Jack Katz is a professor of sociology at University of California, Los Angeles (UCLA) and one of the main names in the interactionist tradition in sociology. He studied at Northwestern University, North Chicago, in the United States, and worked, as a student and later as a colleague, with renowned authors such as Howard Becker and Robert Emerson. In this interview for the
It is important to note that although this interview is focused on Katz’s work on crime, we have tried to explore other aspects with which, throughout his career, he has worked. It is worth remembering that his first book,
We are particularly grateful to Daniel Cefaï, for his encouragement and initial mediation with Professor Jack Katz.
I can tell you “how” that happened, which is what you ask. “Why” is a different question. When I took a job with the sociology department at UCLA, I had to teach something. Before coming to UCLA, in New York I had been studying how prosecutors (government attorneys) prosecute “white-collar crime”. I had expected I’d be writing about some aspects of crime, and that it would be helpful to know what social scientists had written about crime. In order to be efficient in the use of my time, I decided I’d teach criminology. That would give me a practical reason to read the literature, and routine pressure to form opinions that made sense, at least to undergraduates.
What I quickly found was that the emphasis in almost all of the social research on crime was on the explanans, on the supposed explanatory factors, with almost nothing on the explanandum, on the phenomena that the research community wanted to understand and explain. It made no sense to me that people would try to explain something without looking at it closely. Criminologists and sociologists writing about crime looked at police statistics, demographics, neighborhood characteristics, the state of the economy, etc., but not at the people doing crime when they were doing crime.
So I started to collect everything I could find that had some description of people doing crime. There were fragments in the academic literature, more in biographies and journalistic writings, and some collections of police files that had descriptions of what went on when crimes “went down” as seen and related by victims and observers. And, since crime is all around us (my study of white-collar crime had led me to understand that crime is not rare or exotic), I found I could supplement what I could find in published descriptions by calling on scenes I had witnessed and by asking students to do reports on crimes the students had done .
For a couple of years, I gathered these materials, sorted them into types of crimes as experienced by the people doing them, and taught my criminology course. Then I used that as a basis for writing
Excellent study by Bubandt and Willerslev. Why assume that empathy is positive, an inspiration to moral behavior, a basis for conduct that benefits the other? Consider the relationship between sex and love. (Sex is great for thinking. It cuts through so much academic sophistry.) At all historical times and in virtually all segments of society, one can find men trying to get into others physically, and to do that without force it helps to get into others interactionally and spiritually, to see how they are perceiving themselves and to focus on the detailed embodiment of their behavior, the cues they give as to what is moving them. The interaction process that ensues can lead to loving sex, to manipulative abuse, to sex without love, and to an infinite range of alternatives not easily classified in the binary, selfish or caring, helpful or harmful. Indeed, what the participants think is happening on this dimension at the moment will not always be clear, and what seemed clear at the moment will, when regarded later, not infrequently seem to be the opposite or at least not so clear. This ambiguity is natural to sex, the most fundamental process of reproducing social life.
So I would point out not only that empathy can be used to harm others, but also that empathy is routinely used in a range of behavior not comfortably coded as helpful or harmful. In arguing whether empathy is morally positive or sometimes negative in its uses, we risk missing the ambiguity, the openness, the always renegotiable meaning that is at the essence of social life.
It’s striking that philosophies and many forms of social science have wanted to presume that interaction competency, which requires at least minimally seeing yourself from the perspective of others, thus engagement in a process close to or partaking in empathy, has a positive moral character. That says more about the religious aspirations of social science than about the role empathy in social interaction.
The idea that to understand is to forgive has been appealing to many, including to George Herbert Mead, one of the founders of interactionist sociology.
The methodological issues you raise are widespread concerns, so it is important to discuss them. All the phenomena that social science would like to explain, not just crime, present similar challenges. Your question suggests that special methodological problems arise when we study behavior, we do not ourselves conduct. But even for those behaviors that we ourselves do, if for example we rob someone to get data on robbery, we face the problems of generalizing from our experience to that of others, since usually we are not content to offer autobiography. And, until we study what others do when we are not present, we do not know that instances of our own behavior are similar to what others do or experience. So autobiography is no panacea for social science; our basis for extrapolating from our own lives to others’ lives remains problematic.
