View source for Medieval Iceland and Modern Legal Scholarship

<!-- you can have any number of categories here -->
[[Category:Richard Posner]]
[[Category:Medieval Iceland]]
[[Category:David Friedman]]
[[Category:Bloodtaking and Peacemaking: Feud, Law, and Society In Saga Iceland]]
[[Category:William Miller]]
<!-- 1 URL must be followed by >= 0 Other URL and Old URL and 1 End URL.-->
{{URL | url = https://chicagounbound.uchicago.edu/cgi/viewcontent.cgi?referer=https://www.google.com/&httpsredir=1&article=5039&context=journal_articles}}
<!-- {{Other URL | url = }} -->
<!-- {{Old URL | url = }} -->
{{End URL}}
{{DES | des = Richard Posner takes a very different view of how Medieval Iceland works.  (He makes one mistake: government does not have a monopoly on force, it has a monopoly on legitimizing force.) | show=}}
<!-- insert wiki page text here -->
<!-- DPL has problems with categories that have a single quote in them.  Use these explicit workarounds. -->
<!-- otherwise, we would use {{Links}} and {{Quotes}}  -->
{{List|title=Medieval Iceland and Modern Legal Scholarship|links=true}}
{{Quotations|title=Medieval Iceland and Modern Legal Scholarship|quotes=true}}

{{Text |

BLOODTAKING AND PEACEMAKING: FEUD, LAW, AND SOCIETY IN SAGA ICELAND. By William Ian Miller. Chicago: University of Chi- cago Press. 1990. Pp. xii, 407. $29.95.
Professor William Miller of the University of Michigan Law School has written a learned and lively book on the society of medieval Iceland, with special reference to its methods of dispute resolution. He has written the book for two audiences. One consists of the tiny handful of specialists in medieval Iceland; the other of social histori- ans, academic lawyers, political scientists, and other social scientists interested in dispute resolution and in social control more generally (including the foundations of the state). For the second group, the relative simplicity - political, social, and religious - of medieval Ice- landic society lends fascination to the social institutions found in it because they are, as it were, observed under laboratory conditions.
I am a member of Professor Miller's second audience, and so will not try to evaluate the accuracy of his scholarship. I have no reason to question it. I read the two books of Icelandic scholarship listed in Miller's bibliography that seemed most like his book,1 and a compari- son of those two books with Miller's does not suggest, at least to this outsider, that Miller's scholarship is in any way deficient. His book is longer and more detailed than either of the others and contains as impressive a bibliography.
Miller's concern is with the period (A.D. 930-1262) in which medi- eval Iceland was independent. It is this period - between the settle- ment of Iceland (primarily by Norwegians) and the submission of the country to the King of Norway - that is depicted in the sagas, which are our main source of knowledge of medieval Iceland although there
* Judge, U.S. Court of Appeals for the Seventh Circuit; Senior Lecturer, University of Chi- cago Law School. - Ed. I thank David Cohen, Thrinn Eggertsson, David Friedman, Lawrence
Lessig, and Andrew Schapiro for their very helpful comments on a previous draft.
1.JESSEL..BYOCK,FEUDINTHEICELANDICSAGA(1982);JESSEL.BYOCK,MEDIEVAL ICELAND: SOCIETY, SAGAS, AND POWER (1988) [hereinafter BYOCK, MEDIEVAL ICELAND]. Byock, a historian, specializes in and has published a number of articles on medieval Iceland,
besides the two books. A recent addition to the pertinent literature is Henry Ordower, Exploring the Literary Function of Law and Litigation in Njal's Saga, 3 CARDOZO STUD. L. & LIT. 41 (1991).
1495
 1496
Michigan Law Review [VCol. 90:1495
are also law codes and other miscellaneous writings and some archaeo- logical remains. Those of us who remember the television series Monty Python's Flying Circus recall saga Iceland as a bleak, feature- less, monotonous land of drably accoutered and colorless people whose names are strange without being charming - recall it indeed as the quintessence of the dull. Professor Miller has succeeded in dispel- ling this impression by the liveliness of his style and the infectiousness of his own enthusiasm for his subject. That success is a tour de force in itself, for apart from, but of course not unrelated to, the unforgiving land and, as a consequence, sparse population, medieval Iceland lacked everything that made the Middle Ages so colorful a period in Western history. Iceland had no king, court, or nobility; feudalism was at best nascent, which is not to say that the society was egalita- rian, for some men were considerably more powerful than others and some indeed were little better than slaves - some were slaves.2 The Catholic Church was much weaker than in Europe - so weak that for much of the period of independence it was forced to adopt a policy of peaceful coexistence with paganism.3 There were no Jews or heretics. Art, architecture, music, ornament, and decor were nonexistent or ru- dimentary. There were no wars, no armies, no jousts, no chivalry, no courtly love, no philosophy or theology, no theater, and no literature other than the rather matter-of-fact sagas.4 There were no cities, no towns or even villages. There were no stone buildings. It was a simple
pastoral culture.
But of course it is the very simplicity of saga Iceland, a simplicity
almost (though not quite, as we shall see) unique in Western history, that is the fascination - in particular its political simplicity. It was not quite a stateless society; nor is a stateless society apolitical; but the Icelandic state was weaker than the weakest night-watchman state. The formal institutions of government consisted of courts and an as- sembly, which, like the ancient Athenian courts and assembly, were staffed by ordinary citizens rather than by professionals. There were no professionals, though there were law experts who in retrospect can be seen as proto-professional; in effect there were jurors but no judges
2. And this at a time when slavery had all but disappeared from Europe.
3. An unfortunate omission in Miller's book is that the nature of Icelandic paganism is no- where explained; the index entry for "religion" merely refers the reader to the entry for "Church." The omission is repaired in BYOCK, MEDIEVAL ICELAND, supranote I, at 137-64.
4. Professor Miller's translations make them downright prosaic, probably unfairly to the sa- gas' authors because I am informed on good authority that the prose style of the sagas - in their original Icelandic, of course - is elegant, and modem in its precision and brevity. Miller might reply that he is trying to be accurate rather than artistic. Yet when he has a saga character say, "To tell the truth, I could care less if they do each other as much harm as they wish" (p. 259), he makes it hard for the reader to believe that one is dealing with the peers of Homer. (The Penguin translation is a little better: "To tell you the truth, they can do one another as much harm as they please for all I care." LAXDJELA SAGA 173 (Magnfs Magndsson & Hermann Pilsson trans., 1969)). Although Miller praises the sagas for their subtlety of characterization, it cannot be said that he inspires a desire to read them.

