Section 20
Other Selected Paradoxes
and Reflexivities in Law Peter Suber, Paradox of Self-Amendment Table of Contents

• A. Protagoras v. Euathlus
• B. State v. Jones
• C. Self-referring laws
• D. The liar
• E. Circular liens and liabilities
• F. The bootstrap doctrine
• G. Inferences drawn from the fact of the dispute
• H. Reflexivities of sovereignty
• I. Self-amendment
• J. Permissible disobedience
• K. Contract reflexivities
• L. "More of the same"
• M. Breaking vicious circles
• N. To know before we know
• O. Tax reflexivities
• P. Circular reasoning
• Notes

This section is a brief introduction to the abundance of reflexivities and paradoxes in law. The lessons learned under the paradox of self-amendment do not necessarily apply to these, which is one reason to include them now. But all raise the question whether a strictly logical legal system could tolerate them and, if not, whether our legal system ought to continue to tolerate them. And oughts aside, as long as our legal system does tolerate them, empirically-minded legal theorists must take this toleration into account when describing legal reasoning and the respective roles of policy and formal logic in legal decision-making. Reflexivities and paradoxes that present "insoluble" problems to logicians reveal much about legal reasoning when they are "solved" without more ado by courts or legislatures, or imperturbably left unsolved.

I do not have as precise a definition of "reflexivity" as I gave of "paradox" in Section 1. "Reflexivity" is the generic term for all types of logical circularity: the self-reference of signs, the self-application of principles, the self-justification and self-refutation of propositions and inferences, the self-creation and self-destruction of legal and logical entities, the self-limitation and self-augmentation of powers, circular reasoning, circular causation, vicious and benign circles, and feedback systems.[Note 1] The common practice of using "self-reference" to cover all this territory is inaccurate and unfortunate.

Some legal terms use the language of reflexivity to denote what are plainly irreflexive phenomena. For example, "self-authenticating documents" in the law of evidence, and "self-proving wills" in probate, are not documents that, by their own lights, supply all the premises needed to conclude that they are what they purport to be. They are documents that, if they meet some externally stipulated requirements, are presumed to be what they purport to be. "Self-defense" and "self-incrimination" in criminal law emphasize the personal self more than the circularity of the process, while "self-dealing" comes much closer to a reflexive application.

Reflexivities are not necessarily paradoxical. "This very sentence is false" is a classical reflexive paradox (see Section 1); it is true if false, and false if true. But "This very sentence is true" is equally reflexive but not at all paradoxical.

The following is simply a list of some reflexivities that arise in other domains of law. The heads under which they are collected are crude; many of the examples could fit under more than one of them. With a few exceptions, too important to omit, I have tried to avoid examples that have been noted in the literature before. I omit redundant examples that would not present new logical features. For the most part I have also omitted hypothetical statutes, contracts, etc., invented solely to create reflexivity problems.

In order to present a wide range of examples, I have had to limit my commentary and analysis. I am confident, however, that each of these examples, no matter how innocuous it may seem, could sustain the kind of in-depth study that the larger work has given to self-amendment.

A. Protagoras v. Euathlus

This case deserves separate, special treatment as a classical illustration of the reflexive use of the "counter-dilemma" to respond to a "dilemma".

Protagoras taught rhetoric and argumentation, which in ancient Greece comprised the education of a lawyer. Euathlus wished to learn these arts and asked Protagoras to teach him. He said he could not pay right away but promised to pay in full after he won his first case. Protagoras agreed, and taught Euathlus rhetoric and argumentation. Some accounts say Protagoras sued Euathlus for payment immediately upon the completion of the lessons; other accounts say he waited until it was evident that Euathlus was not taking on any cases. The case was heard in the court of Areopagus in Athens.

The judge asked Protagoras why he thought he had a claim against Euathlus. Protagoras argued, "I will either win this case or lose it. If I win it, then Euathlus must pay me, by the judgment of the court. If I lose it, then he must pay me, under our contract. So he must pay me either way."

The judge was impressed, and asked Euathlus to reply. Euathlus had learned his lessons well, and replied, "I too will either win this case or lose it. If I win it, then I need not pay Protagoras, by the judgment of the court. If I lose it, then I need not pay him, under our contract. So I need not pay him either way."[Note 2]

It is said that the court was so puzzled that it adjourned for 100 years. Aulus Gellius reports that Protagoras had a "wily scheme" in bringing the suit in the first place (probably to win in a second suit if he lost the first), and that Euathlus knew he'd escape paradox if he hired another advocate to argue his case but that he wanted to show off his cleverness. Euathlus would be better off hiring a lawyer because, if he won with a lawyer, his victory would be non-paradoxical, and if he lost with a lawyer, he would not yet have won his first case. By adjourning, the court in effect waits for Euathlus to take another case, which is equivalent to a judgment for Euathlus.

There is a small literature on the case, all of it suggesting that Euathlus should have won.[Note 3] Most commentators then observe that Protagoras could have sued a second time and won. A few observe that Euathlus might then sue for malicious prosecution, but they divide on who should win that suit.[Note 4] None deals deeply with Protagoras' "wily scheme", assuming it was to win in either the first or second suit.[Note 5] By bringing the first suit, even if he is certain to lose it, Protagoras guarantees his victory in the second suit. If equity wants to thwart Protagoras' scheme, then holding for Euathlus in the first case does not suffice; it plays into Protagoras' hands.

Many solutions are available to law that are unavailable to logic. For example, the judge could thwart Protagoras' wily scheme by ordering Euathlus to hire a lawyer in the first case. The contract could be reinterpreted, reading its terms in light of the situation or "trade usage". For example, the judge could easily decide that this case will not count as Euathlus's first case, except possibly in future cases looking back. Or the judge could decide that no "meeting of minds" occurred if Euathlus meant his first case with someone other than Protagoras and if Protagoras did not. The contract could be positively reformed in equity. Euathlus could be ordered to pay earnest money while making a reasonable effort to take on another case, or to pay quantum meruit for the time Protagoras had already devoted to his instruction. If Euathlus unduly delayed in taking cases, that might be considered a breach; or if Protagoras was unduly hasty in bringing suit, he might be ordered to wait for a reasonable period. The logical difficulties could be sent to a jury, and no questions need be asked about how the jury arrived at its solution. (This is how the judge in State v. Jones hoped to deal with a similar problem; see Section 20.B below.) The judge could make a forthright appeal to some relevant policy, such as the benefit to future legal education of deciding for Protagoras (or for Euathlus). Without bringing in policies or principles of this kind, in fact, it is difficult to see how the case could be decided, since Protagoras and Euathlus do not give reasons why they ought to win other than the contradiction in the alternative position. If the two arguments are truly equal in weight, then the one with the burden of proof loses; this works against the plaintiff Protagoras.

B. State v. Jones

State v. Jones, 80 Ohio App. 269 (1946) is the only American case that has cited Protagoras v. Euathlus, according to the computer search service, Lexis. It presents a similar paradox and shows that the logic of the Euathlus case can arise in modern law. Jones was a physician charged with performing illegal abortions on each of six women. For at least one of the women, Jacquelin Harris, the only evidence of the abortion was her uncorroborated testimony. The judge instructed the jury (1) that women who voluntarily procured abortions were accomplices of the criminal abortionist, (2) that if Jones is found guilty, then Harris must be found guilty as an accomplice, and (3) that the testimony of accomplices against principals is suspect and cannot ground a conviction without corroboration. These jury instructions were correct statements of Ohio law in 1946.

Jones wanted the second instruction replaced by one to the effect that Harris should, by her own admissions, be found an accomplice as a matter of law. This would have the consequence that her testimony against him would become legally insufficient; hence he could not be convicted of performing that abortion. Jones' motion was denied and the original instruction went to the jury.

Jones clearly wanted Harris' testimony against him to become legally insufficient, and this occurs as soon as Harris is found to be his accomplice. But if she is his accomplice, then he must be the guilty principal. But if he is guilty, then the only evidence against him is the uncorroborated testimony of his accomplice, which is legally insufficient. So should he be convicted?

We can restate the issues as a dilemma and counter-dilemma to highlight their similarity to the Euathlus case. The prosecutor argues that Jones will be found either guilty or not guilty. If guilty, then the state wins. If not guilty, then Harris is not his accomplice and her testimony against him can suffice to convict him, and the state wins. So either way, the state should win. Alternately, Jones should be found guilty. For if Harris is his accomplice, then he is the principal; and if she is not his accomplice, then her evidence against him suffices to convict him. So either way he should be convicted.

Jones also argues that he will be found guilty or not guilty. If not guilty, then he should walk; if guilty, then Harris is his accomplice and her testimony against him is legally insufficient, and he should walk. So either way, he should walk. Alternately, Jones argues that he should be acquitted. If Harris is his accomplice, then her evidence against him is legally insufficient to convict him; and if she is not his accomplice, then he is not the principal. So either way he should not be convicted.

On the prosecutor's argument, it should not matter whether Harris was ruled to be Jones' accomplice or not. On Jones' argument, it should not matter either. But the judge did not instruct the jury that Harris simply was, or that she simply was not Jones' accomplice. The judge told the jury to find Harris to be Jones' accomplice if it found Jones to be guilty of performing the abortion. This puts the jury in a position of returning a self-annulling verdict. If they find Jones guilty, then they must find that Harris was his accomplice, then they must find her evidence against Jones insufficient, then they must acquit Jones. But if they find Jones innocent, then they must (at least may) find Harris' evidence legally sufficient, then they must (at least may) convict Jones.[Note 6]

By hearing and denying Jones objection, the trial judge was certainly made aware of the paradox latent in the instructions, but he felt no compunction about passing it along unchanged to the jury.