The problems you raise have been recognized since social science began. Interviews have their rationale in our desire to know about parts of people’s lives that we cannot personally witness, like how they got to the situations in which we can directly observe them, what goes on in other situations in their lives where we are not participants.
Further, there are special motivations for self-deception, which make relying on one’s own experience or life perhaps more problematic, as a source for making generalizations about social life, than using data on what others experience when we, the researchers, are not present.
In various of my methodological writings I have argued that methodological questions often proceed from a self-defeating focus on an end state of perfect knowledge, which is the rhetorical form for methodological discussion in quantitative research. Researchers who work with quantified data often ask, on a range of 0, no relationship, and 1, perfect correlation, where do my findings lie? Do I reach or exceed a standard of .05, or .01 significance? The use of an image of an end state of perfect knowledge is a helpful convention in fixed design, quantifying research. But it is inappropriate and, worse, self-defeating when used in ethnographic, case study, and other forms of qualitative research.
The appropriate perspective on evidentiary questions that is productive in qualitative research is not only comparing the current state of knowledge or evidence to a future, unobtainable, perfect state of knowledge but comparing what we have learned to what we knew before we started the research. How far have we come?
Now, all the questions you have raised can and should be addressed to ethnographic or qualitative research, but the pragmatically helpful way to raise them is as guides to making further progress. It isn’t useful to agonize about the possibility that an interviewee is inventing a version of what he did in a robbery, for example. It is useful to understand that an interviewee might be making up a story, and to ask questions that make it more difficult for the interviewee to invent an account. If, for example, your questions ask about ‘how’ or sequence... what did you do next, and after that, and after that... rather than about motives, the interviewee is drawn back into the scene and finds it harder to make up a fiction than if you ask “why”, as in “why did you do it?”
Put in other words, you should (almost always) never ask someone what their motivations were. Figuring out why is not their work, it’s your work. What they respond with, when you ask them to tell you “why” they did something, or their motives, will almost always be designed to make a favorable impression on you there and then, in the interview. So, ask “how” not “why”. That takes them back into the scene, and most of the time you will get a very different understanding of what happened. Then, after you collect data on “how” actions were done (like robberies), you ‘test out’ various hypotheses about motivation. Where were the robbers when they started the course of action that led to robbery? In what situation, with whom, doing what, in what kind of ambience, did the robbery project begin? And what did they do after the robbery? Feed the family, help the less fortunate, party?
So, for each of the questions you raise, there are “tricks of the trade,” ways of conducting research that enable us to make progress from our initial state of knowledge. And that’s how we make a contribution to knowledge, at least if we are well-read on what others have found, so that our learning is also an advance in the knowledge of the community of researchers.
Other tricks of the trade include: not relying on the offender’s account alone but getting multiple versions from different people present, including victims; going to the offender’s neighborhood and finding out the relationships he has with others there ... all of these are matters for shaping a research design, where and how you spend your research time. You won’t get to a final state of perfect knowledge, but you will discover that what you had understood about the person or the event needs revision. And that’s what you have to offer as a contribution to knowledge.
It’s actually easy to respond to these questions when we are discussing a given research study. It’s hard to answer them in the abstract. That’s what’s different about qualitative research. Quantitative, fixed design research is shaped to answer methodological questions in the abstract. A sample is good or bad, according to the procedure of randomization used and the size of the sample relative to the population sampled, whatever the substantive questions may be. Quantification, numbering, counting is an abstraction process. If you try to answer methodological questions about qualitative research in the abstract, you’re using the wrong set of standards and you’re torturing yourself unnecessarily. But that means there’s not much more I can say in the abstract that will be helpful responses to these questions.
I won’t repeat what I said above about the difference between “how” and “why” questions, which your question here again raises. But I would emphasize the difference between asking someone how practically they did something and asking them why they did it. You as a researcher can ask a subject to review and correct your description of what they have told you about how they acted, but it’s usually pointless and misleading to ask them to review your account of “why” they did something. That account, in social science, will be based on your review of what you have learned about many subjects. No one subject will have your experience, data, competency to review your explanations, unless they work with you as a collaborator.