 May 1992] SagaIceland 1497
(hence no appeals). The speaker of the assembly was the only salaried official in Iceland. There was plenty of death, including much chop- ping off of heads with axes, but at least there were no taxes (or virtu- ally none: the speaker was paid out of the marriage fee). There was, crucially, no executive arm of government and as a result there were no sheriffs, no police, no soldiers, and no prosecutors. All suits, in- cluding criminal suits, were prosecuted by private individuals, and,
what was the greater innovation in the art of minimizing government (for a number of societies that no one would call stateless have left prosecution even of criminal cases to private individuals5), all judicial
decrees were enforced privately, if at all.
Adjudication was not the only lawful method ofresolving disputes.
The feud was lawful, too, and so (less surprisingly) was private arbitra- tion, in which a dispute was submitted for binding resolution to one or more men selected by the disputants. Family was extremely impor- tant. Because people looked to both near and remote relatives for aid in enforcing decrees and conducting feuds, concepts of kinship were more elaborate and refined than in modem society. Power was impor- tant too. Powerful men were chieftains, to whom lesser men rendered services in exchange for support and protection. Here was the germ of a feudal system; but the elaboration of feudal duties, ranks, and obliga- tions that we find in medieval Europe was missing. Moreover, chief- tainships could be sold - an intrusion of the market that a true feudalist would consider horrifying.
Icelanders were great amateur lawyers and their law codes were as complex and ingenious as most of the rest of their culture was simple and monotonous. Njal, the hero of the best-known saga, was one of these amateur lawyers. The procedures for finding facts were more rational than those in force in most medieval societies, for little reli- ance was placed on supernatural methods. Yet the system was sur- prisingly inflexible on the remedial side. The only sanctions that the Icelandic courts meted out, other than for the most trivial infractions (which were punished by a fixed fine), were outlawry and lesser out- lawry. Outlawry made a man an outlaw in the literal sense: anyone could kill him with impunity. Lesser outlawry meant banishment from Iceland for three years. The inflexible character ofthe legal rem- edies made arbitration attractive, since arbitrators could impose fines or fashion what we would call equitable decrees, tailored to the partic- ular circumstances of the case. Refusal to obey an arbitral decree was punishable like other serious wrongdoing.
This strikingly decentralized system of governance survived for more than three centuries. No one knows just how much violence there was, but the society was not anarchic, though the institutional structure sketched above might seem a recipe for anarchy. Nor did
5. See, e.g., DOUGLAS M. MACDOWELL, THE LAW IN CLASSICAL ATHENS 237-40 (1978).

 1498 Michigan Law Review [Vol. 90:1495
saga Iceland lapse into tyranny, despite the apparent fragility of its institutions. Even feuds were governed by norms (for example, that the killing of an outlaw should not be revenged - should not, that is, occasion a feud) that despite the absence of formal sanctions were
obeyed with some, though far from complete, regularity.
All this - the system of law and order in medieval Iceland and its ancillary institutions such as kinship - Professor Miller sets forth
with admirable clarity and also with a wealth of detail, meanwhile judiciously separating fact from fiction in the saga accounts and lu- cidly recounting from the sagas numerous disputes and their resolu- tion by feud, litigation, or arbitration. The greatest strength of the book is its description of the structure and, above all, the actual opera- tion of medieval Icelandic institutions. Although Miller's focus is on dispute resolution, much of the book quite properly is devoted to describing the social background out of which disputes arose and against which they were resolved, including the system for the ex- change or transfer of goods (through gift exchange, sale, or raid), the subject of a superb chapter (Chapter Three). The second audience of which I spoke at the outset can hardly be assumed to be conversant with the structure of Icelandic kinship, yet that structure is basic to an understanding of the operation of the feud, the lawsuit, and arbitra- tion, and so receives a chapter too (Chapter Five). It is basic because none of the methods of dispute resolution could be implemented effec- tively without rallying potential supporters, and the principles of kin- ship were vital in determining who your potential supporters were and also whom among your potential opponents you might neutralize on
grounds of kinship (pp. 164-67).
Miller's former vocation as a professor of literature stands him in
good stead as he teases out the implications of the saga texts that are the principal data for his analysis, while his present vocation as a law professor may account for the superior crispness and authority of his description of the Icelandic legal process (Chapter Seven), compared to Byock's. 6 The only fault I have to find at the level of description, of narrative, as distinct from the level of analysis, is with the clumsiness of Professor Miller's scholarly apparatus; and it is with this point, pa- rochial or even pedantic as it might appear to be, that I begin my critique.
II
Miller has, as I have said, two audiences. His method of trying to satisfy the demands of the first, the specialist audience, for documenta- tion without losing the rest of us to tedium is inept. It consists of combining the "scientific" method of citing references - that is, citing
6. See sources cited in note 1, supra.

