Justice and Judgment Among the Whomever: An Anthropological Approach to Judging

John Conley - University of North Carolina Law School

Posted in , , , , , ,

Prior to the conference, the organizers asked me for my thoughts on how an anthropologist might approach the problem of studying judging.  Those thoughts follow.  I have subsequently reflected on the discussion at the conference itself, and I conclude this essay with those reflections.

When I think of “judging” as an anthropologist, I think of two of the classics in legal anthropology:  Paul Bohannan’s Justice and Judgment Among the Tiv 1 and Max Gluckman’s The Judicial Process Among the Barotse.2 Both of these books were ethnographic studies of African tribal courts during the latter days of British colonial rule.  The tribal judges were prominent in both books, and when I teach them in law school I ask students to compare the tribal judges’ roles and conduct to what they know about contemporary American judges.  Like American judges, these African judges sometimes maneuvered the disputants into settlements and sometimes issued final judgments.  Unlike American judges, they never wrote opinions.3 So one could understand them only by watching, listening, and attending to all aspects of their behavior, including such apparently trivial details as how they arranged the seating in the open-air courtroom.

I also think—though far less grandiosely, of course—of the work that Mack O’Barr and I did in American small claims courts in the 1980s and 1990s.4 Like their African counterparts, the judges we watched in these courts persuaded, cajoled, threatened, and, when necessary, judged.  Nothing except the basic terms of a final judgment got committed to writing.  Moreover, again as in the African courts, the small claims judges had broad and largely unreviewable authority to impose whatever “law” they saw fit.  In one memorable case that involved a dry cleaner’s liability for a damaged suit, the judge decided to await and abide by the ruling of that august supranational tribunal, the International Fabricare Institute.

Regardless of any methodological preferences or preconceptions we might have had, the absence of written work product forced us—as it had forced Gluckman and Bohannan—to focus on the performance aspects of judging.  Exactly what did the judges say?  Just how did they say it?  What was their non-verbal behavior like?  And a good thing, too: this enforced focus on behavior led us in directions we never would have thought of on our own.  We found a range of judging styles and approaches; in anthropological terms, each courtroom had its own legal culture, created by the judge.  Within these mini-cultures, we found curious echoes of such jurisprudential schools as formalism and realism, as well as applications of the psychological concept of procedural justice.  This is not to say that the judges were consciously enacting these theories, but enacting them they were.

This focus also led to another set of findings, all derived from an initial observation that courtroom interactions are a form of conversation.  (This observation was influenced by an earlier study of British courts, Max Atkinson and Paul Drew’s Order in Court.5)  Practitioners of various forms of “ethnomodology” (including conversation analysts, linguistic anthropologists, and the many species of discourse analysts) have long demonstrated that “doing” conversation is an enormously complex social task, one that requires detailed rules.  Moreover, ethnomethodologists argue, these and other things that we think of as social “rules” are not rules in the sense of preexisting principles that need only to be applied.  Rather, they are “phenomena of order,” apparent rules that people “attend to,” and that “emerge” over and over again as people negotiate concrete social situations.  For example, two separate discourse analyses of actual jury deliberations (by Doug Maynard and John Manzo, in 1993,6 and Robin Conley and me, this year7) have shown, respectively, how principles of justice and standards of proof are produced by jurors in the course of their conversations.  In the small claims courtroom, O’Barr and I saw emergent rules for such judging problems as displaying authority, “doing” due process, showing empathy, and delivering bad news.

But why would these approaches have any relevance for those who study “real” judges?  In a current draft of a paper on the practical impact of “landmark” decisions, Professors George, Gulati, and McGinley8 suggest that the crowd divides into two general camps:  the legal scholars, who parse opinions; and the political scientists and their allies, who count and otherwise slice and dice outcomes.  Why might either camp care about methods of studying judging developed while sitting under a tree in Nigeria or in grimy small claims courts around the United States?

Because “real” lawyers and their clients might care.  The most striking observation made by George et al. is that practicing lawyers and judges as well as potentially affected citizens may not define “landmark” in the same way as academics.  In the case that George’s group focused on, what the eggheads expected to be earth-shaking barely registered on the ground.

This leads to a corollary idea more directly related to judging:  practicing lawyers may define “good” and “bad” judges in very different ways than academics.  Law professors tend to praise written opinions that are logically tight, that deal appropriately with precedent, and, most of all, that validate the professors’ policy preferences.  The authors of such opinions are “good” judges.  Similarly, quantitative studies tend to exalt judges who write lots of opinions (some researchers prefer long ones and others short ones), who rarely get reversed, and who get cited often.

