The Arbiter's Principal: Joint Appointment, the Challenge before the Emperor and the Covert Imperial Party in the Law of Kanly, 9719–10191 AG

1. Introduction

Early in 10118 AG, House Essenmark came before the Landsraad High Council at Kaitain, the Padishah Emperor present, to challenge a finding entered against it by the Judge of the Change in its kanly with House Varnholt. The Judge had found that agents of Essenmark poisoned the cisterns of a Varnholt market town, killing 340 people who belonged to neither House, and had ordered the cession of two Essenmark estates as the penalty. Essenmark had held a directorship in CHOAM at the Emperor's instance since 10086 AG, seven years before the feud was declared. The Council vacated the finding and remitted the cession. The feud ran sixty-three years more and closed in 10181 AG with the loss of the Varnholt fief.

The case is one of eight in the registry period in which the High Council reversed a Judge, and it belongs to the pattern that Oudevar et al. (10247 AG) measured across 203 linked feuds. A Judge of the Change seated early in a kanly between Houses outside Imperial favour cut the odds of bystander death to about a quarter; in feuds where either House stood in favour, an early Judge made no difference. That study showed the division and left its cause open. The registry from which both studies start records how kanly ended and what it cost (Roquefeuil et al., 10242 AG); it does not record why an arbiter's rulings held in one class of feud and not in another.

This paper, completed in 10253 AG, takes up the cause as a question of law. We examine the rules that create a Judge's authority and the rules by which that authority is reviewed, and test what we find against the archive series in which the rules were applied.

2. Sources of Law

The law of kanly is thin in text and thick in practice. The articles of the Great Convention on kanly and the War of Assassins, in the consolidated edition with the standing glosses, give the permitted weapons and the requirement that a feud be declared and conducted in its forms; like every law of the Convention, they open by requiring that the forms be observed (Imperial Chancery, 10102 AG). Two further rules govern the Judge of the Change. A Judge is appointed by the Landsraad High Council and the Emperor, to oversee a change of fief, a kanly negotiation or formal battle in a War of Assassins. A Judge's arbitration may be challenged only before the High Council with the Emperor present. Both rules appear in the consolidated edition (Imperial Chancery, 10102 AG), and we found no amendment to either in the series we examined.

Application is recorded in two series at Kaitain. The commission files of the Judges of the Change (Landsraad High Council and Imperial Chancery, 9719–10190 AG) give each Judge's name, offices and date of seating. The minutes of challenges (Landsraad High Council, 9719–10190 AG) record 47 challenges to a Judge's arbitration in registered feuds, each with the parties, the ruling challenged and the disposition. We use the definition of Imperial favour adopted by Oudevar et al. (10247 AG): a direct Imperial grant of fief or office, a CHOAM directorship held at the Emperor's instance, or a marriage into House Corrino in the twenty years before the declaration.

Commentary has treated the Judge as a neutral instrument of the forms (Vesk, 10232 AG), and the Emperor's place at a challenge as a guarantee of the forum's authority (Pellisandre, 10245 AG). Neither treatment asks what happens when the guarantor has an interest.

3. The Seat: Joint Appointment

The first rule gives the Emperor a share in choosing every Judge. The text does not divide the share, saying nothing of whether the Council nominates and the Emperor confirms or the reverse, and the commission files show both orders in use. What the files do show is whom the joint appointment produced. A Judge was seated in 162 of the registry's closed feuds, and the commission survives for 141 of them. In 51 (36.2%) the Judge held an Imperial office at the date of commission, as a Chancery official, an Imperial magistrate or a servant of the Imperial household. The figure is descriptive; the files give no comparison with how Judges would otherwise have been chosen. A Judge who holds Imperial office is not, by that fact, partial. Many of those 51 Judges sat in feuds in which neither House stood in favour, the class of feud in which Oudevar et al. (10247 AG) found an early Judge effective. But the rule makes no distinction between a feud in which the Emperor is indifferent and one in which a party enjoys the Emperor's favour. Nothing in the text bars a Judge in Imperial office from sitting in a feud where one House holds an Imperial grant, and the files record no recusal on that ground in nearly five centuries.