If you look at “Pissed Off in LA”, you will find the interview guide in an appendix. You will see that I asked students to report on “how” not why events developed as they did.
What is the visible unconscious when drivers become angry while driving? It is the relationship they are maintaining between their feet on the car’s pedals, their hands on the steering wheel, their attention to a projection of themselves as at the end of a trajectory moving on the road ahead, and their attention to how the projections of other cars’ moving trajectories are intersecting with their own. It’s very practical, not mystical, but a process that the driver cannot reflect on as he/she drives. I can’t watch my hands, my feet, the moving “dot” that is the projection of my own car on the road up ahead... all at the same time. I develop an aesthetic, a feeling, a sensibility/sensuality that I use to interrelate all these dimensions of my action, and that aesthetic itself, while felt, is not known in a self-reflective way, not until I get “cut off” by the intersection of moving trajectories up ahead, and draw my unconscious aesthetic embodiment of action to the surface of reflections in an angry outburst. I suppose you might call this a sociological version of magical realism. I must create a super-natural body to negotiate practical life; as I drive I create an aesthetic/sensual intermediation of myself and others, which I rely on but which, if I try to look directly at it, disappears from my grasp, like a ghost.
The visible unconscious exists in the embodiment of action, which, existentially, remains outside of our reflective perception as we act. I can’t look into a mirror and watch how I move my mouth as I speak, not without losing my train of thought. We cannot in real time observe the rhythms of movement we use as we act. Our focus inevitably is on how we are seen (or will be seen... which is the perspective we take when writing) by others. So constantly as we act we put or leave the fundamental, creative basis of our conduct in the dark. You can, of course, use recordings to review how you acted, but then you will be reviewing with and through a body that remains in the dark.
I hope to have the time to write at length directly on the visible unconscious. So far the emotions book is my most extensive attempt. But it is this understanding that distinguishes my work from the sociological interactionists, on the one hand, who generally ignore the unconscious (although it sneaks in with references to “tacit” aspects of behavior); and on the other hand, from psychologists, who treat the unconscious as invisible, to be inferred, not directly described. Freud was more subtle, distinguishing between an unconscious, which could be grasped in reflections, and a subconscious, which could only be described indirectly, through metaphors. But, I argue, we animate metaphors in and through our bodies all the time as we act, and that process is observable.
Yes, in a word. There’s a tricky connection—relationship that’s hard to grasp intellectually and to apply empirically—between the aesthetic/sensual and the transcending narrative meanings of criminal (or any other) behavior. Consider a physical fight. It’s not just a matter of perceptions, emotional vulnerabilities, anticipations of how others will see you now that you’ve been insulted. In order to attack someone physically, you have to create an embodied dynamic, a fluid sequence of movements shaped into a pointed thrust. How do you get there? How do you get yourself into that momentum? The attack, if and when it develops, will be, not an isolated moment but a trajectory, within which the other’s body is hit. That trajectory is a narrative project that goes beyond any moment which might be captured in a snapshot. To the police, to judges all that may matter is the moment of impact, which establishes a crime. To the victim, who may have been unaware of the attack in progress, also the moment of impact may be all that matters. But to the attacker the necessary condition for the attack is getting into a trajectory of movement, within which the moment of impact is a chance moment. So, to get at the causation of the crime, it’s a very different perspective, different from that of the victim or the avenging legal system’s officials, that the researcher must take.
In one sense, the answer is easy. Crime is what people, in forming their behavior, consider criminal behavior, which they do when they alter their conduct in anticipation of being caught and punished, when they sense they are being regarded as criminals by peers, when they work to be seen as threatening by victims, etc.