 May 1992] SagaIceland 1499
them in the text itself (or in the endnotes) by the author's last name and the date of publication and giving the full citation information in a bibliography at the end of the book - with long textual endnotes (sixty-three pages of small print, compared to a text of 308 pages). So there are really three texts: the text itself, studded with bibliographi- cal references and endnote call numbers; the bibliography; and the endnotes. Opening the book at random, I find myself on page 224, where in the middle paragraph I spot endnote call number nine. I go to the back of the book and flip through the notes until I find note nine to page 224, and there I find a reference to page 144, to which I repair only to find another endnote call number, which sends me to a note that contains bibliographical references in scientific form and therefore propels me to the bibliography. Who can read a book this way? Maybe medieval Icelandic studies is such a quiet field that an Ice- landic scholar would be refreshed to encounter a book in his field com- posed d la mode Derrida, but it turns out that many of the endnotes and many of the bibliographical references in both the (main) text and the endnotes are directed at the second audience, the general scholarly audience. The last two endnotes in the book are to game theory and to the comparative study of revenge, respectively, rather than to anything Icelandic. They contain three citations and so send one hopping to the bibliography again.
The scientific system of citation is designed for short papers with a handful of references; you can remember the names as you read, and when you get to the end you can see at a glance what works the author is building from and make a note of any work you want to look up. The system makes no sense for a long book with a bibliography of (at a rough count) almost four hundred separate works.
My bigger gripe is the endnotes. I have not yet read a book in which I thought it made sense to, put the notes at the end of the book rather than at the bottom of the pages. It is too great an interruption to be constantly turning to the back of the book. How much of an interruption depends on the number and length of the notes. Miller's book has more than four hundred notes, which means an average of more than one per page of text or, more vividly, means that the consci- entious reader will be turning to the back of the book four hundred times before he has finished. Once back there, as I said, he will often find himself flipping still farther back to the bibliography before he returns to the main text. The notes aren't short either, for remember that most of the bibliographical information is in the bibliography rather than in the initial citation; thus these are mainly textual notes. Four hundred notes in sixty-three pages means that each page of notes - pages in small print, be it noted - has on average fewer than seven notes, and that means that the average note is substantial. Some of them are a page or more in length, and by the time you've finished

 1500 Michigan Law Review
reading one of these notes and returned to the text you've lost the
thread of what you were reading.
I can imagine, though I don't recall ever having actually seen, a
book in which all the endnotes were (1) bibliographical and (2) in- tended solely to provide research leads for specialists. And in such a book the endnote format wouldn't be troublesome. Miller's book is not of that character. The notes are a melange of bibliographical and textual, and many of the bibliographical notes supply not merely leads for further research but the sources and evidence for the cruces in Miller's argument. The careful reader, specialist or not, wants to see those notes as he reads.
I have gone on at such length on this subject - at the risk of sounding like one of those cranky reviewers who complain about the size of typeface or the quality of the binding or the color of the dust jacket, or who count typographical errors - because the endnote is a rapidly growing plague of scholarly book publishing rather than just an idiosyncratic failing of Professor Miller and his publisher. He has made it even worse by combining it with the scientific method of cita- tion (ironically, since, as we shall see, Miller is no friend of the appli- cation of scientific method to the study of human society) and by the number, length, and heterogeneous character of the notes.
Miller might reply that his notes and bibliography contain much useful information - which is true - and what was he to do with it? As footnotes with full bibliographical information, the notes would have taken up on average almost twenty percent of each page of the book. But it's a lot easier for the reader to glance to the bottom of the page and see whether a note is something he wants to read than to turn to the back ofthe book. And anyway the notes could and should have been pruned. Some of them are unnecessary; for example, the reader need hardly be told that the name of a saga character that Miller translates as "Sam" is not derived from the Hebrew "Samuel" (p. 353 n.27), especially since Samuel is not "Samuel" in Hebrew, it's Shmuel. No one thinks the Icelanders are the lost tribe of Israel. Some notes continue the discussion in text and could have been inte- grated into it. Some deal with recurrent issues relating to the texts of the Icelandic sagas and could have been consolidated and addressed in the introduction.
It is easy when writing a scholarly piece to keep adding notes as new thoughts occur to one or as new sources come to one's attention. But at the end one should go back and interrogate every note care- fully, asking: Is this a qualification or continuation or amplification of the text, and if so can it be worked into the text so that the flow of discussion is not broken? Is it a minor point that can be dropped with- out impairing the integrity of the work? Can it be consolidated with another note? Had Professor Miller undertaken such a regimen, and,
[Vol. 90:1495