The jury found Jones guilty, not guilty-hence-innocent, and not guilty-hence-innocent-hence-guilty.... Jones appealed on the ground that the second instruction was erroneous. The Appellate Court also rejected his claim. It observed that Jones' preferred alternative (that Harris was an accomplice by her own admissions) presupposed Jones' guilt and, accordingly, would violate the presumption of innocence the trial judge was obligated to accord him. Jones' instruction would then lead to a paradox of its own: it would require that he be convicted while at the same time it invalidated the only evidence on which his conviction depended. In a way the higher court found Jones guilty either way, using the dilemma propounded by the state: if Harris is declared an accomplice by the court, then in the same breath Jones is convicted as the principal; if Harris is not declared an accomplice, then her evidence suffices to convict Jones.

Jones is guilty if innocent, and innocent if guilty,[Note 7] just as the Liar's statement ("This very statement is false") is false if true, and true if false. Any judgment of Jones' guilt, or of the truth-value of the Liar's statement, will be subject to reversal. In logic, the reversals oscillate ad infinitum unless some steps are taken to pretermit them entirely, e.g. by changing the usual notions of truth-value, proposition, or other parameters of the Liar's lie. But in law, the reversals only go as far as the human judges and jurors care to take them. The jury's verdict is stable, and appellate court's judgment is stable and, after a point, appeals come to an end.

Jones has never been cited by another case or by commentators, as far as I or Lexis knows.[Note 8]

C. Self-referring laws

Self-reference is common in statutes and legal instruments such as contracts, wills, and treaties. Here is a small list of perfectly harmless examples to show that self-reference per se does not create paradox, incoherence, or meaninglessness.

C.1. Statutes, constitutions, and legal instruments may contain language that explicitly refers to the section or whole of which it is a part, e.g. "Nothing herein shall be construed to prohibit...", "This statute shall be prominently displayed in every hotel room...", " Violation of this treaty shall be referred to arbitration..."

C.2. Laws may apply to themselves without necessarily referring to themselves, e.g. the provisions in a state code on how the code is to be interpreted, any constitutional or contractual amending clause, and the entire Vienna Convention on the Law of Treaties (the "Treaty on Treaties").

C.3. Legal powers may apply to themselves to limit or augment themselves or their users, e.g. a parliament's power to bind itself and future parliaments by a present decision, a sovereign state's power to bind itself by a treaty or contract, a President's power to pardon himself, a legislature's power to delegate legislative power, a contract's power to declare its own integration, a majority's power to establish majority rule, a court's jurisdiction to declare the boundaries of its own jurisdiction.

C.4. Statutes or sections of statutes may be caught in a circle of mutual reference, e.g. A refers to B and B to A, but neither refers to itself. This may lead the human researcher on a circuitous path, but an irreflexive rule of conduct or procedure may be articulated by the rules. The references may form a circle of any size: A refers to B, B to C, ..., Z to n, and n to A. Albert Einstein was probably referring to this phenomenon when he said that the United States Internal Revenue Code was more difficult to understand than the universe.

Many complex statutes contain sections that do nothing but define terms. These sections implicitly refer the reader to dictionaries for the undefined terms used in the definitions, thus putting one indirectly on a very large circuitous path.[Note 9]

C.5. Statutes may as a group inadvertently create vicious circles, e.g. circular liens (see Section 20.E, below)

C.6. Self-reference in language outside law may lead to legal difficulties, e.g. Abbie Hoffman's book, Steal This Book, which was frequently stolen.

D. The liar

The best known reflexivity paradox is the Liar: "This statement is false." The statement is true if false, and false if true. Some variations occur in legal settings.[Note 10] The most direct is the "imperative liar": "Do not obey this law". Alf Ross discusses this variation briefly.[Note 11] In Section 14 I showed the sense in which sunset clauses in statutes and constitutions might be construed to say, after the expiration of the clause, "this law is no longer valid".

The 16th century French courts had to decide how to interpret the testimony of women accused of being witches. According to the prevailing theology, witches were under the control of the devil, and consequently they either always lied or frequently lied. If they always lied, clever interrogation could disclose the truth. If they only lied sometimes, but with devilish ingenuity, then the truth could be hidden forever.[Note 12]

When Urbain Grandier was on trial for invoking the receiving the aid of the devil, his judges had to obtain a ruling from Sorbonne theologians on the reliability of testimony given under oath by one who might be possessed by the devil.[Note 13]

More recently, Melvin Weinberg, the swindler hired by the F.B.I. to devise the "Abscam" sting of members of Congress who accepted bribes, testified under oath that, yes, he did once say that he was the world's biggest liar, but at the time he had been lying.[Note 14]

E. Circular liens and liabilities

Cases often arise in which an estate or fund must be distributed among claimants when A's claim takes priority over B's, B's takes priority over C's, and C's takes priority over A's. The problem takes several, subtly different forms that have received inconsistent treatment from courts. This type of circularity has a surprisingly large literature.[Note 15]

A similar problem arises when an insurance policy declares itself inapplicable to injuries or losses that are covered by other policies. If an injured person has two or more policies that each contain such provisions, then each would become inapplicable. But as soon as they were inapplicable, they would each trigger the other's applicability again, and so on. The policy-holder would either be entitled to nothing or to benefits from at least one policy, to be determined by an unending and indeterminable oscillation of liability.

F. The bootstrap doctrine

Although "bootstrap" is a term used by lawyers to indicate many types of reflexivity, as a term of art "the bootstrap doctrine" designates a court's jurisdiction to determine its own jurisdiction. This means that if in some objective sense a court lacks jurisdiction over someone, the court may acquire jurisdiction over that person by ruling that it has jurisdiction.[Note 16]

The bootstrap doctrine is a prudential and logical necessity. It deters frivolous appeals, disarms otherwise unanswerable objections, saves time and money, and prevents an infinite regress.[Note 17] As such the bootstrap doctrine is a good example of a benign circle, a form of self-justification that is not fallacious or unintelligible and that (while creating some dangers) is clearly preferable to the alternative.

The de jure existence of the bootstrap doctrine also reveals that there is no "theory of types" in our legal system, at least in a court's judgments on its own capacities.

G. Inferences drawn from the fact of the dispute

Parties in disagreement usually debate irreflexively by proposing their own views and arguments and attempting to subvert those of the other side. Rarely do they cite the fact that a disagreement exists as a premise from which to infer a conclusion designed to settle the dispute. If sound, such arguments bypass the merits of the irreflexive reasoning in contention, and settle the dispute dialectically: the dispute, by virtue of being a dispute, ceases to be a dispute and becomes a conclusion.

Divorce cases often hold that the fact that the parties are seeking divorce (or seeking it so vehemently) indicates that the marriage is dead, thus at least partly justifying the decree of divorce. In fault jurisdictions the same inference is more difficult to make, but is still often made: one of the spouses does not want a divorce and is contesting the proposition that the marriage is dead. In Zavin v. Zavin, 366 P.2d 733 (1961) an interesting twist occurs. The husband and wife each alleged outrageous fault by the other, but neither proved any fault. Their heated allegations unmistakably displayed a dead marriage. The court inferred that the marriage was truly dead, from which it inferred that some fault must have existed; but because no fault was proved, the court presumed equal fault. Under the doctrine of recrimination, equal fault prevents divorce, so the court denied the divorce.

In the case of The S.S. Lotus (France v. Turkey), P.C.I.J. Ser.A., No. 10 (1927), the French steamship Lotus collided at sea with the Turkish collier Box-Kourt. In the dispute on liability that followed, France pointed to what it claimed was international customary law favoring its claim. But the Permanent Court of International Justice found that the national courts of the world were divided on the question at issue, and it inferred from this disagreement that there was no customary law at all on the issue. In short, France is wrong because the courts of the world disagree on whether France is wrong. Here the disagreement from which a conclusion is drawn is not that between the parties but between others on the same issue that divides the parties.

In Nassoiy v. Tomlinson, 42 N.E. 715 (1896) the parties disputed whether Nassoiy's commission on a sale of land was 5% or 1%. Tomlinson had already paid 1% and didn't want to pay more. If Nassoiy's claim to 5% was "unliquidated", then his acceptance of the 1% would operate as a settlement of the entire claim. A claim is liquidated (roughly speaking) if the parties agree on its amount. The trial court gave the jury the question whether the claim was liquidated or unliquidated. The Court of Appeals reversed, holding that the question should not have gone to the jury at all, since the very dispute shows that the claim was unliquidated. It held in effect that a disputed question of fact should not go to the jury precisely for being a disputed question of fact of a certain type, namely, a dispute as to whether a claim was disputed.

An activity sponsored by the state is "excessively entangled" with religion, and so violates the "establishment clause" of the First Amendment, if it causes some potential for political divisiveness by creating hostility in the excluded religious groups or encouraging them to compete for government favors. Lemon v. Kurtzmann, 403 U.S. 602 (1971). Can this potential for divisiveness be symptomized by a lawsuit against the state to cease some activity that touches religion? If it can, then plaintiffs win just by complaining; if it cannot, then the criterion is a bit misleading. In Bogen v. Doty, 598 F.2d 1110, 1114 (8th Cir. 1979) a group of citizens and taxpayers of St. Louis County, Minnesota, sought to stop the city board from opening its meetings with a prayer. The court said, "[t]he present litigation and the earlier attempts to seek rescission of the practice demonstrates the political divisiveness of the invocations....However, we do not see this divisive potential as being of the same caliber as...[that] forbidden in Lemon v. Kurtzmann."

Sometimes it is even clearer that courts should draw conclusions from the fact of disagreement when they presently fail to do so. If a judge finds that no reasonable juror could have found for the prevailing party, she may overturn the jury verdict. If the original winner appeals the judge's action, an appellate court will decide whether the jury verdict was or was not so contrary to common reason that no reasonable juror could have held it. Sometimes appeals courts split on this question. When they do, it may seem reasonable for such judges to infer from their own disagreement that a reasonable jury could have found as it did find, but this is not the current rule. Moreover, this change would be difficult to implement. If the appellate judges polled themselves and found disagreement, then they could still only vote to uphold the jury if a majority of them voted to uphold it; sometimes this would require judges voting at the meta-level for an outcome they reject at the object level. But this happens even today when judges who dislike a rule of law nevertheless vote to uphold it on grounds like stare decisis.