But talk to practicing lawyers and you might get a different emphasis.  (I base this generalization on my own six years in full-time practice and 26 as of counsel to a firm and student of the profession.)  First, reversing the priority of their status-conscious academic betters, lawyers may focus more on trial than appellate judges because of per capita impact on their practice.  They are often more interested in how law is administered in their cases than in what law is made in “important” cases.  Second, my guess is that practitioners would identify a much higher proportion of “bad” judges.  If, for instance, one were to rely solely on the testimony of former law clerks now in academia, all federal judges would emerge as hybrids of Oliver Wendell Holmes and St. Thomas More.  But practitioners see some of these same judges as, in varying combinations, lazy, stupid, and mean (to quote my brother, a practicing lawyer).

In identifying “good” judges, practicing lawyers are likely to ask such questions as: How does s/he behave in court?  Does s/he treat lawyers, litigants, and witnesses fairly and courteously?  Can s/he manage a trial competently? Does s/he actually listen or does s/he prejudge things?  Will s/he decide motions in a reasonable period of time, or does s/he ignore them in the hope they’ll go away?  How smart is s/he?

These judicial qualities, I suspect, are very important to those who must live with the judges.  But they do not necessarily emerge from written opinions, nor from aggregations of outcomes.  They would become evident only through an inductive, ethnographic approach to judging that paid particular attention to language.  I wonder what other “phenomena of order” might emerge.  I suspect that important clues to the legitimacy of the legal system might reside in the details of courtroom interactions, and that a significant, “emergent” notion of justice might reveal itself as well.  The point is that, based on anthropology’s experience, including my own, I would not know what hypotheses to test a priori, but would instead expect the unexpected.

An aside: the legitimacy issue was brought home to me when I made my first visit to the European Court of Justice this past summer.  With a group of students, I attended a hearing in a case about import duties (a private company versus the European Commission).    The lawyers argued in German, the judges spoke in French (badly, since none came from France), and I listened alternately in French and English.  The lawyers read prepared remarks and the judges asked no questions; their only utterances were organizational and perfunctory (“We’ll hear now from Herr X”).    Instead, they stared with glazed eyes or flipped through the parties’ written submissions.  I could not imagine what value this presumably expensive exercise had for anyone (except the private-sector lawyer, of course).

After the hearing I discussed it with an ECJ staff lawyer.  I asked, in more polite language, “Why bother with this charade”  I suggested that the court learned nothing and the parties did nothing to advance their cases that couldn’t have been done better in writing.  He didn’t disagree, but said that hearings helped to legitimize the ECJ in the eyes of litigants and, more broadly, to inculcate EU citizenship.  I could not imagine how.  But I thought that if I had been paying closer attention—paying ethnographic attention, as it were—and had a chance to study and restudy my own, linguistic transcript of the hearing, I might have been able to figure out how the court and its constituents were “doing” legitimacy.

All of these considerations lead to the methodological question of how an ethnographer would study “real” judges (federal district judges, let’s say).  One thought might be to collect trial transcripts of cases that had been appealed and do discourse analysis.  That would be interesting.  However, from a sociolinguistic perspective, that would be more a study of transcripts than a study of judicial behavior.  Court reporters perform a subtle kind of interpretation, and their work product is not what a linguist would call a transcript.  Moreover, trial transcripts miss much of what counts as judicial behavior, even in the courtroom.

The only ethnographic solution would be to do what Bohannan and Gluckman did in Africa, and O’Barr and I did in small claims court:  pick some judges, spend days or weeks in detailed observation of their behavior in court (I can’t imagine getting into chambers), tape record it all, and prepare and analyze the detailed, multi-dimensional transcripts that discourse analysts use.  In terms of permission, the watching part would be easy, probably a matter of constitutional right (meddling Institutional Review Boards be damned).  The recording part would require the permission of each judge.  Based on previous experience (in both small claims and, in graduate school, “big” court), some would agree and some wouldn’t.  All would insist on permission from lawyers and litigants.  Perhaps surprisingly, all that I encountered agreed.

Again based on previous experience, it would be a long and demanding project, but by no means undoable.  As models, compare not only my small claims work but Beth Mertz’s 2007 book, The Language of Law School,9which analyzed discourse in several first-year law school classes around the country (and, as I think about it, is an excellent illustration of every point I’ve tried to make here).  For the reasons already advanced, I think such a project would be eminently worthwhile, although probably in surprising ways.  I doubt I’ll do it, but I hope someone will.

Reflections on the Conference

Being an anthropologist, I approached the conference itself as an ethnographic opportunity—I don’t know any better.  It was another day among the “natives,” in this case an array of federal and state judges, trial and appellate, and some academics who are interested in studying them.  Two themes stood out.

