In a previous blog, I promised further comment on Steven Lubet’s ethical critique of Alice Goffman’s fieldwork in her 2014 book, On the Run: Fugitive Life in an American City. Lubet takes particular exception to Goffman’s account of driving one of her key informants around their neighbourhood with a loaded gun, looking for a member of a rival gang to target in a revenge shooting. In his book, Lubet places this alongside other reports from participant observers who had either participated in, or colluded with, criminal activity. He acknowledges that this is sometimes inevitable in maintaining the confidence of informants or documenting the impact of oppressive laws. His examples include WF Whyte’s participation in voting fraud or Laud Humphrey’s facilitation of gay sexual encounters at a time when these were criminal. The question is where a line is to be drawn and who gets to draw it.
Lubet’s position seems to be that only lawyers get to have absolute immunity from reporting past crimes – although not planned crimes or crimes in progress. While he is willing to tolerate a certain degree of ethnographic indifference to criminality, serious crime should be reported. In the book, he also criticises Goffman for not coming forward as a witness to two murders that she reports observing.
This is not quite as straightforward as Lubet suggests. Social scientists are not trained as lawyers or police officers to monitor everyday life for the existence of criminal acts. Indeed, it is well-established that most potential crimes are not turned into actual crimes for very good reasons of system management and peacekeeping. A classic criminological aphorism is that the most powerful actor in the criminal justice system is the front-line officer who makes arrest decisions. Everything else rests on this choice. Field researchers are not socially authorized to make such judgements.
When I was part of a team studying child protection services in the late 1970s, we discussed what we should do if we disagreed with agency decisions about risk or intervention. In the field, we rapidly came to understand that we had no real basis for displacing the professional judgements made by social service, legal and health professionals. It was simply our opinion against their socially authorized expertise. Because we worked across agency boundaries, we did occasionally have information that a key professional lacked – but we avoided disclosing this because the silos were important in understanding the system. I can only recall one exception, where a team member told a social worker that a visiting nurse had relevant information and should be asked for it, without disclosing the information itself. In this context, I do not see any basis for criticising Goffman for not volunteering evidence about two murders. It is for the police to find witnesses: if they fail to do this, or choose not to, then we have learned something useful.
Should Goffman have driven a key informant around with a loaded gun? Lubet thinks this is inherently a criminal act, whether the gun was ever fired or not. I see this much more as a judgement call. Since the incident never came to the attention of any law enforcement agent at the time, we have no way of knowing whether it would have led to an arrest, let alone a charge. A more experienced fieldworker might have been more cautious but we can reasonably ask whether it would be better to have a social scientist drive the car with the hope of discouraging a homicidal act – or another gang member who might expect it to be carried out. Only Goffman is in a position to make that call, knowing her relationship with the informant and the degree to which she might be able to influence him.
Lubet’s position is ethical rather than legal. In England and Wales, at least, there is no general obligation on any citizen to report a crime or volunteer as a witness, unless the incident is related to terrorism. Other jurisdictions adopt different approaches. France requires all citizens to report serious crimes and child abuse. Some US states mandate the reporting of child abuse. Anecdotally, I have been told that a few Western states retain 19th century laws about the duties of citizens to assist in law enforcement, which go back to the days when sheriffs assembled posses to pursue offenders. I have not verified this for the present blog. In England and Wales, merely being present at the scene of a crime is not sufficient to create a criminal offence of aiding and abetting or conspiring. There must be evidence of intentional encouragement or support, which would be inconsistent with the fundamental principles of participant observation.
The core issue, then, is whether social scientists have some kind of moral duty to act as whistleblowers or, more pejoratively, as snitches, grasses or stoolies. Lubet’s position appears to be that accommodating or condoning criminal actions is inherently wrong, although it may on occasion be tolerated. Many participant observers would argue that a greater good is achieved by not volunteering information about criminal actions in order better to understand them. It is not our job to second-guess the street-level decisions of law-enforcers
Wherever the line is drawn, all ethical decisions in research risk contributing to the creation of ignorance. Researchers may forego knowledge that would otherwise be of value to their society. Even within a framework of ethics regulation that makes ignorance structured and systematic, fieldworkers have to make street-level judgements about what IRB or REC policies mean in practice. The line is created by these decisions. The people who make them should be able to expect that any subsequent debate will respect the circumstances under which they have acted.