What if the appellate judges were unanimous that the jury should be upheld and thus disagreed only with the trial judge? If the judges disagree on their first vote and then vote to uphold the jury on the ground of their own disagreement, then what happened to the "reasonableness" of their first vote? If the second vote also splits, will there be a third on whether the disagreement of the second justifies upholding the jury? Will there be an infinite regress of votes? Is there a method of voting that will assure us that the series of votes will eventually converge on a unanimous holding? Is the disagreement of one vote with the prior votes as significant for these purposes as the disagreement of one judge with another?

In State v. Robbins, 189 N.E.2d 641 (1963) an Ohio Appeals Court ruled that three judge panels must be unanimous in bench trials; a vote of two to one to convict would indicate a reasonable doubt that would overturn the majority's vote to convict. The state appealed this ruling however, and this Ohio Appeals Court was reversed by the Ohio Supreme Court in 199 N.E.2d 742 (1974).

The same disagreement about the standing of disagreement exists for juries. Some states permit less-than-unanimous jury verdicts in criminal cases. If a person is convicted of a crime in such a state by a jury that splits, say, nine to three, then has the defendant's guilt been proved beyond a reasonable doubt? Does the mere existence of outvoted jurors (and their doubts) establish a reasonable doubt that should overturn the majority's vote to convict? In Johnson v. Louisiana, 406 U.S. 356 (1972) the Supreme Court faced this question and held that it does not; even a split jury can convict a person beyond a reasonable doubt. But in Johnson, the Supreme Court was itself split as much as it could be, five to four. Is there a reasonable doubt as to what constitutes reasonable doubt?

In Hibdon v. United States, 204 F.2d 834 (6th Cir. 1963) the court held the opposite —that split juries indicate the presence of a reasonable doubt— so firmly that even the defendant's consent to abide by a less-than-unanimous verdict did not suffice to bind the defendant to a less-than-unanimous verdict to convict.

H. Reflexivities of sovereignty

The concept of sovereignty, even the question whether such a concept is necessary, leads to reflexivities of many types.

H.1. Sovereign or parliamentary omnipotence is, from one perspective, the main topic of this book. Can a parliament by any procedure irrevocably limit its future ability to act? Can it abolish itself? Can the Reichstag turn over all sovereign power to Hitler? Are law-makers above the law? Most modern sovereigns decide that they cannot irrevocably bind their successors, as if they had a kind of continuing omnipotence. As the author of 1 Corinthians 15:27 described the sovereignty of Jesus Christ: he shall put all under his feet, but he cannot put himself under his feet. (See Section 5.B, above.)

Can a sovereign bind itself for the present and future in the decision of a case, the enactment of a statute, or the agreement to a contract or treaty? As W.J. Rees summarizes the literature on this question, "All political theorists have found it logically necessary, therefore, either to deny the existence of legal duties on the part of the government so as to be able to maintain its legal sovereignty...or else to deny its legal sovereignty in order to assert its legal duties...."[Note 18]

Many of these issues are beautifully encapsulated in London Street Tramways Co. Ltd. v. The London City Council, (1898) A.C. 375, in which Parliament ruled that it is bound by its own precedents. The self-justification of this rule (precedent) is only part of the reflexivity tangle in which it became involved, for in 1966 Parliament reversed the rule of London Street Tramways. For a more detailed discussion, see Section 15.B above.

H.2. Can a sovereign bind itself? Nations have a greater liberty to renounce contracts and treaties than private parties, a liberty that seems to derive from their sovereignty. This liberty is ironically an incapacity: an incapacity to be bound by anything, whether the acts of others or of itself.

Under some concepts of human beings, each of us is a sovereign in this sense. Epictetus and Sartre probably agree on nothing else than that we are condemned to be free. Can one validly consent to be unfree? Is a contract to make oneself a slave ipso facto void for incompetence? John Stuart Mill says yes,[Note 19] just as U.S. constitutional law says in effect that legislatures cannot use their legislative power to delegate their legislative power (unduly or excessively). One is "doomed to be free" if one's voluntary binges of irresponsibility are not defenses to liability for acts committed while drunk. The "irresponsibility defense" seems to apply only to cases of involuntary irresponsibility (insanity) and not to voluntary irresponsibility or consensual unfreedom.

H.3. Sovereignty and legitimacy. Is a sovereign legitimate because it is sovereign, or is it sovereign because it is legitimate? On a practical level, should foreign governments be recognized because they are legitimate or are they legitimated by recognition? Must sovereignty by its nature jumble our notions of cause and effect, ground and consequence, this way? Can dictators validly declare the legitimacy of dictatorship? Can the people speaking through their democratically elected leaders validly declare the legitimacy of democracy? If so, then anything can be legitimate, and the concept of legitimacy becomes trivial; but if not, or if no kind of regime or practice can legitimate itself, then nothing seems capable of becoming legitimate.

H.4. Circle of sovereignty. If we accept the model of lawful power that holds power to be lawful only if it is granted or authorized by a lawful power, then it seems that we have only three choices for explaining lawful power: (1) powers are authorized by higher powers in an infinite regress, (2) powers are authorized by higher powers in a series that eventually comes to a halt with a power that is self-granted and self-authorized, and (3) a plurality of powers authorize one another in a circle.

The first of these is plainly false; the second and third appeal to circular concepts. The second may posit a halt to the series in an arbitrary monarch, an unreviewable legislature, or a God whose will is wholly inscrutable except for the certitude that She authorized the sovereign to reign.

Moreover, if we reject the model of lawful power that requires it to be authorized or dispensed by another lawful power, it seems we are equally driven to postulate a self-legitimating power. Hence, some circularity in our concept of sovereignty seems unavoidable, just as metaphysicians have never been able to avoid concepts of self-caused causes and self-justifying first principles as the price of avoiding causal and logical infinite regresses.

I believe the United States embodies the third type of sovereignty above, unless "the consent of the people" is taken as a nonfictitious source of power and legitimacy. Sovereignty in the United States is not monolithic, but plural and systematically divided. While the parts may not positively "grant" power to the others, the parts do negatively "review" or "check" and limit the power of the others. To follow the chain of reviewing bodies, starting anywhere, one will be led either in a circle (of greater or smaller size depending on where one began) or to "the people".

H.5. Sovereignty by default. The persistent, self-positing nature of sovereignty is seen in cases of what might be called sovereignty by default. In the sphere of international law, "monists" argue that when there is apparently no international law on a certain subject, then there really is an international law, namely, "leave it to the domestic law of the individual nations." That is, national sovereignty is itself a dispensation of the international legal system. The idea that every visible form of government is subordinate to an invisible higher sovereign, whether it is logic, morality, natural law, the kingdom of heaven, or international law, explains problematic national sovereignty by referring it to a higher sovereign that holds its authority by self-justification. The alternative is an infinite regress of sovereigns, like the gods behind the gods.

The instinct to posit ever higher levels of sovereignty reminds one of Adam Weishaupt who infiltrated the Freemasons in the late 18th century and became its leader. His "weapon" was the hoax that there was a secret, higher body of Freemansons, called the Illuminati, which he supposedly represented. The love of hierarchy and secrecy were both strong among the Freemasons, and Weishaupt's hoax enabled him to win sovereign power as an invader.

A few American cases defer to various species of "higher" law, including international law. A recent case from the Second Circuit, (Filartiga v. Pena-Irala, 1980) asserts that the international law of human rights is, and always has been, incorporated into the "federal common law". (This ruling enabled a U.S. federal court to obtain jurisdiction over an alleged Paraguayan torturer sued in tort by a relative of a victim when both happened to be in the United States, when the alleged tort did not occur in the U.S., and when neither party was a U.S. citizen.) Article 25 of the constitution of West Germany, written by the allies, defers to international law as superior to domestic German law. The supremacy clause of our constitution (Article VI, §2) gives supremacy to the constitution itself but also to treaties made under the authority of the United States. In Missouri v. Holland, 252 U.S. 416 (1920), domestic legislation implementing a treaty was held constitutional although substantially the same legislation was held unconstitutional before the treaty was signed. There is much literature on this case raising the reflexivity questions whether the constitution can declare its own supersession, whether the supremacy of treaties is an indirect method of amending the constitution, and whether supremacy can be shared. A proposed constitutional amendment to overrule Holland and straighten out some reflexivities in the supremacy clause was defeated (see Section 17.B).

H.6. Self-Created Authority. Some authoritative instruments have a self-created authority. This is slightly different from the self-created sovereignty that may or may not be required to halt the infinite regress of sovereigns. A sign at an airport security point warning passengers that they may be searched is backed by the federal common law on permissible consent searches. But it is the sign itself —its existence, intelligibility, and conspicuousness— that gives the notice that makes the search permissible, and thus that makes the sign's text true. Similarly, a conspicuous copyright notice validates itself.

In the law of evidence, excited utterances are admissible hearsay only because the excitement of the speaker is an assurance of reliability. The showing of an exciting event is a prerequisite, but in most jurisdictions the existence of the alleged exciting event may be inferred from the utterance. Hence the warrant for the reliability is self-established.

Nine states have had, at one time in their history, a constitution without an amending clause.[Note 20] Yet each has been enabled to amend its constitution. The authority to do so has been recognized on various grounds, as inherent in the people or the legislature or a constitutional convention. The law and scholarly opinion on the self-creation or validation of such amending power is well-summarized in In Re Opinion to the Governor, 55 R.I. 56, 178 A. 433, 449-51 (1935).

