The total number of gypsies (aka Rom, Romani) is variously estimated as from three to fifteen million. If current scholarship is correct, they are descendants of a population that left northern India more than a thousand years ago. They first appear in European history in the 15th century, showing up at the court of the Holy Roman Emperor Sigismund, where they claimed to be from lesser Egypt in Greece, on pilgrimage as penance for the temporary abandonment of Christianity by their ancestors. Multiple accounts describe them as travelling through Europe bearing letters of safe conduct from Sigismund, letters giving them judicial autonomy, the right to be punished only by their own authorities.
The letters may have been forgeries created by their bearers to protect them from local law enforcement authorities, but they need not have been. Polylegal systems in which different people in the same country were under different laws and legal authorities were common in medieval and renaissance Europe as we will see in later chapters. It is possible that 15th century Gypsies persuaded Sigismund that they were entitled to similar treatment.
Whether or not 15th c. Gypsies obtained a grant of de jure judicial autonomy from a 15th c. emperor, gypsy communities through the centuries have been strikingly successful in maintaining their de facto autonomy, remaining below the radar of the official legal system while imposing their own rules on their own members. It is that fact that makes them of interest for the purposes of this book. Presently existing Gypsy communities provide examples of not one legal system very different from ours but several different systems, legal systems differing primarily not in the rules they enforce but in the mechanisms they use to enforce them.
The rules, in large part common to the different communities, can be usefully grouped into two categories. One consists of ordinary legal rules covering the obligations of gypsies to each other. Swindling or stealing from a fellow gypsy is an offense to be dealt with, swindling or stealing from a non-gypsy comes under gypsy law only to the extent that it creates problems for other gysies.
The second category covered by gypsy law is an elaborate system of purity and pollution, rather like orthodox Judaism on steroids. Its central tenet is that the human body is unclean from the waist down, clean from the waist up. Contact with the unclean is polluting—ŅmarimˇÓ—and the pollution is contagious. Someone who is polluted will find others reluctant to associate with him or even permit him to touch their possessions, providing an automatic enforcement mechanism for the rules against pollution and an incentive to go through the rituals required to cleanse pollution.
Excretion and reproduction, being associated with the lower half of the body, are the subject of extensive rules and restrictions. Among the Vlach Rom, the largest of the Romani cultures, a pregnant woman is expected to eat alone, consume food cooked in her own pots, and after childbirth destroy the garments she wore while pregnant. A woman can pollute a man by skirt tossing—exposing her genitals—obliging the victim engage in costly procedures of purification. The Kaale, the Finnish gypsies, a small population isolated for centuries, carry this attitude even further, refusing to openly admit the facts of human reproduction. They have no institution of marriage; couples that wish to reproduce are expected to first leave their family households, flee a substantial distance away—far enough so that their kin cannot find them and retrieve the woman—and return only when the child is weaned and so no longer requires a visible association with its mother. On returning, the father is expected to show the humility appropriate to one who has violated the norms of his society while the women of the motherÕs generation smuggle mother and child back into the household, where the child will be expected to treat all the women of his motherÕs generation as equally mothers.
One result of the Kaale rejection of sexuality is to eliminate many of the taboos associated with it among other Gypsy groups. There can be no restrictions associated with menstruation since enforcing them would require recognition of the fact of menstruation, and similarly with pregnancy. A Kaale woman living in the household of her (or her partnerÕs) kin conceals the fact of pregnancy until shortly before delivery, and arranges for it to happen somewhere outside of the household—in modern times in a maternity hospital.
The marimˇ rules, while differing in detail across Gypsy communities, are similar in their general form. What differs much more is the system for enforcing both those rules and other parts of gypsy law. Among the Vlach Rom, the descendants of the gypsies enserfed for four centuries in Romania, disputes that cannot be settled at a lower level are submitted to a kris, a court proceeding in which all adult male members of the community (in some cases female as well) are permitted to take part. Details of how the kris functions vary across accounts and probably across communities, in particular to what degree the ŅjudgesÓ produce a verdict and to what degree they simply function as chairmen presiding over an open discussion. Whatever the details, the common pattern is that the dispute is argued until consensus is reached; that consensus is then imposed on the parties to the dispute, if necessary enforced by the threat of ostracism via a sentence of marimˇ.
