreement to do so.
It is quite common to be grateful for the opportunity to speak, to express the contentment and relief caused by the interview situation, such as saying goodbye to the researcher with the phrase: "It was an honor to be able to speak with you."I believe that a common element appears in field research carried out among poor populations: granting an interview is a privilege and a mark of distinction.Many times, in the neighborhood on the outskirts of São Paulo where I researched, I was called by people who asked me: "Won't you want to talk to me too?"; "And to my place, won't you go there?"; or I was forced to face unpleasant situations with offended people because I had gone to their neighbors' houses, not their own.Distinction, importance, and identity are elements that the researcher and the interview situation confer on the interviewers, and these aspects become more present when it comes to members of the dominated sections of the population.Societies relegate those who are dominated to silence: their lives are spent in anonymity, they have no voice, and their identity is not recognized, except in very specific situationsin general, they are just one among others (in the factory, in the INPS or bus queue, when voting in elections, and so on).The act of asking someone to describe their life contains (for the interviewee, even if the interviewer may not be aware of it or may not intend to do so) a recognition of individuality and respect for that life.After all, it is not just any life that they are being asked to report, but a particular life, their own life.The interview gives voice and removes from anonymity a life marked by suffering, anguish, and struggle: It is a life whose story can only be marked by suffering because, in addition to being an integral part of it, it is what, in some way, confers dignity.
In what other situations does a dominated class member have their identity recognized?In what other situations are importance, respect, and dignity given to their lives?In what other situations are they called by name, and are their complaints heard and considered?In our society, these situations are not very common.This recognition can appear in political discourse, especially in those of a populist nature-the pat on the back, the handshake, and the phrase "I'm counting on your vote."It can also be present in the practice and rituals of popular religion.However, more often than not, the identity of a member of the dominated classes is recognized only when they threaten order-when committing a crime, a poor person is challenged as an individual; part of the violence that is carried out against them is precisely that when judging them guilty, identifying them by class and name, and making them respond individually. 6 believe that this recognition of identity is one of the aspects present in the interview situation and can give satisfaction to the interviewee, even explaining their willingness for the relationship.It frequently appears in the informants' speech.For example, the same man who started talking about the issue of truth and lies ended the account of his life with the following observation: "It's a pleasure to talk.You know, I could write a book if I told my life exactly!Down to the smallest details." However, this type of recognition cannot always be expressed or, rather, understood.How often have I been forced to answer the question: "But why do you want to know my life?I'm just a random person; I don't even know how to speak properly!"Countering it was difficult.How can one convince them that their lives are meaningful?It is certainly easier to believe that one is under suspicion (the most significant problem a researcher faces when arriving in a poor neighborhood is convincing its residents that they are not tax inspectors or the police).Furthermore, how do you convince someone without a voice that they can speak?Numerous times at the end of the interview, I heard them apologizing "for anything, you know, we don't understand the things" and so on.However, I believe that these apologies, which are often one of those formal phrases spoken for the sake of speaking, are said in a sufficiently ambiguous tone to the point of not being able to hide a hint of satisfaction.It is evident, for example, in a dialogue that occurred between an interviewer and an interviewee after the latter had apologized for everything wrong that could be said, for not knowing how to speak, for not understanding things, etc.: "If there is a mistake, please correct it, as we are not used to talking like that" (interviewee).

"No problem! I really liked what you said; you have some very beautiful (researcher).
"Did you really think so?As you know, I never studied, I can't read, everything I say is a gift from nature" (interviewee).