H.7. Self-Destroying Authority. Some authority is self-destroying. My favorite example is the ex parte annulment of marriage. A proceeding is ex parte if it is "one-sided" in the sense that only one of the parties whose rights is being adjudicated is present in court. The permissibility of ex parte divorce is an exception to the normal rule that courts must have personal jurisdiction over a party before it may adjudicate her substantive rights. The fiction that marriage is a thing (res) that travels with each spouse is used to justify the ex parte divorce. The court granting the divorce has jurisdiction in rem over the thing, marriage, not jurisdiction in personam over the spouses. But while this might make some sense for divorce, it makes no sense at all for annulment. While divorce simply terminates an existing marriage, annulment declares that there was never a marriage or that it was void from the beginning. In our system ex parte annulments are permissible under a theory of in rem jurisdiction. So when a person goes to court for an ex parte annulment, the court adjudicates the rights of the absent party by virtue of its in rem jurisdiction over the "marriage" which is in the courtroom. If the annulment petition is granted, then the marriage never existed; the court's decree subverts and eradicates the jurisdictional basis of the petition. The ground of standing is self-destroying. Decrees of ex parte annulment should be self-overruling, but they are not.

Article 38(1)(d) of the statute creating the International Court of Justice acknowledges the law-making significance of scholarly writings (for international law). What if a scholar wrote an essay deploring this provision, arguing that only the acts of states, not those of private individuals, should affect international law? What if the issue became controversial, and scholarly opinion divided on the wisdom of the provision? What if this paradox of self-repeal became controversial?

The presumption of death that arises from a seven year absence is rebuttable by the person's reappearance. As a presumptively dead person, does the returnee lack standing to prove her reappearance, say, in the face of a charge of fraudulent impersonation? (The dead usually lack standing.) Or is legal life dispensed in order to prove legal life, the presumption held rebutted in order to permit the rebuttal?

H.8. Renvoi. Some statutes and contract terms tell us which law should govern injuries or breaches. For example, if a defective tire explodes and causes a car to hit a pedestrian, we turn to these statutes to ascertain whether a suit by the pedestrian against the tire manufacturer should be decided under the law of the state where the accident occurred, the state where the tire was manufactured, the state where the manufacturer is incorporated, or some other place. If a contract is violated, a term in the contract usually tells us whether to decide rights and remedies under the law of the state where the contract was made, where one of the parties is incorporated, where the violation occurred, or some other place. Since the laws of different states and nations can differ significantly, these choice-of-law rules can be decisive.

Suppose we are citizens of nation A, and file suit against a foreign corporation in one of our own courts. The local choice-of-law rules tell us to apply the law of nation B. We consult B's law and find that its choice-of-law rules direct us to apply the law of nation A. A conflict or circle in these "pointers" to applicable law is called renvoi. Because it occurs frequently, renvoi has a more specialized vocabulary. If nation B sends us back to nation A, that is remission; if it sends us on to nation C, that is transmission.

Renvoi can be prevented if, when we are directed by A's law to look at B's law, we look only at B's substantive law, not B's choice-of-law rules as well. This is the solution adopted by a majority of courts today. Another solution, called "partial renvoi", is to accept the renvoi from B back to A, but once back in A to look only at A's substantive law, not A's choice-of-law rules again. The "non-solution" of looking always at both the substantive and choice-of-law rules of a jurisdiction is called "whole renvoi". In theory whole renvoi can lead to an infinite (that is, unending) loop. But because we are dealing with law, not software, it never does so in practice. Some court always changes the rule from whole renvoi to partial renvoi, at some point in the burgeoning regress, and terminates the renvoi.[Note 21]

The general practice today is to adopt different solutions for different kinds of cases. For example, whole renvoi is usually used in cases of divorce and title to land, and must be used in tort claims against the federal government. This means that the exact injury or complaint, in its full historical context, must be subsumed under traditional legal categories before we know which kind of solution to renvoi to use, and hence before we know which state or nation's law applies to the case. But many concrete injuries are borderline cases that are put into different categories by the laws of different places. Hence we have a meta-renvoi paradox: sometimes we must classify a wrong in order to know what law applies, and we must know what law applies in order to classify it.[Note 22]

Because renvoi problems are actually solved without incurring a genuine infinite regress, they are especially good examples of problems that are paradoxical for formal logic but not for law. Something important about the nature of law is shown merely by the fact that they are soluble in law, like the paradox of self-amendment. Fiori Rinaldi concludes his discussion renvoi by saying[Note 23]

These clashes arise from adherence to formal logic and can generally be resolved only by 'breaking out' of logic itself.

Laurence Goldstein follows J.C. Hicks in comparing the logic of renvoi to the logic of the irreflexive liar:

A. Statement B is false.
B. Statement A is true.

But Goldstein uses the legal solubility of renvoi problems to explain how it is that law can solve paradoxes that logic cannot.[Note 24]

[T]he crucial difference is that in the case of [statements A and B] as in the case of any statement, it makes no sense to stipulate that one of them is true, since to say that a statement is true is to say that it corresponds in some way to some state of affairs in the world, and we cannot stipulate how things are in the world. But we can stipulate what laws are to prevail and this is in fact just what we do do.

I. Self-amendment

Self-amendment appears in many contexts other than constitutional amending clauses. If a written contract stipulates that only written modifications will be effective, may that provision be modified orally? It is reflexive either way: if it is orally modifiable, then self-excepting, and if not, then self-applicable. Can a contract's no-waiver clause be waived? Why can a will's no-revocation clause be revoked? Why can a will's no-contest (in terrorem) clause be contested? (See Section 9.C above.)

In American states that allow home rule, the "pyramid of power" may be inverted by a vote of a lower level. Is the lower level thereby made an apex power, or an apex power only subject to the power of the apex of the original pyramid?

How can a sovereign release a subject from subjection? If a parent tells a child, "you are now completely free of my commands," then there is a sense in which the child owes its liberty to that parental dispensation; if the parental fiat is revocable, then freedom cannot be conferred by another.[Note 25] This problem can become very real and troublesome. As England's former colonies and dominions won their independence, by war or negotiation, England wanted a way to seal their independence by law. In 1931 it adopted the Statute of Westminster, solemnizing its intention never to legislate for former colonies again without their express consent and request. That was direct and thorough; the problem is that the Statute of Westminster is a statute and can be repealed, which could mean that the independence of the former colonies and dominions is revocable by England.[Note 26] This was quickly noticed by courts in the former colonies. It appeared that one generation of English people had done its best to surrender this right to legislate, but had done so with frustrating incompleteness, since any later generation could restore it. This was a world-class cartoon of the child with flypaper on its fingers trying to shake it off. Nothing that England could do, it seems, could give the colonies full legal independence, for if it were done in law, then it could be undone; and if were not done in law, it would not be lawful. England was learning that it is paradoxical to command another to be free or even to offer another their freedom as a gift.

One tempting solution was to interpret the Statute of Westminster as irrevocable. That would guarantee that the emancipated countries would stay emancipated. But it would also contradict the independence of the English people, their sovereignty in their own country, and deny them the power to change their own laws. More succinctly, the English Parliament for these reasons cannot bind its successors irrevocably. Following this line of reasoning Parliament ruled in 1935 that the Statute of Westminster could in principle be repealed. British Coal Corporation v. The King, A.C. 520. So the paradox of liberation remains.

To avoid the problems of revocable independence, many former colonies have deliberately inserted some irregularity or procedural defect into the ritual of liberation so that they could say they owe their independence to peaceful revolution, not to the Statute of Westminster.[Note 27]

The equivalent problem arises in U.S. law in the Philippine Independence Act of March 24, 1934, in which the United States agrees to give the Philippines independence in exchange for making a constitution for themselves according to U.S. specifications. In Section 10 (a), the United States promises that, when a suitable Philippine constitution is ratified, "the United States shall by proclamation withdraw and surrender all right of possession, supervision, jurisdiction, control, or sovereignty then existing and exercised by the United States in and over the territory and people of the Philippine Islands...." The Philippine Independence Act, like the Statute of Westminster, is merely a statute; if Congress cannot bind itself irrevocably, then the statute can be repealed by the United States at any time. If after a certain time repeal would have no effect on the independence of the former dependent, which is almost certainly the case, then legal formalism cannot explain it.

J. Permissible disobedience

The phenomenon of de jure toleration or justification for acts that depart from requirements of valid law presents a neat contradiction. Reflexivities enter in different ways. Disobedience of certain kinds may be permitted because the legal system as a whole is perceived as self-limiting, because a statute is self-excepting, because the disobedient act is self-justifying, because the act is self-concealing and unreviewable, or because the act itself, while disobedient to existing law, has law-making power and so amends or repeals rather than merely violates.

The decisions of a common law judge inconsistent with relevant precedent (consciously overruling it, or simply disregarding it) are in an important sense departures from valid law, but they become law themselves. The precedents they violate are amended or overruled pro tanto or to the extent of the conflict. In international customary law, a dearth of custom may exist on a certain issue but all of it may point in the same direction. An act departing from that custom may be a delict or the beginning of a new custom —a violation or an amendment.

Some laws may mischievously be interpreted to mean, in effect, that certain acts are forbidden only until committed. Suicide is the classic example, but there are many more interesting examples. The current international law on the minerals of the deep seabed arguably says that the minerals extracted are the property of the taker while the minerals in place are the common heritage of mankind and not to be taken. Some commentators have suggested that it would be illegal for an oil company to drill for oil in the deep seabed, but that any oil so taken would legally belong to the oil company. The situation is similar to drug laws that penalize sale and purchase, and perhaps possession, but not use.

Much law and commentary exists on the problem of "implied repeals" of old statutes by new statutes, old statutes by new cases, old cases by new cases, and old law of all kinds by new interpretations. These cases show the importance of distinguishing the type of law-making act employed in the earlier and the later laws in conflict. Otherwise we will never clearly answer the question whether new acts inconsistent with existing law are violations or repeals.

Laws may be self-excepting when they require their own publication in order to take effect and then are not published on account of some intrinsic feature. This happens occasionally with administrative regulations or state statutes on obscenity. In order to avoid the vagueness problems that might invalidate them under the First amendment, they must be graphic and explicit. But that makes them obscene to many people. In one case a publisher refused to print the statute because he thought he was prevented by the very statute to be published. The paradox is tightened when we realize that, for reasons of notice and due process, the statute is of no force until it is published.