Ostracism provides one example of how an embedded legal system like that of the gypsies, one that exists within a territory under the rule of a state with much greater resources of coercion than the community possesses, can function. Refusing to associate with someone is not illegal, so the verdict of the kris can be enforced without violating state law. Other possible verdicts, up to and including capital punishment, have in the past been imposed, but become less workable the greater the ability of the state legal system to enforce its own rules on the embedded commuity and the greater its interest in doing so. Thus an embedded legal system that operates without a state sanction is more restricted in its means of enforcement than one that possesses that sanction, as Jewish communal authorities in the medieval period frequently did and as 15th c. gysies at least claimed.
In contrast to the Vlach Rom, the Romanichal, the largest of the British gypsy communities, enforce their rules not by communal pressure and the threat of ostracism but by a feud system, a less developed version of the same legal system as saga period Iceland. A Romanichal gypsy who believes his rights to have been violated responds by demanding, with threats of violence, compensation. Both parties know that if rights as defined by the norms of that community have been violated, the violatorÕs friends will be reluctant to support him, the victimÕs friends willing to support him. That makes it in the guilty violatorÕs interest to offer compensation or, if unwilling or unable to do so, to remove himself from the neighborhood of the victim, just as an Icelander outlawed for failing to pay the fine imposed on him by a court would leave Iceland. As with any well functioning feud system, while the incentive to obey the laws or norms is provided by the threat of private violence, actual violence is the exception rather than the rule.
The differences between the Vlach Rom and the Romanichal, each of which is inclined to view its institutions as the original and the otherÕs as a corrupted version, are not limited to different enforcement mechnisms. The Vlach Rom organize themselves into extended families, and practice marriage by purchase; the Romanichal function as independent individuals, and practice marriage by elopement.
The Kaale provide an example of a third approach to enforcement. The relevant unit is the household, not, as among the Romanichal, the individual. All households are considered peers, and there exists no mechanism above the household for peacefully settling disputes. Conflict within the household is settled internally, in a society in which authority is centered in male elders. Violation of marimˇ rules in ways that affect other kin groups leads to a loss of status and honor by the group whose member is responsible, providing an incentive to prevent such violations by enforcement within the household.
Conflict between individuals of different households, if sufficiently serious, leads to duels between the parties, under rules designed to reduce the risk of death or serious injury. If death or serious injury does occur, in the context of a duel or otherwise, the result is a blood feud between the kin groups of the parties. In principle, the feud can be pursued by any member of the group one of whose members was killed or injured against any member of the other group, although in practice women or children are unlikely to be targeted, the responsible individual in the other group particularly likely to be. Successful retaliation does not end the feud and there is no equivalent of the court procedures or arbitrated settlements that terminated Icelandic feuds; successful retaliation merely exchanges the position of the two kin groups.
While in theory these rules could lead to a succession of killings, in practice feud mostly takes the form not of violence but of avoidance—of all members of each kin group by all members of the other.
Gypsy communities share a common ancestry, a common language, a common set of rules of cleanliness and pollution, but have several quite different approaches to enforcing the rules of the community. That raises an obvious and interesting puzzle. Why do the Vlach Rom have, and the Kaale and Romanichal not have, the institution of the Kris? Why do the Vlach Rom not have and the Kaale and Romanichal have, in somewhat different forms, the institution of feuding?
The authors of Chapter 3 of Gypsy Law, whose description of the Romanichal system is reflected in my description above, offer an interesting, if conjectural, solution. They argue that the feud system depends in part on the ability, when all else fails, of one party to walk out on the other party and the dispute; absent that, the risk of serious violence becomes too large. Gypsies have, for most of their history, been a migratory population, but there is at least one large exception, the period of four centuries during which the Vlach Rom were enserfed in Romania. Serfdom meant the loss of mobility. Hence they argue that the feud system represents the original institution, the Kris and associated institutions an adaptation to the special circumstances of serfdom, arising among the Vlach Rom and adapted from institutions of the Romanian peasantry.