The first was the sharp distinction between trial and appellate judging.  Trial judges engage in complex conversations.  Many times these conversations are live, oral, and public, as when a judge conducts a trial, a motion argument, or a sentencing hearing.  The judge must not only apply the law in real time, but must also manage the human interaction.  Most trial judges realize that their job in an adversary system is to stay as far in the background as possible while letting the parties produce the trial.  The lawyers are frequently repeat players with whom the judge will have to deal again and again.  This task—“doing trial judging”—strikes me as remarkably challenging, both intellectually and sociologically.  And it must be done while looking over one’s shoulder, since the whole performance is subject to public review, (sometimes unfair) recharacterization, and (sometimes withering) critique.

Appellate judges, by contrast, spend most of their time manipulating written texts.  As they described it at the conference, their work consists largely of reading briefs, memos, cases, and statutes; drafting, circulating, and reading opinions; and attempting to persuade colleagues in written (and, increasingly, electronic) communications.  Live conversations in the form of oral arguments and conferences with colleagues are the exception.  With the exception of the oral arguments, the work of appellate judges is done in privacy and relative anonymity.  When they do perform publicly, they—in contrast to their trial colleagues—are expected to be the stars, interrupting and hectoring the lawyers with questions that sometimes seem intended to show off as much as to focus the arguments.

These functional differences brought home to me the point that studying judging encompasses two quite different projects.  To study the performance of trial judges is to study public behavior and public documents.    It is thus inherently doable, even if extraordinarily labor-intensive.  Appellate judges, however, do most of their work in a secret world that is seemingly impenetrable to ethnographers or others who rely on direct observation. One can study only the judgments and opinions they are required to release to the public; the performance that underlies these carefully crafted documents is immune to scrutiny.

The second observation derives from the first: judges have difficulty being anything but judges, and the trial-appellate distinction carried over into the conference discussion.  The trial judges, by and large, conversed.  They asked and answered questions and seemed eager to have external, scientific perspectives on what and how they are doing.  The appellate judges acted like, well, appellate judges (with apologies to the exceptions).  They monopolized the floor, interrupted, and sometimes went on, and on, and on.  There was nothing unpleasant about this; on the contrary, the whole event was unfailingly cordial. It was just that the appellate judges stayed in courtroom character.  When they discussed their behind-the-scenes work, they saw few problems that were worth studying. They have things under control (aside from the rare colleague who doesn’t work well with others and writes gratuitous dissents).  We academics will never get in, of course, but what would be the point?  (And academic idolatry of federal appellate judges only eggs them on.)

So I came away from the conference with mixed feelings.  Trial judges want to be studied, and there are many ways to do it, with ethnography well-positioned to play a role.  But at the appellate level, it seems that we will be limited to the analysis of outcomes.  As it usually does in legal scholarship, “empirical” will continue to mean “quantitative.”


Copyright © 2010 Duke Law Journal.

John Conley is a professor at the University of North Carolina-Chapel Hill Law School

This material is based upon work supported by the National Science Foundation under Grant No. SES-0946437.

  1. PAUL BOHANNAN, JUSTICE AND JUDGMENT AMONG THE TIV (Waveland Press, Inc. 1989) (1957).
  3. See id. at 250 (“{T}here is no written corpus of judicial precedents {among the Barotse}.”); id. at 288 (noting that “the absence of written precedents . . . limits judicial logic”).
  6. Douglas W. Maynard. & John F. Manzo, On the Sociology of Justice: Theoretical Notes from an Actual Jury Deliberation, 11 SOC. THEORY 171 (1993).
  7. Robin H. Conley & John M. Conley, Stories from the Jury Room: How Jurors Use Narrative to Process Evidence, 49 STUD. L. POL. & SOC’Y 25 (2009).
  8. Tracey George, Mitu Gulati & Ann McGinley, The (Relative) Unimportance of Case Law, 103 NW. U. L. REV. (forthcoming 2010) (manuscript at 1), available at http://weblaw.haifa.ac.il/he/Events/eveFile/The%20Relative%20Unimportance%20–%20Gulati.pdf.


  • I’m not sure I agree with your characterization that “acting like judges” leads to a more conversant group of trial judges. Wouldn’t appellate judges, who have to convince colleagues, be expected to be more conversant than trial judges, who make unilateral decrees? Granted counsel are likely to re-appear before a trial judge, but so too (and to greater degree) are colleagues likely to re-appear on a panel.

    Posted by dbl230, 03.31.10 

Post a Comment (all fields are required)

You must be logged in to post a comment.