International law forbids intervention in the internal affairs of another state (with some exceptions) about as clearly as it forbids anything. The value promoted by this prohibition is the self-determination of the people of the state. Currently international law is very much unsettled on the question whether intervention is permissible to further self-determination, e.g. by aiding democratically minded insurgents against a dictator's army.

K. Contract reflexivities

Contract law contains many types of reflexivity, several of which have already been mentioned under other rubrics.

K.1. Reflexive v. irreflexive theories of contract. In the text (Section 7.A) I call the view that contracts bind ex proprio vigore or by their own strength, the "reflexive theory" of contracts. The alternative is an irreflexive view that contracts bind only if a rule of public law external to the contract states that contracts bind. Only the reflexive view can be used in social contract theories of government and their variations. The irreflexive view is supported for ordinary contracts, but not decisively, by the fact most nations do have rules of public law describing when contracts bind.

K.2. Overweening contract terms. On the irreflexive view of contract, contract terms may create a reflexive hierarchy by purporting to jump out of their low level of subordination and control action at higher levels. If contracts are private agreements regulated by public law, then the private agreement cannot, without paradox, pretend to overrule, circumvent, or supersede the public law that regulates it. Insofar as contract terms succeed at this overweening, they create reflexivities in the hierarchy of legal types.

The relation of treaties to the environing international law is slightly different from that of contracts to the public that regulates them. Treaties create public law between the parties and are frequently taken as evidence of the international law that regulates them.

Sample contract terms raising this problem: "This contract will bind the two companies even if the signing representatives had not the authority to bind their respective companies." "This contract is integrated."[Note 28] "The signatories of this contract are competent to bind themselves and so bind themselves." "If this contract is rescinded, the other party may demand arbitration." "No conditions precedent apply to this contract." "This contract will be deemed accepted by the offeree through any language to the effect that the offer is refused."

Can a treaty require nations to become signatories? Probably not. But can a treaty require its signatories to ratify the treaty in accordance with their domestic procedures for treaty ratification? Article III of the Kellogg-Briand Pact of 1928 requires just that. Does a disobedient nation violate the treaty or, precisely because it hadn't yet ratified, escape its strictures?

A logical analogue of this problem occurs in games with rules that feign to regulate non-playing and meta-playing behavior, e.g. that impose a game penalty for refusing an invitation to play or that refuse to recognize forfeits.

Many contracts and treaties contain provisions defining what shall constitute a breach, repudiation, and rescission. May one breach or repudiate these terms even in principle? If one repudiates these terms (in a way included or not included in the repudiation definition), can one claim that one may then breach the breach provision without penalty? Can the contract's repudiation provision always outflank attempts to repudiate it?

K.3. A Classical Problem. A popular medieval puzzle can be analyzed as a contract. Socrates stands in the market and publically offers to pay one drachma to the first person to approach him and tell him a true statement. Plato approaches before anyone else and says, "You will not give me the drachma." Result?[Note 29]

An interesting variation occurs in Don Quixote.[Note 30] Sancho Panza became governor of Barataria and had to judge its hardest cases. A Baratarian bridge had a gallows at one end; those crossing it had to state their business and liars were hanged immediately. A mischievous traveler once said that his sole business was to be hanged. Sancho Panza decided the case by use of legal devices that entitled him to ignore all the logical difficulties. He appeals to a presumption in favor of mercy when the judge is confused. Note that while Sancho tried to avoid the reflexivity of the problem, his solution is itself reflexive; he draws a conclusion settling the dispute from the undecidability of the dispute.

K.4. Circular Contracts. Some tripartite contracts take this form: A promises to do something for B, B for C, and C for A. This circle of duties is the first reflexivity, though it is not logically different from the two-step circle of reciprocity in an ordinary bilateral contract. A stalemate breach may occur: each is willing to perform but none wants to go first. Stripped to its logical bones, mutual promises of "I will do x if you will" create a contract, but the circle of conditions makes it undetermined and perhaps undeterminable whether the parties are actually bound to do x. Finally, the closing of the circle is apparently not a condition of the duty to perform: if A breaches with regard to B, then B must still perform for C, and C for A.

K.5. A Vicious Circle. Scholars have frequently noted the circularity of the 19th century rule that a promise is not binding unless given in exchange for a binding promise. The decline of the doctrine is a rare example of legal change inspired by the desire to avoid paradox. If the doctrine is taken seriously, then a vicious circle would prevent the creation of any contracts. Similarly, if a person must give some valuable "consideration" to make a contract, then what consideration is given in a contract to do what one is already obligated to do? One theory was that to give the other party the contract right to compel one is itself consideration. But if so, then the contract creates the consideration and the consideration creates the contract.[Note 31]

L. "More of the same"

In contemporary English, double negation is equivalent to affirmation, but in Chaucer's time it expressed emphatic negation. Is capital punishment a cancellation of the initial wrong or a doubling of it? When a wrong is repeated, especially in the reflexive manner of victimizing the offender, is the original wrong negated or doubled? Is "an eye for an eye" the model of simple justice or simple barbarism? For Kelsen, acts of violence are either crimes or punishments; hence, repetition can neutralize. For Bentham, it depends on the circumstances; crimes and punishments are both evil and the question is whether the good produced by the punishment outweighs its evil.

Some courts have held that the independent negligence of each of two defendants in one accident is the equivalent of no negligence and no liability by either. For example, if A negligently supplies B with a car lacking brakes, but B negligently fails even to apply the brakes and hits C, then neither A nor B is liable to C. Sanders Systems Birmingham v. Adams, 117 So. 72 (1928). If A negligently fails to signal for a left turn, and if B negligently fails to look but wouldn't have seen A's signal even if she did look, and hits C, then neither A nor B is liable to C. Rouleau v. Blotner, 152 A. 916 (1931).

If sexual harassment of female employees is unlawful sex discrimination under Title VII of the Civil Rights Act, then is the equal harassment of male and female employees a double violation or no violation at all? Courts have not yet faced these facts, apparently, but dicta in earlier cases suggest that it would be a double violation.

The old doctrine of recrimination in divorce law, now generally abandoned, insured that doubly broken marriages in which both spouses were at fault were less likely, not more likely, to be legally dissolved than marriages in which only one spouse was at fault.

A later will inconsistent with an earlier will revokes the earlier will. If A is given money in an earlier will, and is given a different amount in a later will, then does the second gift supplement or supplant the first? Is more money "inconsistent" with less money? See Gould v. Chamberlain, 68 N.E. 39 (1903).

In international law the absence of any analogue of centralized police power has given rise to a "common law" of retorsion: a violation of international law may be answered by the victim state with a retaliatory act that would otherwise be unlawful. The doctrine is unclear on the question whether retorsions are lawful acts of reprisal or unlawful acts that are tolerated for prudential reasons. Can two wrongs make a right, or only satisfaction?

If incompetents can only make voidable contracts, why are they considered sufficiently competent to exercise their option to void? Is the second decision competent as a matter of law? Has the law created a fictitious competency in order to avoid a self-negating method of protecting the incompetent?

Federal courts have used new quotas to remedy the harms of invidious quotas, and new gerrymandering to remedy the harms of invidious gerrymandering. This may be called a legal theory of inoculation. Why is the altered repetition of the offense a remedy in these cases, when no form of battery is thought a good remedy for mugging, beating, or rape? Is affirmative action a doubling of unlawful discrimination on sex and race lines or a justifiable remedy for earlier instances of that discrimination? To conservatives, the replication is additive (two wrongs); to liberals subtractive (a right). If the point is to treat people equally, affirmative action will be additive, a new offense; if the point is to undo the effects of past discrimination, affirmative action will be subtractive, compensation. Do conservatives and liberals differ in their logic or only in their political principles?

In states with anti-gambling statutes, prosecutors frequently try to apply the statutes to those pin-ball machines that reward good scores with free replays. These prosecutors evidently believe that free replays either encourage or constitute gambling. But if pin-ball is legal at all, then only the gift of replays is allegedly illegal. The prosecutors, then, believe that "more of the same" can be too much of a good thing, and rather than multiply legal acts, actual cross the line of legality. See 89 A.L.R.2d 815. Like many other "more of the same" cases, a quantitative change becomes at some point a qualitative change, just as lowering the temperature of water by degrees gives us nothing but cold water for a while, and then suddenly gives us ice.

Should a free society tolerate intolerance? The First Amendment has been interpreted to require tolerance even of those who oppose its values. But many have advocated that we draw a line at opposition to that underlying value, and cease tolerating the intolerant. Do the two camps disagree on the meaning of freedom, the strategies of preserving or maximizing it, or the logic of self-application?[Note 32]

Is ordinary compliance with law an offense? What if the laws are unjust? A commonly heard justification for nonviolent civil disobedience is that it is the most just response to unjust law. This position sees violent disobedience and compliance to unjust law as additive, not subtractive, in the cumulation of evil.

M. Breaking vicious circles

From a normative standpoint, circular causal chains may be self-bettering or self-worsening situations. The latter are usually called vicious circles (but that term applies equally to causal and logical circles). Agents within the causal vortex are usually helpless to break or reverse the circle; on the contrary, the viciousness of vicious circles lies in the fact that those caught in them have reason or cause to contribute to their perpetuation. In a wage-price spiral, the employer and employee equally seek their interests in the self-worsening trap.

Courts have frequently acted as if they were uniquely situated to break such vicious circles, as agents outside the circle with strong equity powers able to act quickly and in conformity to the needs of the special situation. Residential segregation leads to school segregation and vice versa until a vicious circle spins of itself. This circle was broken in Swann v. Charlotte-Mecklenberg Bd. of Education, 402 U.S. 1 (1971). Inadequate education for blacks leads to poor performance on job application tests, which leads to worse jobs, lower incomes and less adequate education. This circle was broken in Griggs v. Duke Power Co., 401 U.S. 424 (1971). District lines that deprive voters of equally weighted votes lead to unrepresentative legislatures, which lead in turn to the perpetuation of such lines. This circle was broken in Wesberry v. Sims, 376 U.S. 1 (1964). Indigent criminal defendants who could not pay their fines formerly had to stay in jail until they could do so or else "work them out" in jail at a miserly rate. This not only prevented such defendants from earning the money to pay their fines, but also kept them poor. This circle was broken in Williams v. Illinois, 399 U.S. 235 (1970).