With the exception of some conspiracies and white-collar crime, officials in the criminal justice system focus on action of short duration. That is a great provocation for researchers, who can show the varying relationship of those short, officially relevant moments in the lives of alleged offenders and the longer courses of action in which people understand themselves to be engaged in criminal, or deviant, or transgressive life. (I’m hedging on “criminal” here because, once you go to ‘the other side,’ as Becker put it, “criminal” becomes a term of varying descriptive accuracy.) Alice Goffman’s
This is the rich turf that subject-based studies of crime can open up. We should not shut down the research possibilities by adopting the temporal limits of official concerns. Someone living as a “badass” or “wiseguy,” or con artist, or pursuing “action” is likely to be posing the question of when he is doing crime and when he is not doing crime, as a problem for his associates. The question of temporal duration is not one for us, as researchers, to conclude by fiat, at the start of research, by following some theoretically deduced logic. I can specify the formula, “criminal behavior begins, for the purpose of research, when the people we study begin to organize themselves, and shape their social relationships, for criminal purposes.” But that’s just a formula. What’s great about this perspective on crime is that it opens up research onto infinitely diverse, often inherently ambiguous, ways that people themselves sense this question in their lives and work out answers in interaction with others. It’s not unlike drinking alcohol and being an alcoholic. The ambiguities—you think you just drink, others think you’re an alcoholic; you think you’re an alcoholic so you don’t drink, so you don’t really get evidence any more that you are still an alcoholic; you realize that for years you were an alcoholic when you thought you were not—should not be excluded from research by a definition.
You touch on what may be the most difficult issue for criminology. If you consider white-collar crime, the relationship between poverty and crime starts to become problematic. I won’t venture comments about the corruption of the political and business classes in different countries, but I would suggest that bribery, in all its forms, is deeply woven into the maintenance of elite status on a worldwide basis. Some bribery is called lobbying or contributing to political parties. Some white-collar criminality is so routinely done that it is not considered crime, yet. Think of cheating on exams by university students. Universities are supported by the government. Cheating on exams keeps students qualified to receive a government benefit. In the US, we might wake up one day to read in the newspaper that college students, having been found to be cheating on exams (which is a massive reality), are being prosecuted criminally for defrauding the government. Recent scandals about parents bribing college officials to admit their children to colleges is a step in that direction.
Now, if you keep white-collar crime in mind, the relationship between poverty, or lack of economic opportunity, or discrimination, or lack of privilege, and crime, takes on a different complexion. Robberies are the kind of crime that poor people can do. And they are more often prosecuted than are, for example, tax frauds, which, according to studies I’ve seen, could be prosecuted for upwards of half of tax filers, especially for business owners, who can easily mix business with personal expenses. And who cost their victims (all who benefit from government programs that depend on tax receipts) more than robbers cost their victims.
I would make two points about the presumed relationship between poverty and crime. First, it is sustained by a trap that left or progressive social thinkers fell into from the earliest days. Somehow social scientists think that by arguing that poverty causes crime, they are going to get support from the public for reducing poverty, rather than support for those who would discredit the poor as criminal. Sociology seems unable to escape that trap.
Why? That’s the second point. Avoiding the trap requires much more creative research than researchers will take on. Consider: to argue that poverty causes crime, criminologists rely on government statistics, government definitions of crime, government bureaucracies that produce for them the data they will use... to criticize the government’s policies on poverty. That’s the trap. Or worse, qualitative researchers do not use government statistics, they study a few poor people in detail and rely on the reader’s stereotypes/prejudices to understand that there is a general causal relationship between poverty and crime.
How refreshing it would be to go beyond scandals (like Odebrecht, the Panama Papers, etc.) and show how subtly and diffusely crime, understood as behavior people do in a way that guards against its perception as criminal, is woven into middle class and elite ways of life. I think then we might finally begin to escape the trap of presuming a relationship between poverty and crime. And we might begin to see how what is officially regarded or punished as crime is part of an ideology, which criminologists and progressives generally if inadvertently help sustain, that keeps the poor morally discredited and the elite protected by a culture of deference.
The way out of the trap is not through research that shows that the poor are innocent or criminal only by necessity—why think that the poor are more saintly than the non-poor?; nor that the police falsely accuse the poor. It is through research that show the “angles” and “connections” and deceptive economic strategies of the middle class and rich, and that shows what the police do not do.