 May 1992] Saga Iceland 1501
having drastically reduced the number and length of his notes, placed the survivors at the bottom of the pages where they belong, he would have produced an even more readable work than he has.
III
A more serious problem with Bloodtakingand Peacemakingis the book's relation to theory. The book is at once too theoretical and not theoretical enough. It is too theoretical in that it gestures toward two largely useless (for Miller's proper purposes) bodies of theory that happen to overlap. The result is some bad patches of jargon, some laboring of the obvious, and some superfluous bows to the norm of political correctness for which his university is notorious.7 One of the bodies of theory is the historical and, particularly, the anthropological literature on dispute resolution, and it is indeed a mine of interesting information about societies that resemble medieval Iceland in the weakness of their state institutions and resulting emphasis on substi- tute institutions such as the feud and arbitration; but the literature in question is for the most part a- or pretheoretical, occasionally ventur- ing interesting generalizations but usually emphasizing the uniqueness of the particular society under consideration. (Anthropologists, like historians, have a vested interest in the particularity of the individual culture that they have taken such pains to understand.) General knowledge is not the only knowledge, even if you don't agree with William Blake's dictum that to generalize is to be an idiot. But knowl- edge without an organizing theory is difficult to deploy, and Miller seems stumped by what to do with the anthropological and historical parallels that he has found.
The second body of theory that Miller invokes, albeit sparingly, is the discourse of "postmodernism" in its aspect of critique of the En- lightenment - and hence of functionalism, economics, and other sci- entific or at least systematic approaches to the understanding of society. Postmodernism derides objectivity, observer independence, the mutual intelligibility or intertranslatability of different cultures, the transparency of thought to language, the idea of progress, and other alleged presuppositions ofWestern thought. The world is a text, to be interpreted rather than deciphered, "interpretation" being un- derstood as (at best) the forcible imposition of meaning by individuals trapped in their subjectivity, their ethnocentrism. Abstraction kills. Our only hope of understanding is through the making of ever "thicker" descriptions. Postmodernism maintains an unbroken vigil for signs of racism, sexism, and ethnocentrism, particularly when the
7. See, eg., Doe v. University ofMich., 721 F. Supp. 852, 863 (E.D. Mich. 1989) (invalidat- ing, because too broad and too vague, the University of Michigan's regulation of hate speech and other forms of discriminatory harassment).

 1502 Michigan Law Review [Vol. 90:1495
oppressor race is white, the oppressor sex male, the oppressor ethnos Western.
I am being pretty derisive myself, and in my soberer moods I ac- knowledge that we have something to learn from postmodernism (and of course from history and anthropology - two fields, by the way, deeply invaded by postmodernism), that there is, in short, knowledge besides scientific knowledge. But it isn't easy to maintain this judi- cious approach in the face of Miller's employment of postmodernism. All he seems to have gotten out of it is an unpleasant and unhelpful, though happily very intermittent, jargon; a desire to dress common sense in the language of theory, in other words to mystify common sense; a conviction that the world is a very complicated and mysteri- ous place resistant to comparative study aimed at uncovering regulari- ties and constants analogous to those found by science in nature; and a fear of being thought judgmental.
Such phrases or propositions in Miller's book as "the negotiability of significations" (p. 3), "[t]he philology of residence" (p. 115), "[flostering was the social construct within which the circulation of children was comprehended" (p. 123), kinship was "an organizing metaphor" (p. 154), "[t]he balanced-exchange model, to give it a name, served as a kind of constitutive metaphor" (p. 184), "[p]ower is a difficult concept, at least since Foucault made it one" (p. 245), "power did indeed have a strong discursive component" (p. 246), and "[c]an the 'private' as an analytic category exist unless it is paired with and distinguished from 'public?'" (p. 305) show how the post- modernist style can lead an author away from clear-headed analysis of social institutions and into a terminological miasma.
In the last quotation Professor Miller is trying to save his Ice- landers from the clutches ofthe libertarians, who admire the privatiza- tion of law enforcement and find it exemplified in the society depicted in the sagas.8 However, the idea that "private" is not a meaningful term in a society without a "public" sector is, unless amplified, non- sense parading as paradox. The libertarian point is after all a simple though potentially misleading one: that the history of Iceland shows that a society can maintain some minimum of law and order without having what we would consider the minimum condition for law and order - a governmental monopoly of force. Medieval Iceland had barely any government, and what government it did have had no mo- nopoly of force. The point is not touched by Miller's rhetorical ques- tion. Maybe what he means, however, is that if there is no coercive government in a society, there will be other coercive institutions to take its place, and these - a numbing traditionalism, perhaps, or in- timidation by grandees - may curtail liberty just as much as does a
8. Actually, one libertarian. See David Friedman, PrivateCreationandEnforcement ofLaw: A HistoricalCase, 8 J. LEGAL STUD. 399 (1979).