The state of Idaho formerly favored men over women as administrators of estates. It argued that this decision was not arbitrary because it reduced the workload on probate courts in deciding who is most fit to serve as administrator, i.e. it is not arbitrary because there are some advantages to arbitrary preferences. This vicious circle would justify any kind of discrimination. (It saves court time to decide contract cases in favor of the richer party, and police time to arrest only black people.) This circle was broken when Idaho's preference was struck down in Reed v. Reed, 404 U.S. 71 (1971).

Before the Social Security Act was passed, states did not want to tax the employers in their jurisdictions to provide a relief fund, for that would have driven them out; or they did not want to be the first or only state to tax employers. In upholding the constitutionality of the Social Security Act, Justice Cardozo recognized the need to break this vicious circle with federal action. Stewart Machine Co. v. Davis, 301 U.S. 548 (1937).

A good example of a court creating a vicious circle when it had an opportunity to break one is Warth v. Seldin, 422 U.S. 490 (1975). In that case the plaintiffs claimed that they were too poor to buy housing in Penfield, New York, on account of that town's zoning ordinances; hence they claimed to be victims of unlawful discrimination. The Supreme Court held, in effect, that the plaintiffs were too poor to have standing to complain of their poverty. Because they were so poor, they would be unable to buy housing in Penfield even if the zoning ordinances were changed; hence they were not actually injured and lacked standing to complain.

N. To know before we know

A recurring type of vicious circle occurs when an inquiry designed to yield knowledge of x is triggered, or can only be triggered, by knowledge of x. If we could not know before we knew, then the mere logic of the situation would prevent us from ever starting the inquiry. Another form of the same circle occurs when we must make a decision based on knowledge to be gained from the inquiry triggered by our decision.

Normally Congress has the power to investigate any matter on which it may legislate. It cannot legislate on wholly intra-state commerce, so it cannot investigate wholly intra-state businesses. But it may investigate a business to determine whether it is wholly intra-state. This investigation seeks its own justification. Either it will be a permissible investigation in any event, or it may in principle discover that it had no right to discover anything. Either Congress knows before it knows (perhaps corrigibly) that it may investigate that business, or it possesses a second type of valid investigative power. U.S. v. DiCarlo, 102 F.Supp. 597 (1951).

Again, the investigative power of a legislative committee is normally limited to the scope of the committee's legitimate legislative interest. If a witness refuses to answer a question she considers irrelevant to that legitimate interest, then the committee could formerly hold her in contempt, as if the committee alone could decide such relevancy and as if it knew before it knew that it was legitimate to compel the answering of the question. Bart v. U.S., 203 F.2d 45 (1953), rev'd 349 U.S. 219 (1955). The dissent in Gibson v. Florida Legislation Committee, 371 U.S. 539 (1963) explicitly recognizes the "know before we know" nature of the problem: the witness' answer may be compelled only by a claim of relevancy that "requires an investigating agency to prove in advance the very thing it is trying to find out."

The court was just as explicit in Douglas v. California, 372 U.S. 353 (1963) when it said of a California rule giving indigents the right to counsel on appeal only if the appellate court thought the indigent's case had merit, that "[t]he appellate court is forced to prejudge the merits before it can even determine whether counsel must be provided."

Injunctions are normally given as remedies only if the plaintiff can show that irreparable harm would otherwise result to her. When the injunction halts an action that the plaintiff could sue about if it occurred, which is usually the case, then the granting court is really deciding the merits of the plaintiff's future case before the future case materializes.

In judicial investigations problems arise that are similar to those created by legislative investigations, e.g., when a witness testifies to the grounds of her own competence, such as possession of personal knowledge, or testifies to the existence of her own privilege. In criminal trials witnesses may invoke the Fifth Amendment privilege against self-incrimination without triggering any inquiry into the merits of the invocation. Obviously if witnesses could be compelled to give evidence that certain testimony would incriminate them, to justify invoking the privilege against self-incrimination, then the privilege would be pointless. So the privilege in effect makes certain claims self-warranting, or self-insulating and unreviewable. This does not stop courts from denying a witness' claim of self-incrimination occasionally, e.g. if immunity has already been granted. But normally the privilege is the witness' virtually for the asking, as if we knew before we knew that it was due —or as if we had decided that it is better all in all to err on the side of indulgence than stringency. The normal rule is that the claim of privilege is sustained unless it is "perfectly clear" that answering the question "cannot possibly" incriminate the witness. Hoffman v. U.S., 341 U.S. 479, 488 (1951), quoting Temple v. Commonwealth, 75 Va. 892, 898 (1881).

Despite what Hoffman and Temple make perfectly clear, this rule about self-incrimination does not extend to a president's claim of "executive privilege". Richard Nixon lost his argument that executive privilege was his for the asking, and that no court could inquire into its propriety or use compulsory process to investigate its justification. Nixon v. Sirica, 487 F.2d 700, 708-16 (D.C.Cir. 1973).

Witnesses may be required to testify to enough contextual information for the court to decide whether the invocation of the privilege against self-incrimination is justified. Judge Learned Hand noted the paradox of this in U.S. v. Weisman, 111 F.2d 260, 261-62 (2d Cir. 1940): "Logically, indeed, [the witness] is boxed in a paradox, for he must prove the criminatory character of what it is his privilege to protect because it it criminatory."[Note 33]

In principle courts can limit the privilege to those questions which, if answered truthfully, could yield self-incriminating testimony. But in practice they rarely do so, for to make this a regular practice would encourage prosecutors and jurors to infer criminality from the invocation of the privilege against self-incrimination, which would defeat the purpose of the privilege. It is difficult to prevent the inference of criminality; at least counsel cannot argue to the jury that the witness has incriminated herself by invoking the privilege against self-incrimination.[Note 34] Griffin v. California, 380 U.S. 609 (1965). But if a man made a contract with his employer not to disgrace the company, and then invoked the privilege against self-incrimination in a court, a jury can decide whether the invocation was a breach of the contract. Loew's Inc. v. Cole, 185 F.2d 641 (9th Cir. 1950), cert. den. 340 U.S. 954 (1950).

Similarly, a statement to a doctor or lawyer is privileged only if made to the doctor or lawyer in her professional capacity. But this can often not be determined without actually hearing the statement. Nevertheless, the determination is made without hearing the statement.

A stark "know before we know" paradox occurred in Matz v. U.S., 158 F.2d 190 (1946). A man prosecuted for bigamy argued in defense that he was only married to one woman; his first marriage was defective, hence void, and only his second marriage was valid. The state offered the second "wife" as a witness against him, but the man objected under a statute disqualifying a wife to testify against her husband. The man's objection would have to be sustained if the woman were really his wife, but that was precisely the question at issue in the trial. For if his first marriage had been valid, then his second would have been defective. Hence the admissibility of evidence on a certain question could not be decided until that very question was answered, requiring the judge to know before he knew. His ruling evaded the vicious circle by substituting a less vicious circle that still required him to know the man's guilt before the trial determined it: he ruled that he, rather than the jury, had to decide the validity of the first marriage.[Note 35]

The Supreme Court has original jurisdiction over ambassadors and cases in which states are parties (Article III, §2.2). Suppose a borderline example of such a case came before the Court as if entitled to original jurisdiction, e.g. an allegedly discharged ambassador, or a suit between state administrative agencies. The Court must decide whether it has jurisdiction before it decides the merits, but it can't decide jurisdiction without, in effect, conceding original jurisdiction. Logically similar but less politically dramatic problems can arise for any court, since every court's jurisdiction has boundaries.

This "know before we know" problem is solved by the device of "special appearance" in which a court can hear arguments for and against its jurisdiction. Only if the court decides that it does have jurisdiction will the parties be required to make "general appearance" for the adjudication of their rights. This does not prevent all reflexive paradoxes, however. In the case of the Supreme Court's original jurisdiction again, if the borderline case raised the question whether a certain woman was an ambassador, then the question of original jurisdiction could not be decided until the merits had been decided.

One reason why the bootstrap doctrine (Section 20.F above) is controversial is that if a person comes to court under special appearance solely to deny the court's jurisdiction over her, the court may convert the special to general appearance and assert general jurisdiction over her. York v. Texas, 137 U.S. 15 (1890); Chicago Life Ins. Co. v. Cherry, 244 U.S.25, 30 (1917). In short, to prevent the reflexivity problem of deciding the merits of a case before deciding jurisdiction, and knowing before it knows, courts use the device of special appearance. But use of it in conjunction with the reflexive bootstrap doctrine creates the alternative reflexivity problem of self-validating judgments of jurisdiction by courts that can, in effect, use dissent as consent to a court's jurisdiction.

Bail is excessive under the Eighth Amendment if it is greater than necessary to assure that the defendant will appear for trial. The risk of flight is the major factor in setting bail, and that risk may not be inferred from the indictment, even if it specifies an offense (e.g. conspiracy) that suggests that the defendant is able and likely to flee. Stack v. Boyle, 342 U.S. 1 (1951).

An interesting example of a reflexivity that was obviously enacted to avoid a "know before we know" paradox is Federal Rule of Evidence 104 (a). It provides that none of the Federal Rules of Evidence applies to the judge's determination whether offered evidence is sufficiently relevant to be admitted, or whether a witness is qualified or privileged. Without Rule 104 (a), evidence inadmissible under law would have to be admitted so that the judge could decide its admissibility —or else the judge would have to know before she knew in each case.

Effective-date clauses raise "know before we know" paradoxes insofar as they tell us authoritatively, before they are effective, when they and the statutes to which they attach will become effective. (See Section 14.C for more discussion.)

