Some thoughts on Goffman’s On the Run

After initially receiving mostly adulatory press (which it continues to receive, including a Malcolm Gladwell essay in the current New Yorker), the sociologist Alice Goffman’s urban ethnography On the Run: Fugitive Life in an American City has started also to receive more negative attention. At the New York Times, Alex Kotlowitz raised some tentative ethical concerns in an otherwise positive notice; at Slate, Dwayne Betts sharply critiqued the book’s simplifications, in his view, of urban communities, while also wondering about the evidentiary support for some of its empirical claims; and most recently, at The New Inquiry, Christina Sharpe has called into question the morality of the entire project, positioning it within the history of “a sociological tradition that subjects black life to scholarly scrutiny” and, more specifically, within West Philadelphia’s “long and contentious history” with the University of Pennsylvania. (Goffman began researching the project as a Penn undergrad.)

I read the book earlier this summer and thought it might be useful to round up some of the marginal notes I made into a blog post (some of which I think I also discussed or referenced on Twitter while I was reading, although certainly not comprehensively). I should say that I found the early press around the book off-putting, insofar as the media lavishly praised Goffman for having the audacity to spend a few years in a neighborhood that plenty of people just, well, live in, and insofar as this coverage trafficked in the shock appeal of a young white woman living among black men, which is to say, trafficked in any number of longstanding white supremacist tropes, whether intentionally or otherwise. Still, those were issues with the coverage, not the book itself, and given the attention it’s received, and given that my academic focus encompasses criminal justice and policing, I figured it was something I should read, whatever I ended up making of it. 

By the end, I found it occasionally thought-provoking but generally problematic, and to be honest, I wasn’t really sure what to make of it. My formal academic training is in law and history, which have different standards of evidence and argumentation than sociology and especially than ethnographic approaches to sociology, so I wasn’t always sure where the concerns I had about the book were reflective of my own disciplinary perspective and where they would also hold true by the internal standards of sociology. So, I put the book aside, although I have been glad to see a more critical conversation developing that I hope can help me better understand how to situate the work within its own discipline. This type of urban ethnography — whether it should be done at all, by whom, how, etc. — is the subject of longstanding and contentious internal debates within sociology and anthropology (see for instance Loic Wacquant’s critique of a series of three late ’90s urban ethnographies, including one by Goffman’s dissertation adviser Mitchell Duneier), debates with which I’m familiar but by no means expert in, so I will be especially curious to follow how Goffman’s book ultimately gets situated within those conversations.

But Goffman’s book is aimed not only at sociologists but also at a general audience — the University of Chicago Press has marketed it similarly to the way a trade book would be marketed, with NPR interviews and the like, and it’s being released in paperback by Picador — and so, in that sense, I do feel qualified to ask some questions about the book, if not the underlying study. I appreciate the critiques that Betts, Sharpe, and others have made of the book and the ethics of carrying out this project at all; they raise serious questions that are important to grapple with for scholars who do this kind of work and/or who advise students embarking on ethnographic projects. My questions here generally fall into the category of more local queries about the book’s use of evidence. In other words, even on its own terms, does the book adequately support its arguments and claims?

The cause of many of my specific questions below is the overall issue that the book is unevenly footnoted — some empirical claims get a footnote or an in-text source reference of some kind, but a lot don’t. Perhaps this was an editorial decision by the press to make the book more saleable to a general audience, or perhaps, again, I’m imposing the (perhaps overly) comprehensive footnoting norms of legal scholarship onto a different genre, but in any event, the inconsistent sourcing makes the book’s claims difficult for other scholars to evaluate or follow up on. (I should also note that the prior undergraduate and graduate work on which the book is based may well have more thorough footnoting that addresses the below concerns for academic purposes, but that wouldn’t, in my view, resolve the concerns as to general-audience readers who only encounter the book.)

Hospitals as adjuncts to law enforcement?

One of the book’s central constructs is what Goffman calls “concerted avoidance” (37) — the tendency of men with criminal records and other people with reasons to fear the police to avoid public places like hospitals, worksites, and funerals. In what I found perhaps the most troubling anecdote in a book full of them, Goffman reports that she witnessed a 14-year-old boy who “broke his arm while running from the police” have a neighbor set his arm for him at home because he felt he could not go to the hospital (152-53).

However, I found unclear what precisely Goffman’s claim was here: that her subjects avoid the hospital in the aftermath of specific encounters with the police, or whenever possible, or literally at all times. I am familiar with police visiting particular individuals in the hospital to question them about shootings or other incidents, and it may be that the police visit the hospital on the pretext of those kinds of visits so routinely that, in effect, they maintain a near-constant presence at the hospital. But it’s not clear that that’s exactly what Goffman is claiming, because at times, she seems to suggest that the police routinely run warrants checks on literally the entirety of the hospital visitor log for some unspecified period of time (see 34: “The officers told me they had come to the hospital with a shooting victim who was in custody, and as was their custom, they ran the names of the men on the visitors’ list”). If so, I wondered, why would hospital staff routinely turn over the entirety of their visitor records, absent some specific investigation? (For that matter, I wondered, what type of visitor records do hospitals maintain exactly, for how long, and to whom and under what circumstances are they available generally?) How would such a practice interact with medical ethics and health privacy law, not to mention just the mundane realities of running a hospital without constantly interrupting whatever the staff are doing so they can turn over paperwork to visiting law enforcement?