 May 1992] Saga Iceland 1503
modem social-welfare state, the libertarian's bete noire. That would be a valid and important point. If it's Miller's point, I wish he'd say SO.
An earlier entry in my litany of quotations illustrates, once its con- text is restored, Miller's tendency to wrap a simple and correct, some- times even an obvious, point in postmodemist cotton: "Because power was so intimately linked with reputation and specifically with the rep- utation for having power (that is, power did indeed have a strong dis- cursive component), its loss, was often gradual, requiring both a slow cumulation of discomfitures and a consequent community reassess- ment of one's standing relative to others" (p. 246). This is an obscure and plethoric way of saying that political power, being latent force, will often persist after the actual willingness or ability to apply force has ceased, because it will take time for knowledge of that cessation to filter out. The United States could continue to exert power over other countries for quite a time after its officials secretly agreed never again to employ force in aid of foreign policy.
Readers of this review who are familiar with my writings in the economics of law may think my criticism of Miller's jargon is a case of the pot calling the kettle black. I admit that economics has a jargon, that the jargon is overused (I hope not by me), and that it is exces- sively mathematized (surely not by me, for I lack the capacity to in- dulge in that particular vice). There is a difference. The distinctive language of economics consists of defined terms, such as quasi-rents and marginalcost and marginalutility and Paretosuperiorityand com- parativeadvantageandconsumersurplusandelasticityofdemandwith respect to income, and the definitions are precise. The distinctive lan- guage of postmodernism does not consist of defined terms, is evocative rather than precise, imparts no economy of expression, and if on occa- sion it has a certain quasi-poetic charm - the sort of thing that initi- ates find in Heidegger - Miller's book is not one of those occasions.
Some of Miller's points may not be worth making at all, as distinct from being made clearly. For example, he tells us that "[1]aw never eschews violence either in early Iceland or in modem industrial soci- ety" (p. 232). That is true, because it is always possible that force may have to be used to compel obedience to a legal judgment, and physical force exerted against a resisting organism is violence. It does not fol- low, however, as Miller appears to believe, that in a comparison of early Iceland to modem us, there really isn't that much to choose be- tween on the score of law and order. In his conclusion, worrying the question whether Iceland was a more or a less violent society than ours is, Miller returns to the role of force in a modem legal system and asks whether saga Iceland merely seems violent in comparison to our society because of
the fear and anxiety we imagine we would feel at the prospect of having no state to enforce our rights for us or to protect us from those bent on

 1504 Michigan Law Review
[Vol. 90:1495
enforcing their own? In other words, does their culture seem more vio- lent because the responsibility for actually doing acts of violence was more evenly distributed than it is now, there being no state agents to delegate the dirty work to or to claim a monopoly on the dirty work? [p. 304]
These questions suggest that a bloodfeud society merely makes trans- parent the conditions of our own society. In the same vein Miller says' elsewhere that "there was no state apparatus to disguise better the relationship between the power of the parties and control of legal pro- cess" (p. 256), that "[t]he sagas do not show people continuallyliving with the anticipation of violence, rape, or expropriation that many American urban dwellers must live with daily" (p. 304), and that "[e]arly state formation, I would guess, surely tended to involve redis- tributions, not from rich to poor, however, but from poor to rich, from weak to strong" (p. 306). The qualifications ("early," "I would guess") should be noted, but the flavor of his discussion is that the modem state, including the modem states of the West - including the United States - may not be much of an improvement on an Iceland- style bloodfeud culture (with slavery) other than in the disguise and the mystification of power the more effectively to exploit the powerless.
Miller's own account of Icelandic society belies this attack upon the idea of progress. A lot of bystanders (granted, not all are inno- cent) get killed in these feuds; a lot of pawns get sacrificed. "Gunnar and Njal are substantial householders; they are neighbors and good friends. Their wives, however, have been bitter enemies ever since a quarrel over seating arrangements at a feast. Insults were exchanged there and over the course of the next few years so were killings of slaves and servants" (p. 284). When Gunnar and Njal are off together, Gunnar's wife takes the opportunity to send a member of her house- hold to kill a servant of Njal's household. Miller then describes how, to preserve their friendship, Njal accepted a cash settlement from Gunnar. "The darker side of Gunnar and Njal's settlement is that it was little more than a symbol of peace, a formalized reaffirmation of their friendship. Their wives, Bergthora and Hailgerd, the real dispu- tants, continued to fight; within a year Brynjolf [the killer of Njal's servant] was dead at Bergthora's [Njal's wife's] bidding" (p. 289).
At the risk of being labeled ethnocentric - a risk to anyone who suggests that the West may have progressed in the last one thousand years other than in techniques of oppression - I suggest that we have come a long way in domestic dispute resolution since Njal's day. If Miller wanted to refrain from making any normative judgments, that would be fine by me. But, particularly in his concluding chapter, he seems to be doing something else - burnishing his postmodernist cre- dentials, establishing his political inoffensiveness, by denying that any- thing in his book should be thought to give aid and comfort to