I think comments like these, where ambiguity is present-in this case, the simultaneous recognition of inferiority, importance, and personal dignity-are the most precise portrait of a research relationship.No relationship is univocal; there is always a relationship of forces in which sometimes one aspect prevails, and sometimes another prevails.Even the interviewer's position of knowledge/power is experienced ambiguously, both by the interviewer, who knows but asks, and by the interviewee, who does not know but informs.This appears, for example, in situations like the following: I arrived with two colleagues at the house of a person we had contacted and were supposed to talk to.This person was not there but had left a message with the grandmother indicating the current location, which we knew where it was.The lady gave us the message and asked if we knew the indicated place.We said yes, thanked her, and left.When we were on the street, the lady said, "I'm sorry I asked if you know where it is.People like you know everything, I know, but there are times…" In short, I would like to suggest that the interview relationship is complex, in which varied and often opposing elements always influence each other.A woman's statement made to Ecléa Bosi illustrates in an exemplary way the multiplicity of aspects that may be present: -"I'll tell you everything because we are what we are… These are things that have already passed.Afterwards, I had so much happiness!"-"But telling you the difficult parts, that struggle, it feels like I'm telling a very dear person.I'm happy to tell you.I'd like to make those comments… I don't know… say very important things to you …" -"When we confess, we have to tell the whole truth.So, as I'm telling you the whole truth, I didn't have Christmas at my house" -"That night, I'm going to tell you, we don't even talk so frankly in the confessional; I cried all night without stopping" -"Who would have thought that one day I would open the book of my life and tell everything?And I'm grateful for that: it's good for us to remember.God bless you."(Bosi, 1979, p. 53, 60, 66, 74, 76) Testimony/Truth/Confession/Outburst/Help (to others or oneself)/Information.All these elements mix in an informant's speech and characterize it.In addition to triggering the discourse, the researcher's power situation marks some aspects.However, it does not account for everything occurring in the relationship-from a moment on, it takes on another character, and the interviewee sets the tone.What they say is not a statement that will only have meaning for those who request it to discover something; it is significant-and in a sense also a discovery-for those who provide it, for those who live (and relive) it.
I believe that this aspect clarifies one of the main characteristics of a research relationship in social sciences.The "object" of the investigation is not a neutral and passive object that can simply be observed: their testimony is, first and foremost, a significant action for themselves; it is an action experienced and an action experienced not in isolation but in a relationship with another, the one who triggered the action.However, the researcher is neither neutral and passive in this relationship nor just an observer who collects their material.The researcher is included at all times in the action that takes place.Firstly, it is to them that the report, the confession, and the outburst are addressed.However, perhaps their participation and inclusion become more apparent when the interviewee asks for clarifications, information, opinions, and demands, in short, an answer.An exchange relationship is established, and it is reasonably common for situations to arise in which, at the end of the interview, the informant says: "Is it over?Well, now I will interview you.You'll answer some questions." The interviewee provides the statement requested but receives things in return, ranging from the opportunity to talk about doubts and anxieties to information on objective subjects.But not only that.The statement did not exist and was ready to be said; it is constructed as it is said.All of this makes the interview relationship basically a learning relationship: both the researcher and the interviewee discover, learn, and reflect.The information produced-the recorded interview, for exampleis the result of this exchange, of this common learning where the most varied elements may have interfered.It cannot be produced twice similarly since it results from a relationship and how it happens.
The relationship of exchange and learning is present, clearly, in situations of open interviews and participant observation, and, in my opinion, this ends up somewhat mitigating the violence of the exercise of power that is established over the interviewee.However, this does not happen when applying questionnaires or conducting surveys.In these circumstances, the rigidity of the questions, the pre-established order, and the range of alternatives do not allow the interviewee practically any active participation in the relationship.I believe this is the situation in which the violence of the exercise of power in the practice of research in social sciences presents itself in the crudest form: the interviewee is not allowed anything other than to submit to answer what is asked of them.What is said beyond that is irrelevant; it is not noted or considered.Now, it is time to look at it from the researcher's angle.How, after all, do they behave in this relationship?How do they react?How do they manage these aspects?

III
Focusing attention on the researcher, I believe that there are two aspects to be considered: on the one hand, their subjective reactions during their stay in the field and their relationship with the informants; on the other, their attitude toward the collected material which, as already seen, includes a series of subjective and intimate aspects of the interviewee, which were produced in a scientific and by no means neutral relationship.