4. The Challenge before the Emperor

The second rule reaches further. A Judge's arbitration is final unless challenged, and a challenge lies only to the High Council with the Emperor present. The texts do not say whether presence is a vote, a seal on the Council's ruling or a witness to it (Pellisandre, 10245 AG). In practice the distinction made little difference. The minutes of the 47 challenges record no ruling entered against a view the Emperor had expressed in the hearing. Of the 47 challenges, the Council upheld 39 rulings and reversed 8. In 7 of the 8 reversals (87.5%), the House that gained stood in Imperial favour. Among the 39 upheld rulings, the House that a reversal would have helped stood in favour in 14 (35.9%). The difference is unlikely to be chance (exact test p = .015), though the reversals are few, and several Houses appear in more than one challenge, so the test is descriptive. The Essenmark challenge of 10118 AG resembles the other six, each a finding against a favoured House that was vacated while the feud continued.

Eight reversals in nearly five centuries are too few to account by their own weight for the division Oudevar et al. (10247 AG) report. One reading is that their force lay in what they taught. A favoured House facing a Judge knew that any adverse ruling went, on challenge, to a forum in which its patron sat, and a Judge knew it too. On that reading the rule need not be invoked often to set the terms on which every Judge in a favoured feud worked.

5. The Covert Party at Arrakeen, 10191 AG

The forms of kanly assume that the parties are named. A declaration names two Houses, and every later step, from the herald's renewal to a challenge before the Council, runs between them. Favour, as the registry period knew it, enters a feud from outside the declaration, through a party's patron. The Atreides assumption of Arrakis and the Harkonnen attack on Arrakeen in 10191 AG take this to its limit, and lie just outside the period both earlier studies counted. A Judge of the Change, Liet-Kynes, oversaw the change of fief; he was also the Imperial Planetologist and so the Emperor's own servant. When the Harkonnen attack came, Sardaukar fought in it under Harkonnen livery. The Emperor was then not a patron of one party but an unnamed combatant. The Atreides–Harkonnen enmity, older than the registry and conducted by both Houses as kanly, offered no form through which that participation could be brought before an arbiter, because every route of review ran to a forum at which the Emperor sat. We make no claim here about what the Judge did or might have done at Arrakeen. Our concern is the law, which gave him no procedure for it.

6. Comparative Assessment

The Great Convention's ban on atomics against persons held through the same centuries with no arbiter of comparable standing, and the inquiry files record no finding that a House used atomics against persons. On the reading of Vantanza et al. (10240 AG), its sanction lay in the family stockpiles of every Great House, so that a breach exposed the offender to all the others, and no ruling was needed to make that sanction available. Matters were referred for fact-finding, but the ban did not depend on a ruling to hold. Kanly's restraint is of a different kind. It is administered, not self-enforcing, since a feud is bounded only when a Judge can enter findings of forbidden means and make them hold. That design works where the arbiter's principal is disinterested, and the registry shows it working in most feuds. It fails where the principal is interested, because the law gives the principal a hand in appointing the arbiter and a seat at the arbiter's review. Formal combat between champions, which Imhoff (10236 AG) traces to before the Corrino settlement, was the quickest ending (Roquefeuil et al., 10242 AG). It still ran under a Judge's oversight, so it offered no escape from the arbiter's principal, only a shorter route past him.

7. Conclusion

The division found by Oudevar et al. (10247 AG) has a candidate doctrinal source. The Emperor shares in appointing every Judge of the Change and must be present at every challenge to one, and the law contains no rule for withdrawal when the Emperor's favour, or the Emperor, is interested. What this reading does not settle is which rule carries the effect. The commission files would allow the favoured feuds to be split by whether the Judge held Imperial office; if the effect of an early Judge returns where the Judge did not, the seat matters; if it does not, the challenge rule, or favour acting by a route outside both rules, is the likelier. That analysis needs no new archive, but a second question does.

The High Council's minutes of challenges for 10193–10219 AG, from the years in which House Atreides held the throne, are held in the Imperial Archives and have not been released. They might show whether the Council ever heard a challenge in which the Emperor's interest was direct, and whether anyone proposed that the Emperor withdraw. Until they are read, the question of whether the forms could adapt to an interested Emperor remains open.

The Essenmark challenge of 10118 AG, read this way, was not an aberration in the Council's practice. A favoured House carried a Judge's finding to the one forum where its patron sat, the finding fell, and a feud that had already killed 340 outsiders ran on for sixty-three years. Every step was within the forms. Whether the finding fell because the Emperor sat, or because the Council would have vacated it anyway, is what the minutes cannot yet tell us, and what the split of the commission files could begin to test.