I have no analogous explanation for the difference between feud as practiced by the Romanichal and feud as practiced by Kaale or, more generally, the individualism of the Romanichal and the household based society of the Kaale. But Marti Gronforss, on whose account of the Kaale mine is based, does offer an interesting suggestion linking Kaale blood feud to Kaale non-mariage.
Mating among the Kaale is exogamous, pairing within the household strictly forbidden. The recognition of a serious and stable relationship created by sexual partnership and reproduction would create kinship links between households, subverting institutions based on autonomous and coequal households. Under institutions of non-marriage, one partner resides in the household of the otherÕs kin as a sort of resident alien. Hence blood feud involving the household of one partner does not concern the other or his kin as participants or potential targets. If a blood feud arises between the households of partners the result is the immediate and complete severing of the connection between them, with the one who was dwelling in the household of the otherÕs kin returning to his or her own kinÕs household.
Embedded Law: Problems and Solutions
There are at least three different ways in which a community can enforce its own legal system despite the superior access to force of the legal system of the state within whose territory it is located. One is by delegation; the state legal system subcontracts the job of enforcing legal rules on the members of the community to the communal authority, and delegates some of its authority for the purpose. Gypsy communities in the past have claimed and may perhaps have possessed such delegated authority, but contemporary communities, with very minor exceptions, do not.
A second approach is for the community to restrict the mechanisms by which it enforces its legal rules to acts that do not conflict with the rules of the state legal system. An example is the use of the threat of ostracism by the Vlach Rom to enforce the verdict of a kris.
The effectiveness of that threat depends on how easily the exiled gypsy can function outside of his community. The marimˇ rules (and similar rules in other societies) provide a mechanism for isolating the members of the community. Gaije, non-gypsies, do not know the marimˇ rules and so do not and cannot obey them. It follows that they are all polluted, unclean, carriers of a contagious disease, people whom no Rom in his right mind would willingly choose to associate with; when and if such association is unavoidable it must be taken with great care. The gypsy view of gaije, reinforced by the gaije view of gypsies as uneducated and illiterate thieves and swindlers, eliminates the exit option and so empowers the kris to enforce gypsy law by the threat of exclusion from the only tolerable human society.
A third approach to enforcing an embedded legal system, also employed by gypsy communities, is to use control over information to substitute for control over physical force. I started this chapter by reporting a range of estimates for the world population of gypsies. That the estimates range over almost an order of magnitude is not an accident. Gypsies do not wish to be controled by gaije. It is hard to control people if you cannot count them, and it is hard to count people when there is no one to one correspondence between person and name—Gypsies treat a name, more generally an identity, as fungible, property belonging to the extended family to be used by any member who finds it useful. By this tactic and others, modern gypsies make it difficult for the states that claim authority over them to monitor and control them, and so increase the range of alternatives available to gypsies and gypsy law.
 Gypsy Law, p. 28. This book is the basis of almost all of the information contained in this chapter.
 This is presumably the origin of ŅEgyptianÓ and hence ŅGypsyÓ as labels for the Rom.
 This account is based on research done by Marti Gronforss, mostly from 1976-8, and reported in Chapter 7 of Gypsy Law.
 This account is based on Chapter 3 of Gypsy Law, written by Thomas Acton, Susan Caffrey, and Gary Mundy.
 Gypsy Law pp. 228-30 gives examples of North American courts making some attempt to take account of Gypsy law.
 Reference from the book to restricting gaije to front room, covering furniture, etc.
 One implication of this is that a more tolerant attitude by non-gypsies towards gypsies may destabilize the gypsy system by making exit easier, as may a more tolerant attitude in the other direction.
 For a more detailed description of the ways in which Gypsy communities restrict information about them and control over them by the states within which they live, see Gypsy Law pp. 52-3.