Starting with the subjectivity of the researcher, what seems to exist is a general tendency on the part of social scientists to treat as undesirable or as "folklore" of field research the emotions and discomforts they feel when trying to understand and experience the universe of " others" with whom they are relating.Descriptions of situations of this type are common in anthropologists' notebooks.Roberto Da Matta (1978, p. 4) has already drawn attention to the importance of "incorporating oneself into the field of official routines, already legitimized as part of the anthropologist's training, those extraordinary or charismatic aspects, always ready to emerge in every human relationship."This would be the area of anthropological blues, "that of the element that insinuates itself into ethnological practice, but that was not expected.Like a blues whose melody gains strength by repeating its phrases so that it becomes increasingly perceptible.In the same way that sadness and longing (also blues) insinuate themselves into the fieldwork process, causing surprises to the ethnologist" (Da Matta, 1978, p. 6).
Why not consider the times when the researcher was absolutely depressed and helpless when faced with a desperate report interrupted by tears?Alternatively, the times when, after hearing a woman's anguished outburst about her relationship with her husband, who drank and beat her almost daily, felt like simply saying to her: "Ma'am, work hard, leave this man, and go mind your own business!,"and the times it was said.Alternatively, the researcher's unpleasant feeling of loneliness when closing their house door and sitting down to write down the day's observations, unable to digest what had happened without anyone to talk to.Alternatively, the anger felt when unwillingly involved in a fight between political groups in the community where the study was being conducted.Or even the constant question: why not directly take a side in this fight, since the researcher has a very clear position, instead of pretending to everyone that everyone is right?How can one interview a snitch and a grassroots activist with the same "neutrality" When having contempt for one and admiration for the other?
In general, all these types of questions are not addressed, and, in truth, they cannot even be considered if the aim is to maintain the belief in the "scientific neutrality" of the researcher.It forces the researcher to be objective, not to get involved, to distance themselves from their research object and to make this distance clear to them, never to interfere, and to create a situation that is as neutral as possible.This makes it impossible to consider that it is not that simple and that omission and distancing are, first and foremost, specific forms of participation.
Even if one does not intend to make a psychological consideration, it is worth remembering that the relationship that provides data for knowledge in social sciences is a relationship between people who face each other as subjectivities and where all the attitudes of one are reflected in the other and are taken into consideration.For this reason, I believe that the attempts at distancing and asepsis recommended by many field research manuals do not have the expected results, namely, objectivity and noninterference.The interviewer's silence when faced with the interviewee's doubts and questions and their refusal to provide answers and opinions are not attitudes that isolate elements, but which, by isolating, include.Thus, the silence and reticence of the researcher, in addition to being embarrassing for them and causing great anguish in the interviewee, can contribute, for example, to triggering a series of fantasies and suspicions about them or to the production of "guilty," defensive, or "lying" speech.I do not believe, in short, that maintaining distance and silence is less interfering than getting closer, expressing opinions, and providing answers.Participating or refusing to participate have, in one aspect, the same effects on those surveyed: they are attitudes that interfere and are taken into consideration.One can choose to do one thing or another, even considering the effects one wants to achieve, but what cannot be imagined is that a situation isolates the production of effects.I believe that silence can be an excellent research technique, for example, to see how a specific population reacts to anguish, to the unknown (which does not mean that it is not a violent technique).However, silence and objectivity/ neutrality are certainly not the same thing.
However, I believe that the issue of objectivity and the "scientific neutrality" of the researcher goes even further.Somehow, the researcher too-and especially they-is treated as an object, not as a subject who relates to other people and who, as a human being, has emotions, sensations, and feelings.They are required to forget and disregard as obstacles to actual knowledge their entire, let's say, emotional side.The truth is that for those who believe that permanent neutrality and objectivity are necessary to carry out research and that these are possible, it must be tough to put their emotions under observation.It is preferable to deny them.Perhaps because the consensus on what scientific practice is is so strong that anyone even slightly committed to it prefers to leave some issues behind the scenes instead of making them public.