As Betts notes in his Slate review, the book doesn’t name any specific hospitals where such practices go on, and when Betts called up a Philadelphia police officer to ask, he denied that police would do this at all. Perhaps that’s to be expected, but I would have also liked to know if Goffman spoke to anyone at the hospital about this issue. I would imagine that at least some doctors might be troubled at the idea of deterring whole swaths of a community from seeking needed care. Alternatively, perhaps Goffman’s aim was to suggest simply that her subjects believe that the police conduct this type of ER surveillance more than the police actually do, which would still be an interesting finding about the effects of pervasive surveillance but would be a different finding.

Anyway, when I looked into this a little further I was glad to note that Sarah Brayne, a sociology graduate student, is apparently doing further fieldwork on police information gathering in an urban ER, and will be curious to read her findings. (Brayne also has an article doing systematic quantitative testing of Goffman’s “avoidance” theory.)

Police brutality and admissions thereof

Both the Kotlowitz and Betts reviews point to some claims about the Philadelphia police for which Goffman does not identify a source, such as the claim that they use a software program inspired by a documentary on the Stasi (see 20 — Goffman writes that this program was developed by a “local FBI officer” but I think she is implying it’s used by the Philadelphia Police Department; if so, she doesn’t explain the mechanism of transfer — whether through FBI sharing of technology or a training program or a joint task force or whether her claim is that this is now simply standard police software).

Here are two more policing-related claims about which I had questions. First, Goffman says that Philadelphia’s warrant enforcement “consisted of two men who sat at a desk” until a new “Warrant Unit” was established in the 1970s (18-19). She doesn’t provide a footnote or source for this information. I don’t doubt it, but I wanted to know whether there was a particular book, article, or other source where I could read more about this transition.

Second and more concerning to me, Goffman reports: “In interviews, Warrant Unit officers explained to me that … violence represents official (if unpublicized) policy, rather than a few cops taking things too far … a number of officers told me that they have orders from their captains that any person who so much as touches a cop ‘better be going to the hospital'” (72). I don’t think of myself as particularly naive about police officers, and I don’t doubt that police might have this attitude or say these kinds of things among themselves, particularly in Philadelphia. I am, however, surprised that “a number of officers” would volunteer such a politically and potentially even legally problematic direct quote to an undergraduate researcher (as Goffman was during some parts of the study) or even to a graduate researcher (as Goffman may have been during these interviews, although I don’t think she ever specifies when in the project they occurred). More information would be helpful: How many interviews did she conduct with police officers? When? How did she secure these interviews? What specific promises of confidentiality, or understanding of the project, did she give these officers? Compared to her exhaustive account of how she got to know her main subjects, her interactions with the police are not much addressed in her otherwise lengthy Methodological Note.

My confusion about that quote deepened when I noticed that for some other quotes in the book, although not this one, Goffman clarifies that she is paraphrasing. For instance, in a footnote to a different quote, she makes the startlingly casual aside, “As with other quotations, it should be taken as a close approximation of the wording and sequence, not a recording” (274 n.4). Is that how we should take all quotations in the book or just some? Why is this important caveat buried in a footnote? Because most of the book’s quotes aren’t footnoted at all, there’s no way for a reader to know which ones fall into this category and which are transcriptions, and moreover, Goffman says that she has destroyed all of her field notes.

Who is the book’s audience?

While reading, I grew increasingly frustrated by my impression that the book is aimed at the naivest possible audience. On the one hand, as noted above, it makes some striking claims without identifying a source; on the other hand it laboriously defines for the reader such obscure urban jargon as “blunt” (“a cigar hollowed out and filled with marijuana,” 121). The effect is to imply a reader who will believe anything because they know nothing. As Sharpe suggests in her review, the book implicitly assumes an audience that has no previous familiarity with the phenomena under discussion — it implies, as has much of the media coverage around it, that Goffman is revealing a world unbeknownst and even unknowable to readers until she revealed it. In other words, it implies that the book likely won’t (or more perniciously, can’t?) have any readers who come from or live in neighborhoods like the one under study. But even more, it also implies that it won’t have any readers who already know anything, for whatever reason, about policing or the criminal justice system. I am not sure if assuming a completely naive audience is just a conventional trope in ethnography, and would like to set aside some time to read some other ethnographies on subjects that are stereotypically more “familiar” to a university press’s imagined readers, such as Paying for the Party — on Midwestern sorority girls, who have probably also encountered some blunts in their time — to see how and whether the tone is different. But I’d be curious to hear from readers more thoroughly read in this genre. Even if it is a conventional trope, it seems especially problematic when combined with the already less-than-optimal racial optics of this project.


The book introduces its argument as follows: “A new social fabric is emerging under the threat of confinement: one woven in suspicion, distrust, and the paranoiac practices of secrecy, evasion, and unpredictability” (8) (my emphasis; see also 197: “a new and more paranoid social fabric is emerging”). As I understand it, the book describes events that Goffman witnessed as early as 2002 and 2003 (see 218-19 referencing events in those years), and mass incarceration more generally picked up long before then (it’s usually dated to the 1970s, sometimes earlier, and was certainly up and running by the 1980s and ’90s). Thus, I didn’t understand why the book is framed as an argument about a “social fabric” that “is emerging” (presumably now, in 2014). How long has it been emerging? When will we know it has emerged? Why, given the pervasiveness of surveillance that Goffman documents elsewhere in the book, doesn’t she think it’s already emerged? But, maybe this is just my historian’s obsession with periodization.

2 thoughts on “Some thoughts on Goffman’s On the Run

Comments are closed.