 May 1992] SagaIceland 1505
libertarians, or to whiggish liberals, and even trying to redeem medie- val Iceland for feminism by pointing out that the status of women was somewhat higher there than elsewhere in medieval Europe (p. 305).
I said Miller was "gesturing" in the direction of the historical and anthropological literature on dispute resolution and the postmodemist critique of Western thought, by which I meant to suggest that these bodies of theory play only an incidental role in his book. He plainly doesn't believe in the unintelligibility of other cultures. But the ges- tures have two bad effects. One is to complicate what is after all the simple and straightforward account that Miller wants to give us. The other is to deflect him from other scholarly approaches that might have helped him to tell an even more interesting story.
IV
The heart of the book is an account of how people keep greed and violence within bounds despite the absence of a state security appara- tus. The essential restraining device is the threat of retaliatory vio- lence by the potential victim,9 but to be a fully credible deterrent it requires that the victim have allies of some sort - otherwise there would be no (or yery little) deterrent to murder. Hence the impor- tance not only of kinship, but, as Miller also emphasizes, of rhetoric. Even kin may be reluctant to risk their necks on your behalf, so you have to be able to persuade (not prove: the factfinding procedures were not that developed). That is, you must be able to put the offender in the wrong in their eyes and demonstrate the importance of retalia- tion to the future security of the kin group. Kinship has a double (actually triple, as we shall see) significance in a system of revenge; it expands not only the ranks of the potential violators but also the range of potential targets. If X kills Y, Y's family may decide to go after X's brother rather than after X himself - maybe the brother is not as well protected as X. In other words, responsibility is collective ("joint and several," as the lawyers say); this in turn gives people an incentive to police their kin, lest a kinsman's misbehavior lead to retaliation against them rather than just against him. Miller also stresses, as have many other writers on feuds, the importance of a sense of honor. Shame, the reaction to being dishonored, helps overcome fear, making it more likely that a victim will retaliate if attacked or abused - and if he doesn't retaliate he will be easy game and the system of revenge will fail to keep the peace. Hence the importance of keeping score and the connection of the sense of honor with notions of exchange, balance, and reciprocity.
The difficulty with an economy of threats is that it is likely to be highly unstable, although there are counterexamples such as the nu-
9. "The continuance of good relations was assisted greatly by the fear of feud." P. 187.

 MichiganLawReview [Vol. 90:1495
clear balance of terror. Deterrence rarely works perfectly, so there will be some misconduct and therefore some retaliation, and each act of retaliation may be perceived by the original aggressor, or by his kin, as an act of aggression inciting retaliation against the retaliator. The feud is born, and it can spiral out of control. The incentive to form a group powerful enough to deter retaliation is therefore strong, and the competition to form such groups may result in a monopoly of force after all, and hence in the formation of a state. That this did not hap- pen in Iceland for three centuries suggests that the Icelanders must have had norms or institutions that played the same role in the blood- feud system as graphite rods play in the core of a nuclear reactor: to slow down the chain reaction.
One was law. Legal judgments were not self-executing, and if the convicted defendant thumbed his nose at a judgment the plaintiff
would have to rally his kin to enforce the judgment by force, much as if he had decided to retaliate directly against the defendant for whatever wrong had touched off their dispute. But, as Miller explains, a legal judgment might have enough suasive force to make it easier for the plaintiff to rally his allies and also easier for the defendant's poten- tial allies to beg off, thus tending to isolate the defendant and so vindi- cating the plaintiff's decision to go to law rather than to fight (pp. 238- 39). Miller also notes that the bilateral character of Icelandic kinship (Icelanders reckoned kinship through both the father and the mother, whereas in many societies kinship is figured only through the father and in a few only through the mother) made it more likely that a disputant would have kin on both sides of the dispute, and these kin caught in the middle were naturals to try to make peace between the disputants (p. 265). Arbitration was another method of mitigating the ferocity of feuds, although its utility, like that of law, was limited by the fact that a man who seemed too willing to litigate or compromise his disputes might get a reputation for being afraid to fight - might, thus, lose honor, and by doing so invite future aggression against himself.
Miller describes well this basic logic of a stable revenge system (stable in the sense that it doesn't either explode into anarchy or col- lapse into tyranny) and offers many illuminating illustrations ofits op- eration, emphasizing the subtle rhetorical and strategic skills required to maneuver effectively in such a society as well as the norms and institutions that secured its stability. So far, so good. But I think he could have further enlarged our understanding of the Icelandic sys- tem, and of the problem of social order generally, with a more eclectic research strategy.
The legal system that Miller describes resembles that of classical Athens. There prosecutions even for such serious crimes as murder and treason were initiated and conducted by private persons ("de-