Perhaps it would be healthier and produce more effective results if social scientists, instead of apologizing and trying to settle accounts with the methodology and procedures of the exact sciences and with the objectivity and externality that are possible, would admit more clearly that the nature of the object of study is different and that, therefore, the research procedures must be different.It is a human relationship on the one hand and a laboratory relationship between a person and an object on the other.Furthermore, here, the very language inherited by the human sciences is already uncomfortable: it is difficult for me, for example, to designate an informant as an "object" of study.
Although the statement of these differences is a kind of commonplace in social science methodology work, I believe that it has not been taken to its ultimate consequences, at least in most cases.What seems to occur more frequently is that a great effort is made to adapt the techniques and methods of research and observation in a laboratory to a situation in which the "objects" are people, social life, or culture rather than looking for other methods.If this is not the case, why insist on the investigator's neutrality, deny their emotions, and rule out several aspects that could "interfere" with the results?"I believe the researcher's emotions should never be denied (because they are always present).However, beyond that, I believe they should be carefully considered, as they can become an essential instrument for knowledge.What I imagine may consist of the specificity and originality of the field research method in social sciences is precisely the fact that the researcher uses themselves as a research instrument and a source of observation.It is considering, for example, the situations they can provoke and the emotions and sensations they feel as essential sources of information.It is making an effort to identify the elements that are coming into play (including those of their own subjectivity) instead of trying to dismiss them as undesirable.It is to include the anthropological blues in routine observation.
Misunderstandings, shocks, and discomfortsalways unpleasant from a subjective and emotional point of view-are perhaps the most frequent situations in understanding the universe of "others."However, elaborating on them requires the researcher to always pay attention to their own emotions and sensations, considering their anguish, sadness, and astonishment as significant.I could multiply here examples in which chance led me to understand aspects of daily life in the outskirts that I would have difficulty capturing in an interview due to lack of reference: witnessing two children playing, a mother scolding her son, the gossip of two neighbors, the conversation between two friends at the bar after work or a man's astonishment at an observation that was obvious to me, but that he could not understand.Alternatively, the shock that some attitudes caused me allowed me to capture aspects of the way of life that are not talked about, as they are not conscious to the point of being able to put them into words.For example, the perception that the relationship between people who live on the outskirts of São Paulo is rude, harsh, direct, and without nuances and that silence is one of its most striking characteristics.Silence means the absence of an exchange of opinions and ideas and the unconditional acceptance of a relationship pattern transmitted by tradition (regarding sexual roles and age roles).One could say that it is a relationship marked by violence, which is not physical, but a silent violence, the violence of silence, of omission, of prohibition, of ignorance.However, what I want to draw your attention to is that I am the one who classifies violence in proportion to my shock and the contrast with what I recognize as an acceptable or desired relationship: for the people who experience it, it is just the normal, natural standard.

IV
After considering all these aspects, it is worth asking: What remains as raw material for the social scientist's interpretation work?What should their attitude be after leaving the field and sitting in their office with interviews and observations to analyze?I think these questions can be addressed from at least two aspects, which I would like to consider, albeit briefly: from an ethical point of view and a theoretical and methodological point of view.