O. Tax reflexivities

O.1. Pyramiding Taxes. If an employer pays an employee's income tax as an extra benefit, then the amount of the tax paid is considered to be extra income to the employee. Because it is income, it is taxable. If the employer pays the tax due on that extra increment of income, and then on the next increment due on her second payment, and so on, then an infinite series is created. However, the series converges and the exact amount of tax due can be calculated, which is always more than the employee would have to pay if she paid her own taxes directly. Hence this perquisite is rarely used.

Whether taxes pyramid in this way is a matter of policy. They are allowed to do so in the case of employers and employees. But to encourage leases in which the lessee pays the lessor's taxes on the leased property, §110 of the Internal Revenue Code (IRC) stops the pyramiding by stipulating that payment of the lessor's taxes is not considered income to the lessor.

Pyramiding down and up in oscillation would be possible if the late-filing charge or other penalties could be deducted in an amended return. The larger deduction than originally taken would decrease the tax due, decreasing the penalty, decreasing the allowable deduction, increasing the tax due, and so on in a series that would probably always converge. IRC §162 (f) prevents the deduction of such penalties.[Note 36]

O.2. Mutually Deferring Provisions. Under IRC §1015 (a) the taxpayer's basis on the sale of a gift, if a gain, is the donor's basis, and if a loss, is either the donor's basis or the fair market value at the time of the gift, whichever is lower. Suppose the donor buys something for $200 (donor's basis), gives it to the donee when it is worth $100 (fair market value), and the donee sells it for $150. Then the sale is both a gain and a loss. If the sale is considered a gain, then we use the donor's basis ($200), yielding a $50 loss. If the sale is considered a loss, then we use the lower of the two bases ($100), which yields a $50 gain. If the sale is a gain, then it is a loss; and if it is a loss, then it is a gain. The Code does not solve this problem, but Treasury Department Regulations do. Regulation §1.1015-1 (a) cuts the Gordian knot by allowing no deduction and imposing no tax on such transactions, making it immaterial whether they were gains or losses.

O.3. Imputed Income. If one performs a service with a market value for oneself, one may be said to have "earned" the value of the service as income. Such "earnings" are called imputed income, and are traditionally not taxed. For example, a housepainter who paints her own house is not taxed on the value of her work, nor is the homeowner who lives in her own house taxed on the imputed rent. But an insurance agent was once taxed on "commissions" that he "earned" on policies that he bought from his company for himself. Commissioner v. Minzer, 179 F.2d 338 (1960). Minzer claimed that the "commissions" were subtracted from the purchase price, giving him a discount and no income. The IRS claimed that he performed a service for compensation. Minzer paid the price of his policies without the commission, whereas other buyers pay both the price and the commission. Because no commission was paid, none could have been received as income in the normal sense. But just because none was paid, one was "earned" as imputed income —a benefit with a cash value in the market place was received in exchange for labor. Because Minzer was dealing with himself, he earned a commission because he paid none, and he was taxed on it.

Other questions about self-dealing from outside tax law: Why are contracts with oneself void? In a slave-state where slaves were chattels, could a slave take title to herself?

P. Circular reasoning

Judges and lawyers are not immune to the fallacies of circular reasoning, and many types frequently occur. But what is more interesting is circular reasoning that is permitted or required by law, without regard to its logical validity. We've seen many cases under other heads above; here are some others.

P.1. Mutual Corroboration. In order to minimize the risk of collusion in divorce cases in fault jurisdictions, many states have statutes that require that a spouse's fault be shown with evidence other than her admission of fault. In Husband v. Wife, 253 A.2d 63 (1966) the wife denied committing adultery or confessing to adultery to her husband, but he presented circumstantial evidence that she had both committed adultery and confessed it to him. The circumstantial evidence of the wife's alleged confession was legally insufficient taken alone, but the court reasoned that the circumstantial evidence corroborated the husband's in-court testimony, which in turn corroborated the circumstantial evidence. Together these mutually corroborating stories showed (the court held) that the wife's denial was perjured, and hence that she was guilty by a sufficiency of evidence. The court unabashedly concluded, after a string of citations, that the wife's "perjury impels acceptance of the husband's testimony concerning her admissions. We accordingly have a situation in which the indisputable circumstantial evidence corroborates the husband's testimony, which in turn corroborates the circumstantial evidence." Whether the statute designed to minimize collusion required this result is one question; another is whether the court should have inferred something from the fact of the dispute, namely, that a contested divorce does not present a large risk of collusion.

A generalized version of the problem above occurs whenever a mass of evidence is found sufficient when the bits comprising the mass are separately insufficient as a matter of law. The support that each bit provides for and receives from the others, by completing a coherent picture, is commonly and probably harmlessly found to be persuasive. But logically it is difficult to distinguish from the mutual proof of two uncertain propositions. See Peters v. U.S., 408 F.2d 719 (1969), dissent: "Adding hearsay to hearsay [even in administrative hearings where hearsay is admissible] is like adding zero to zero which still equals zero."

In 1967 the Supreme Court struck down a Texas statute that barred partners in crime from testifying for one another, and thus permitted mutual corroboration "for what it is worth". Washington v. Texas, 388 U.S. 14 (1967). The statute, and many like it in other states, was evidently designed to prevent the simple mutual proof of two false or uncertain propositions, an evil to be feared because it is made attractive by the self-interest of the witnesses. But the Supreme Court held, in effect, that the circle does not prove falsehoods without the help of a jury, and that juries are not likely to be fooled and in any event are entitled to believe.

P.2. Arguments for Which Winning is Losing. Occasionally a party argues in such a way that winning, ironically, means that she loses or should lose. Sometimes this is due to counsel's short-sightedness in constructing arguments, but sometimes it is due to a logically perverse set of circumstances from which counsel must try to escape. The best examples of the latter are Protagoras v. Euathlus and State v. Jones. Here are some examples of the former.

The Court of Appeals of New York once held, in a case of first impression, that infants had no cause of action when born maimed for injuries sustained while a fetus. Dobner v. Peters, 133 N.E. 567 (1921). Thirty years later the same court wanted to reverse itself. It argued that Dobner was no longer good law because the "chief basis" of its holding —the lack of precedent— no longer existed. Woods v. Lancet, 102 N.E.2d 691 (1951). The precedent that then existed was Dobner!

Many cases present roughly the following situation. When a parent dies, a child contests the validity of her parents' marriage in order to cut out the surviving spouse and increase her own share of the decedent's estate. The argument succeeds, the marriage is found void, but the child is thereby found illegitimate and (in some jurisdictions) ineligible to inherit. Beamish v. Beamish, 9 H.L.C. 274, 11 Eng. Rep. 735 (1861).[Note 37] This situation is more common when the dead parent is the father, since it is more common for states to bar illegitimate children from inheriting from their fathers (absent an acknowledgment of paternity or a posthumous paternity action) than from their mothers.

The film, The Amazing Dr. Clitterhouse, presents an argument for which losing, apparently, is winning. Edward G. Robinson plays a doctor who leads a criminal band in order to study their physiology before, during, and after criminal actions. He eventually kills his rival, Humphrey Bogart, in order to avoid exposure and to test his own reaction to the ultimate crime. When caught he resists his lawyer's attempt to establish an insanity defense. After his lawyer argued insanity anyway, he took the stand and boldly asserted his sanity. The jury acquitted, saying that only an insane man would claim sanity in a murder case.

The combat flyer, Orr, in Joseph Heller's Catch-22 lost by the same logic. Driven insane by combat, he was eligible for non-combat duty on the grounds of insanity. But he had to ask for it. Those who ask, however, encounter the legalistic "catch" that "anyone who wants to get out of combat duty isn't really crazy."[Note 38]

Notes

1. I proposed this broad definition, and used it to govern the selections, in my "A Bibliography of Works on Reflexivity," in S.J. Bartlett and P. Suber (eds.), Self-Reference: Reflections on Reflexivity, Martinus Nijhoff, 1987, pp. 259-362. [Resume]

2. The earliest version of this case is reported in Aulus Gellius, The Attic Nights of Aulus Gellius (c. 150 C.E.), trans. John C. Rolfe, 3 vols., Loeb Classical Library, Harvard University Press, rev. ed., 1946, vol. I, Book V, x, pp. 405-09. If the story is true, then this account was written roughly 600 after the events. (Nothing is known of Euathlus; Protagoras of Abdera lived from 480 to 410 B.C.E.) Diogenes Laertius repeats the story about a century later in his Lives of the Eminent Philosophers, trans. R.D. Hicks, Loeb Classical Library, G.P. Putnam's Sons, 1925, Book 9, Chapter 56, pp. 468-69.

Different accounts of the case change the facts in small ways. The most salient changes to the logic and law of the verdict are the following.

In Aulus Gellius, Euathlus paid half his fee before the lessons began, and made the contract to pay the other half. The contract stipulated that Euathlus would pay the other half "on the day when he first pleaded before jurors and won his case", which is much less ambiguous than "when Euathlus won his first case". After making the contract, Euathlus delayed in taking any cases and Protagoras sued from impatience. Euathlus knew he'd be better off hiring a lawer, but wanted the satisfaction of winning the suit and the argument. The case went to a jury, whose members postponed their verdict to "a distant day" fearing that any verdict would "annul itself".

In "Dilemmas and Circles in the Law," Archiv für Rechts- und Sozialphilosophie, 51 (1965) 319-335, Fiori Rinaldi says that Protagoras initiated the contract, not Euathlus. Protagoras saw that Euathlus needed training but could not pay, and "came to his rescue by offering to give the necessary tuition...." Protagoras proposed that the contract turn on Eualuthus' first victory, making it difficult for Protagoras to decry the hardness or ambiguity of the terms. Rinaldi suggests (at p. 324) that the original Greek judge held for Euathlus.

In Ilmar Tammelo's Outlines of Modern Legal Logic, Franz Steiner Verlag, 1969, p. 122, Euathlus "abstains from going into legal practice" after the completion of his lessons.