Actually, my current book project, which is a study of neighborhoods in Hollywood as they took shape and changed from 1870 to 2010, is very much about the different meanings and impacts of crime in social areas that differ greatly in socio-economic level, ethnicity and immigration status. Yes, it’s a very rich area for research.
I would take this opportunity not to summarize my work but to draw attention to issues in the area you address that I find to be real, significant, and virtually untouched by research. One is the layered nature of crime in a city’s public life. Unfortunately in the US, polemical debates about the “Broken Windows” theory have obscured the way that different forms of criminal activity become layered on others in city neighborhoods and streets. When, in the late 1960s, pornography entered publicly accessible life in the city, in the conversion of neighborhood theaters to show places, prostitution street walking and the open marketing of contraband drugs on nearby streets soon followed. Conversely, as internet technology moved pornography from the city’s public venues to private homes, prostitution and open drug marketing declined. The relationship is not a simple cause and effect but there appears to be a layering process, in which one form of behavior regarded as criminal or deviant gets layered on others. How does that happen? Does the layering reverse?
Second, criminology needs to exploit the potential in the area of social research and theorizing known variously as game theory, prisoner’s dilemma, tragedy of the commons, social cost, or problems in the organization of collective action. Economists and political scientists and social psychologists work with these ideas, and criminologists might also. Consider the collective effects on a neighborhood of street robbery in the area. On the one hand, a pattern of robbery in an area discourages personal investment. I’m thinking of a particularly painful account by a farmer in Oaxaca—someone I knew in LA as a brilliant crafts person but who longed for life on his property in Mexico—who, in taking his season’s crop to market, loses it all to highway bandits. Compare the value of what the robbers get, which probably is a discount from legal market price, with the cost to the farmer, which is not only lost profits but the disincentive to plant again, given the likelihood that his investments will be again rendered worthless.
So here’s one important calculation that criminologists might make. In some forms of crime, the offender gets less than the victim loses. In other forms, like tax fraud or corruptly inflated invoices submitted to government agencies for payment, what the offender gets equals what the victim loses. In other forms of crime, the calculation is more complex. A contractor bribes an inspector to ignore defects in construction, so that the contractor can save a significant amount of money, an amount beyond the size of the bribe; and the building collapses, causing serious injuries to multiple victims.
In street robberies, insofar as the offender does not rob cash, he typically gets much less than the victim loses. The purse snatcher throws away the purse and many of its contents, but the victim must pay to replace them all. So, that is a persistent undermining of life in neighborhoods with high robbery rates. It’s in effect a shrinking of the overall pie in the neighborhood. Not just a transfer of assets from some to others, but an interaction which produces less value than before it began.
On the other hand, a high robbery rate keeps potential buyers and renters out of a neighborhood. So, in poor areas with high street crime rates, the individual offender benefits all renters by, in effect, keeping their rents down. For decades in US cities, there have been stark differences in real estate value between areas in close proximity. What we call “gentrification” is often the overcoming of these differentials. The differentials are sustained by a culture about the neighborhood, a reputation about the area, that is sustained by crime. In such areas, crime is a way that the selfish lone offender “gives back” to the local community.
So, yes, let’s continue to develop research on the relationship between crime and the fabric of local life in cities, but let’s make bridges to other, economic ways of thinking to develop the possibilities.
It sounds like you have made a very good start on this work. Once individual criminal acts have become part of the fabric of social life in a city area, people take that fabric into account as they shape their behaviors toward each other. What I like especially about your examples are the unstated, “pre” assumptions underlying the interaction. The street vendors, without saying so explicitly, anticipate that potential customers will anticipate that they are approaching them to start a crime; and so they negate, in the way they approach potential customers, a line of action that never materializes. These examples begin to show the three dimensional, deeply rich, never stated but always relied upon fabric of street life in some city areas. I’m eager to read more on this research!