 May 1992] Saga Iceland
1507
nouncers") and tried before panels of citizens chosen at random. There were no professional judges and no appeals. There were no law- yers as such, though orators such as Demosthenes hired themselves out to assist 'the litigants. Enforcement of judgments was by public officers, unlike the situation in Iceland; the panels of jurors were larger; and there were other differences between the two systems. But the parallels are striking - they include the heavy use of banishment as a sanction and a propensity to protracted and repetitive litigation by "feuding" factions - and there is an extensive literature on the opera- tion of the Athenian system, including the role of forensic oratory, that Miller could have consulted with profit.1°
The ancient Greek society that most resembles saga Iceland, how-
ever, is not Athens; it is the society depicted in the Homeric epics.
(And there is the same problem, as in the sagas, of disentangling fact
from fiction in works ofimaginative literature.) Homeric society1" has
only rudimentary governmental institutions, and exhibits the same
emphasis that we find in the sagas on revenge as the principle of social
order. Some years ago I tried to extract the basic system of social
order from the Homeric epics much as Miller does with the sagas.12
But my approach to revenge, the feud, and related institutions and
practices differs from Miller's in emphasizing the utility of economics
as a tool for understanding the operation of a revenge system.' 3 Eco-
nomics is also the approach used by David Friedman in his article on
Iceland.' 4 Recently - too late (I assume) for Miller to have seen the
work - an Icelandic economist has chimed in with an interesting ex-
planation for the stability of saga Iceland's system of governance.' 5
There is an economic literature on other "stateless" regimes that have
managed to maintain a modicum at least of social order, such as the mining communities that sprang up during the California Gold Rush
at a time and in a place where government authority was essentially
6 nonexistent.'
10. See, for example, 2 ROBERT J. BONNER & GERTRUDE SMITH, THE ADMINISTRATION OF JUSTICE FROM HOMER TO ARISTOTLE (1938), other references in my book, RICHARD A. POSNER, LAW AND LrrERATURE: A MISUNDERSTOOD RELATION 110 n.57 (1988), and THOMAS C. BRICKHOUSE & NICHOLAS D. SMITH, SOCRATES ON TRIAL (1989).
11. By which I mean the society Homer depicted, not the society in which Homer (or who- ever the author or authors of the Homeric epics were) lived.
12. Richard A. Posner, The Homeric Version of the Minimal State, 90 ETHICS 27 (1979), reprintedin RICHARD A. POSNER, THE ECONOMICS OF JUSTICE 119 (1981).
13. See also Richard A. Posner, Retributionand Related Concepts of Punishment,reprinted in POSNER, THE ECONOMICS OF JUSTICE, supranote 12, at 207 and POSNER, supra note 10, at 25 ("Revenge as Legal Prototype and Literary Genre"). Recent work in the same vein includes Alan P. Hamlin, RationalRevenge, 101 ETHICS 374 (1991).
14. See Friedman, supra note 8. Miller cites this article, but only to show that libertarians have taken an interest in Iceland; he does not discuss it.
15. THRINN EGGERTSSON, ECONOMIC BEHAVIOR AND INSTIUTIONS 305-10 (1990), dis- cussed infra text accompanying note 17.
16. See, eg., ROBERT C. ELLICKSON, ORDER WITHOUT LAW: How NEIGHBORS SETTLE

 Michigan Law Review [Vol. 90:1495
Economists are interested in incentives and hence in deterrence, and Miller himself recognizes the centrality of these factors to the so- cial system of saga Iceland. He puzzles over the relation between honor and feud, but quickly recognizes that the key is deterrence: "honor is the ability to make others believe that you will indeed be tough the next time" (p. 303). Economics provides a lens for seeing the forest rather than just the trees. Honor, feud, reputation, balance, exchange, reciprocity, legitimacy, power, rhetoric, and the other fea- tures of medieval Iceland that Miller describes are complex, elusive, and nuanced phenomena. But they can also be seen in a simpler, less variegated, less impressionistic light. Remove the coercive state (or fail to institute it in a new society, as happened in Iceland), and people still want to eat, and live, and reproduce, and this results in the culti- vation of attitudes, demeanors, and values that maximize survival (broadly defined) in the stateless setting by creating an equilibrium based on, but also limiting recourse to, threats of violence. All this is clear enough but perhaps thin; the challenge to the economist is to explain the differences in attitudes, demeanors, and so on, and result- ing differences in norms and institutions, across stateless societies.
The stabilizing and destabilizing features of a revenge system are particularly interesting because they explain why such a system lasts as long as it does last and why it collapses when it does. The system seems, as I have already suggested, inherently unstable because people have an incentive to form ever larger protective associations, whether based on kinship or on geographical propinquity, until at last one of them achieves a monopoly of force and the state is born. Economics helps us understand why this happened so slowly in Iceland. First, Iceland didn't have any significant foreign enemies - there were no threats of invasion until Norway began throwing its weight around toward the end. So the benefits of a large protective association were fewer than they otherwise would have been. Second, the country was very poor, which made it difficult for anyone to support an entourage of retainers, feudal-style, who in exchange for food and shelter would place an armed force at the disposal of their liege; lacking that armed force he had little to offer in the way of protection to other people in the society. This is to exaggerate some, since Iceland had chieftains, but the pitifully small forces at their disposal - a handful of relatives, dependents, and clients taking the afternoon off from farm work - limited their power to the execution of piecemeal revenge. None was able to offer a king's peace to the entire society.
In other words, a community needs an economic surplus to sup- port specialists in coercion. Once the surplus appears, however, the stateless state's days are numbered. Eggertsson suggests that toward
DIsPUTEs (1991) (reviewed in this issue by Professor David Friedman- Ed.); GARY D. LIBE- CAP, CONTRACTING FOR PROPERTY RIGHTS (1989).