First of all, I believe that it is when interpreting and publishing data that researchers can exercise their power most violently.At this stage, the people who gave the interviews and provided the data no longer have any interference, and it is only up to the researcher to decide what to do with them.The researcher has the option of, for example, making public an outburst or a confession, which was made to them for absolutely personal reasons, although this may not have been said.It is up to them to decide whether or not to use intimate data about the interviewees' lives.I do not believe that the fact that the interviewee knew that the researcher was collecting information for research is sufficient to serve as justification for publishing this type of data, just as I do not believe in the veracity of the phrase: "After all, they spoke because they wanted to."Things are not that simple: the interviewee may indeed have spoken because they wanted to, but one cannot forget the context of the interview's power, nor the fact that, often, their character is none other than that of a person asking helps someone else who they believes can help them.Outbursts and confessions arise in this context and have a significant and subjective character for the person who makes them, which has nothing to do with providing information or explanations.I believe, therefore, that the long confessions, outbursts, intimate reports, and the bulk of the data provided during a life story are elements that are part of and characterize a personal and intimate relationship, a relationship of friendship and trust (not investigation), whose meaning is exhausted in itself and, for the informant, at the moment it occurs.It is personal and non-scientific information.The researcher must take responsibility for not forgetting the position of power they find themselves in and the number of situations that can arise from there.For all these reasons, I believe publishing data provided in situations of confession, venting, asking for help, or something like that is unethical.This is not to get to the ridiculous point of saying that the researcher should forget what they heard in confession or pretend that situations like these did not exist.These circumstances are part of their relationship with the informants and are data that help understand their lives and universe.However, when it comes to publishing personal stories, I believe there is a big difference: using it as data is one thing, as text is another.Including emotions and feelings in the research observation routine, I believe, is a necessary attitude since they are really part of the relationship that is established in the field, and considering them contributes to the understanding of the reality that is being studied.However, I insist again that considering data for interpretation and understanding does not mean publishing them.
Finally, one would have to think about the interpretation of the data.Although I consider it necessary and urgent to have a more in-depth discussion from a theoretical and methodological 7 See, in this line, the observations made by Ruth Cardoso in a manuscript entitled Notas para Discussão, presented at the Seminar on Discourse Analysis - USP -1979(Cardoso, 1979).
point of view concerning the analysis of discourses and representations ("qualitative" data), it is not the objective of this work to follow this line.What it intends to do is simply provide some indications about the "nature" of data collected in "qualitative" research, which I believe can contribute to this discussion.First, it is worth remembering that interviews, testimonies, and field observations, however rich they may be, do not constitute evidence or an explanation.They are data, i.e., the raw material to be worked on.They need to be interpreted and explained to acquire a meaning for knowledge.In addition, they require work that is inseparable from a theoretical effort and that in no way resembles the compilation of data or pure and simple description.
Furthermore, when beginning interpretation, I believe it is essential to be clear about the "nature" of the data one has in their hands.They are not "objective," external, and univocal data: they are not linear and the product of an experiment where interferences were isolated, but the result of several factors that cannot be ignored.Both what was said (and which may be recorded or written) and what was observed and felt are data produced at different moments of a relationship in which the most varied elements came into play.Among these, there is a background provided by the knowledge-power device with all the developments that we tried to describe; the fact that it is a complex relationship that includes situations that produce confession, outbursts, requests for help and information; and the fact, finally, that it is a relationship between human beings, it would be better to say, an interrelationship, where subjective and personal aspects are obviously present.
Considering the context in which the data were produced, it is not difficult to see that these are multifaceted, fragmentary, contradictory, multidetermined discourses.Ignoring these characteristics during analysis could mean impoverishment, if not distortion.Any interpretation based on this type of data is probably a straitjacket to which the researcher subjects the discourse; each speech can probably be read and punctuated in different ways. 7I do not think one can escape this trap when analyzing, but it may be possible to avoid some more frequent risks.The biggest of them, and perhaps the most common, is trying to seek coherence in speeches they cannot have because it does not belong to them but to interpretation.Alternatively, the opposite risk, as common as the previous one, is denying the data any coherence of its own and subjecting it (often after being cut up and divided into items) to a pre-established model constructed from outside, intended to be proven or disproved.
I believe that a possible way to avoid the risks mentioned is, on the one hand, to respect the data as they are, trying to read them in all their heterogeneity and without distorting them with fragmentation imposed from outside (which may be logical, but not necessarily accurate); and, on the other hand, seek clues for interpretation in its fissures and contradictions.

1
Presented at the IV Annual Meeting of the National Association of Postgraduate Studies and Research in Social Sciences, "Popular Culture and Political Ideology" Working Group, Rio de Janeiro, October 29 to 31, 1980.Originally published in Ciências Sociais Hoje 1 -Trabalho e Cultura no Brasil.Anpocs -CNPq, Recife/