The most challenging aspect of
Actually I have done bits of work along that line, with an essay on “intimate massacres,” which covers a sub-set of the attacks that in popular culture are lumped together as “rampages” or “mass shootings,” and the explanation of episodes of anarchy, which popular culture discusses as “riots” or “rebellions.” My contribution there was to develop the concept of “epiphanies of invisibility” which I found at the turning points into and out of the events that were discussed as the “Rodney King riots” or “uprising” in Los Angeles. I was living near some of the areas that became anarchic in a few days in the spring of 1992, I knew people living in several of the neighborhoods where looting and arson were developing, and, without then having any idea what I would do with the data, I observed on the streets and talked with residents about their experiences. An epiphany of invisibility is a fragile state in which people understand that so many people are so visibly committing crimes that, effectively, none is visible to law enforcement or to each other as criminals. It is this causal contingency that explains why, when background conditions of inequalities, racial oppression, histories of police brutality remain constant, episodes of anarchy rise and fall within, typically, 3 to 5 days.
I had thought to write another volume along the lines of
When you ask about the relationship of criminological research and politics, perhaps the most significant theme is not the political implications of the studies that are published but the lack of a robust research program on white-collar crime. The upshot of this imbalance is a persistent association of immorality with poverty, which really makes no sense, but which seems to be irresistible not only for the right but also the left, the latter taking street or common crime (tangible property theft, interpersonal violence) as indispensable for arguing publicly for government aid to the poor and to minorities. Sociologists seem inevitably to get sucked into the rhetoric of looking for the “root causes” of common crime, as if they will somehow win the battle of controlling the political interpretation so that the public understanding of the relation of crime and poverty will be helpful rather than hurtful to the poor.
The most challenging task for criminology is not opposing repressive Right wing governments. Although that is an enormous challenge for social researchers, people who are not social researchers face that challenge as well, and often in more desperate ways. For social researchers the greatest challenge—the most radical step—is to break free of conventional definitions of crime and deviance, and to study crime without the presumption that it is concentrated among the poor. That intellectual challenge is more specific to the research community. In my lifetime, there was a brief phase, in “the Sixties,” when radical breaks from conventional thinking were possible. In social research on crime, the works of Becker, Kitsuse, Lemert, Matza and some others came close to launching a program of studies that would be free of conventional moral associations of crime and poverty, but that possibility, which ran along with the development of ethnomethodology, and was sustained by trends of novel thinking in popular culture (e.g., Ken Kesey’s Cuckoo Nest), the civil liberties/anti-psychiatry movement, gay liberation, the questioning of the ontology of racial identity, etc., has fragmented into isolated social movements and as an overall context for creative thinking, disappeared. I wish someone would study how that historical moment of intellectual liberation from conventional morality emerged and closed down.
Now I’ll try to respond more directly to your question about studying robbers and other street criminals (street level drug dealers, property thieves) from the offenders’ perspective. Recently I have written a critique of “hot potato criminology”. Today, when social researchers closely describe the lives and criminal ways of offenders, they are well-aware of the political risks they take. Unless they laden their writing with political boilerplate (rhetorical passages that are irrelevant or even inconsistent with their data but identify the author with and link the text to a particular camp) that makes their texts unusable by those who would advance repressive authority, ethnographers of crime may be attacked as “blaming the victim” and serving the enemy.
Alice Goffman’s book,
Still, in the process she identified what is arguably the most important next step to take in criminological research, which is into the determination of the collective as opposed to individual nature of the phenomena. Her subjects were especially “sticky” people. It was virtually impossible for others to stay close to them without becoming implicated in their criminality. Her appendix, in which she describes herself as drawn into what appears to be a murder plan, is just one in a wealth of evidence that shows how family members, lovers, neighbors, service providers, and others either become aides to robbery, contraband drug dealing, and associated violence, or must turn on the offenders and cooperate with the criminal justice authorities.
Not all lines of criminal behavior are equally sticky. Some crimes can be carried out quietly, hidden from family, friends, even from victims who may not realize what they have lost. Perhaps, in the current Brazilian context, the challenge for crime researchers is with the especially sticky forms of crime in both the poorest neighborhoods and in the most elite echelons of society. The cost of naturalistic investigation of sticky crimes at either social extreme is extreme vulnerability to personal political attack.