 May 1992] Saga Iceland
1509
the end of Icelandic independence the Church was able at last to col- lect a heavy tax yet unable to keep the bulk of the revenues out of the coffers of a handful of major chieftains. Six of them grew to the point where they could and did engage in civil war on a respectable scale, at which point the population was happy to turn to the King of Norway
for protection and Icelandic independence ended.17
I mentioned Miller's penultimate endnote, in which he refers to
game theory, the science of conflict and conflict resolution. He cites a single work by Robert Axelrod' and explains that Axelrod had found in a series of computer games that the optimal strategy for achieving cooperation in a decentralized system is "tit for tat"; that is, if you act noncooperatively toward me, I'll retaliate in kind, but I won't escalate the retaliation. Miller gives Axelrod a polite brush off 19 but con- cludes, "I am not a student of game theory, so the points I am making here may well have easy answers I do not know about" (p. 374). This won't do. Game theory is, as I said, the science of conflict and conflict resolution, so if you're going to write about that subject it behooves you to study the relevant science at least to the extent necessary to satisfy yourself that there is nothing in it that you can use.20 Game theory can be forbiddingly mathematical, but it doesn't have to be, any more than economics has to be, and William Miller could learn the rudiments of game theory2' in a good deal less time than it took him to learn Old Icelandic.
Equipped with these complementary bodies of knowledge that I have noted, Miller will be poised to make even more important contri- butions to our understanding of law, Iceland, and social order in his future articles and books than he has made in the excellent book under review. And it is indeed excellent, though I have emphasized my areas of disagreement. And while making qualifications I should like to ac- knowledge that social science is not the only route to the understand- ing of human social behavior. Literature is another route, and the sagas are literature. But with all due respect for Professor Miller, he is not an author of literature, or even a translator of it; and though trained as a literary critic, he does not write as a literary critic. He writes as a social and legal historian, and in this role he could, I re-
spectfully suggest, benefit from widening his scholarly horizons.
17. E GERTSSON, supranote 15, at 309-10.
18. ROBERT AXELROD, THE EVOLUTION OF COOPERATION (1984).
19. "Axelrod's analysis of computerized strategies for the Prisoner's Dilemma game is re-
markably suggestive, but does not seem to be readily applicable to the Icelandic feud without considerable qualification." P. 374.
20. I say this who shouldn't, because I have written about conflict, too, without equipping myself with the rudiments of game theory.
21. See generally ERIC RASMUSEN, GAMES AND INFORMATION: AN INTRODUCTION TO GAME THEORY (1989). For pertinent applications to law, see ELLICKSON, supra note 16, at 296 (index entry for "game theory").

 1510 Michigan Law Review [Vol. 90:1495 V
I end with a few comments on the genre of scholarship, and of scholar, to which this book and its author belong. In the brief period since Miller's book appeared, there have also appeared Professor El- lickson's study of extralegal dispute settlement,22 Professor Grey's study of the poetry of Wallace Stevens,2 3 Professor Rosenberg's study of the effects of landmark legal cases, 24 and Professor Cohen's study of the legal and moral regulation of sexuality in ancient Athens. 25 Ellick- son and Grey are law professors, Rosenberg is a political scientist, and Cohen, though law-trained, is the chairman of a department of rheto- ric. These are all excellent books published by distinguished univer- sity presses despite what might appear to the pedantic to be, in each case, a mismatch between the author's professional credentials and his subject matter. Miller's book exhibits the same phenomenon, for he is a law professor and an ex-English professor writing Icelandic social history (law is by no means the principal focus of the book).
What is going on? This is the age of specialization. Were we not told by Max Weber almost a century ago that "[l]imitation to special- ized work, with a renunciation of the Faustian universality of man
which it involves, is a condition of any valuable work in the modem
26 world"?
I think a number of things are going on:
1. The academic-legal market for books other than textbooks is
small, in part because law professors, being habituated to excellent li- brary facilities, are not natural book buyers. As a result university presses, which generally do not publish textbooks, are reluctant to publish law books that do not spill over into other fields. So spillover
books are overrepresented in these presses' law lists.
2. Law teaching, because it is by academic standards well paid,
undemanding, and noncompetitive, attracts a number of bright people whose true love is for a different subject. A fraction of these turn out to be highly motivated scholars despite their having an easy berth.
3. Being subject rather than method, law is amenable to study by people in other disciplines, such as economics or political science or even literary criticism, or by lawyers employing the tools of those disciplines.
22. See ELLICKSON, supranote 16.
23. THOMAS C. GREY, THE WALLACE STEVENS CASE: LAW AND THE PRACTICE OF PO- ETRY (1991).
24. GERALD N. ROSENBERG, THE HOLLOW HOPE: CAN COURTS BRING ABOUT SOCIAL CHANGE? (1991) (reviewed in this issue by Professor Stephen L. Carter - Ed.).
25. DAVID COHEN, LAW, SEXUALITY AND SOCIETY: THE ENFORCEMENT OF MORALS IN CLASSICAL ATHENS (1991).
26. MAX WEBER, THE PROTESTANT ETHIC AND THE SPIRIT OF CAPITALISM 180 (Talcott Parsons trans., 1958).

 May 1992] SagaIceland 1511
4. Specialization begets a demand for generalists and interdis- ciplinarians. The proliferation of arcane subspecialties creates a space for bridge builders and translators who will meld approaches or turn them to problems outside the research paths of the specialist practi- tioners. Legal training is not the worst preparation for playing this role.
5. Most important I think, the rapid progress of the social sciences in recent decades, and the increasing emphasis on theory in fields such as literature and history that had for so long resisted it, have multi- plied the opportunities for bringing other fields besides law itself to bear upon law. The result is a growing and improving interdiscipli- nary legal scholarship honorably represented by Bloodtaking and Peacemaking.
}}