In "Protagoras versus Euathlus: Reflections on a So-Called Paradox," Ratio, 19 (Dec. 1977) 176-80, Wolfgang Lenzen says that Euathlus contracted to pay "if and only if" he wins his first legal case. Although Lenzen's source of the paradox is J.L. Mackie, Truth, Probability and Paradox, Oxford University Press, 1973, p. 296, Mackie uses "when", not "if and only if". Lenzen says that after making the contract, Euathlus "does not dream of taking part in any court-case".

Judge Putnam in State v. Jones (see Section 20.B) said that Protagoras stopped teaching Euathlus when he decided Euathlus was proficient. Euathlus protested that he was not finished; Protagoras then sued. The case did not go to a jury, but to a panel of judges who were so perplexed that they adjourned for 100 years. [Resume]

3. See Aulus Gellius, Rinaldi, and Lenzen from note 2 above. See also E. Schneider, Logik für Juristen, Berlin, 1965, at p. 152; Ilmar Tammelo, Outlines of Modern Legal Logic, Franz Steiner Verlag, 1969, at pp. 122-23; John Bryant, "The Paradox of the Lawyers," Midwest Journal of Philosophy, 4 (1976) 1-2; W.K. Goossens, "Euathlus and Protagoras," Logique et Analyse, 20 (1977) 67-75; Raymond Smullyan, What is the Name of This Book? Prentice-Hall, 1978, pp. 213-14; Lennart Aqvist, "The Protagoras Case: An Exercise in Elementary Logic for Lawyers," in Wlodzimierz Rabinowicz (ed.) Tankar Och Tankefel, Uppsala, 1981, pp. 211-24. [Resume]

4. Lenzen, op. cit., thinks Euathlus should win the third suit. Schneider, op. cit. thinks he should lose it. [Resume]

5. The only other wily scheme that occurs to me is a design to make vivid a proposition for which he was famous and, indeed, that brought him pupils and income. In Seneca's words,

Protagoras declares that one can take either side on any question and debate it with equal success —even on this very question, whether every subject can be debated from either point of view.

Seneca ad Lucilium Epistulae Morales, vol. II, trans. Richard M. Gummere, Loeb Classical Library, Harvard University Press, 1953, Epistle 88.43, p. 375. Since this proposition would only become vivid in the lawsuit if Euathlus had been sufficiently well-trained to pose the counter-dilemma to Protagoras' dilemma, the wily scheme might have included the demonstration that Protagoras is a good teacher, a claim on which he often boasted (see e.g. Plato, Protagoras, 328.a-b). If anyone thinks this scheme is a bit too wily to attribute even to Protagoras, then I will acquiesce. [Resume]

6. One might well have solved this problem by giving Jones a directed verdict, that is, taking the issue from the jury and ruling as a matter of law that that Harris was Jones' accomplice and hence that there is no legally sufficient evidence to sustain the charge that he performed Harris' abortion. Many judges, however, will wait to see what the jury says before directing a verdict, preferring that acquittal come from the jury than from the bench. To a lawyer, then, Jones may seem less a problem in logic or self-annulling verdicts, than a problem in the timeliness of granting a directed verdict. [Resume]

7. More precisely, it is being found guilty that implies that Jones should be found innocent, although this does not affect the paradox. However, in going the other direction, being found innocent does not strictly imply that Jones should be found guilty, only that he may be found guilty. Being found innocent means that Harris's testimony could suffice if the jurors found it credible, not that it must suffice. Hence, one might say that only Jones' conviction leads to self-annulment, while his acquittal is self-consistent. To logicians that would usually suffice to acquit Jones. The only problem with this solution in law is that (under 1946 Ohio law) Jones is guilty! [Resume]

8. However, when abortion was illegal and women who voluntarily procured abortions were considered accomplices of the abortionist, many accused abortionists found themselves in Jones' position and wanted jury instructions to the effect that the woman who requested the abortion was an accomplice as a matter of law. These requests were usually denied, as they were in Jones. The jury is to decide who is an accomplice, and the burden is on the defendant to show that the woman was an accomplice. See Wigmore On Evidence, footnote 12 to §2060 (vol. 7, p. 444) for cases. [Resume]

9. C.I. Lewis said all dictionaries are circular in this sense, by defining the words with other words that are defined in the same work. The series ultimately closes in a circle, and one criterion of merit for dictionaries is that the circle be as large as possible. Clarence Irving Lewis, Mind and the World Order: Outline of a Theory of Knowledge, Charles Scribner's Sons, 1929, p. 82. [Resume]

10. J.C. Hicks, "The Liar Paradox in Legal Reasoning," Cambridge Law Journal, 29, 2 (November 1971) 275-291. This article is more on logical paradoxes in law than on paradoxes of lying in law. [Resume]

11. Alf Ross, "On Self-Reference and a Puzzle of Constitutional Law," Mind, 78 (1969) 1-24 at p. 19.n. [Resume]

12. See the advice for witch-prosecutors in Heinrich Kramer and James Sprenger, Malleus Maleficarum (original 1486), Dover Publications, 1971, Third Part. [Resume]

13. Richard Popkin, History of Skepticism from Erasmus to Spinoza, University of California Press, 1979, at pp. 181, 287.n.38. [Resume]

14. "The FBI's Show of Shows," Time Magazine, 116, 9 (September 1, 1980) 14ff, at p. 19. [Resume]

15. In chronological order the main articles are: Anon., "The Three-Cornered Priorities Puzzle," Virginia Law Review, 8 (1922) 550; White, "A Problem in Priorities," Ohio Law Reporter, 25 (1926) 116; Anon., "Priorities Between Mortgages and Mechanics Liens," Yale Law Journal, 36 (1926) 129; Orlo Kellogg, "Priorities Puzzle Under the Ship Mortgage Act," Washington Law Review, 2 (1927) 117-21; Anon., [Note on circular liens], Virginia Law Review, 15 (1928) 90; A. Kocourek, "A First Rate Legal Puzzle: A Problem in Priorities," Illinois Law Review, 29 (1935) 852; Benson, "Circuity of Lien —A Problem in Priorities," Minnesota Law Review, 19 (1935) 139; Anon., "Circuity of Liens —A Proposed Solution," Columbia Law Review, 38 (1938) 1267-78; Anon., [Note on circular liens], Harvard Law Review, 67 (1953) 358; Anon., "Circuity of Priority and Liens under Section 67(c)(1) of the Bankruptcy Act," Yale Law Journal, 66 (1957) 784; Richard Peck, "Federal Tax Liens —Their Removal or Foreclosure, Priority Thereof, and the Problem of Circuity of Priorities," Nebraska Law Review, 38 (1959) 163-72; Grant Gilmore, "Circular Priority Systems," Yale Law Journal, 71 (1961) 53-74; R.H. Stern, "Priority Paradoxes in Patent Law," Vanderbilt Law Review, 16 (1962) 131; Alan L. Tyree, "Circular Priorities in Secured Transactions," American Mathematical Monthly, 87 (1980) 186-93. The best review of the literature and summary of approaches for dealing with circular priorities is Grant Gilmore's 1961 article. [Resume]

16. Technically, the jurisdiction of a court to determine its own jurisdiction is subject to res judicata and direct review, but not collateral attack. See Anon., "Res Judicata and Jurisdiction: The Bootstrap Doctrine," Harvard Law Review, 53 (1940) 652-660. See Peri v. Groves, 50 NYS 2d. 300, 308 (1944); Baldwin v. Iowa State Traveling Men's Association, 283 U.S. 522; 118 ALR 1518; Charles S. Collier, "Judicial Bootstraps and the General Welfare Clause," George Washington Law Review, 4 (1936) 211-242. [Resume]

17. The infinite regress is as follows. If a defendant challenged the jurisdiction of court #1 over her, and court #1 could not itself decide the issue of jurisdiction, then the question would have to go to an appeals court #2. If the defendant challenges the jurisdiction of court #2 over her, then the question passes to court #3, and so on. To avoid this it is necessary that some court in the series have the jurisdiction to decide its own jurisdiction; in our system this power is given to virtually every court. [Resume]

18. W.J. Rees, "The Theory of Sovereignty Restated," Mind, 59 (October 1950) 495-521, at p. 518. [Resume]

19. John Stuart Mill, On Liberty (Original 1859), Hackett Publishing Co., 1978, pp. 101f. [Resume]

20. See Section 18, note 8, above, and Appendix 2, below. [Resume]

21. For more discussion of renvoi from a logical point of view, see Fiori Rinaldi, "Dilemmas and Circles in the Law," Archiv für Rechts- und Sozialphilosophie, 51 (1965) 319-335 at 330f; J.C. Hicks, "The Liar Paradox in Legal Reasoning," Cambridge Law Journal, 29, 2 (November 1971) 275-291, at 284-286 and citations at his footnote 47; and Laurence Goldstein, "Four Alleged Paradoxes in Legal Reasoning," Cambridge Law Journal, 38 (1979) 373-91, at 380f. [Resume]

22. See Rinaldi, op. cit. at 321 and 321.n.8. [Resume]

23. Rinaldi, op. cit. at 333. [Resume]

24. Goldstein, op. cit. at 380-81, emphases in original. [Resume]

25. Ross, op. cit. at pp. 21ff. [Resume]

26. There is a very large literature on this problem. See e.g. W.N. Harrison, "The Statute of Westminster and Dominion Sovereignty (II)," Australian Law Journal, 17 (1944) 314; Geoffrey Marshall, Parliamentary Sovereignty and the Commonwealth, Oxford University Press, 1957, Chapter VI; Ilmar Tammelo, "The Antinomy of Parliamentary Sovereignty," Archiv für Rechts- und Sozialphilosophie, 44 (1958) 495-516, at pp. 510-12; Geoffrey Marshall, Constitutional Theory, Oxford University Press, 1971, at pp. 61ff. [Resume]

27. J.M. Finnis, "Revolutions and the Continuity of Law," in A.W.B. Simpson (ed.), Oxford Essays in Jurisprudence, Second Series, Oxford University Press, 1973, pp. 44-76, at 52. [Resume]

28. A contract is integrated when it is the complete agreement of the parties and is not, for example, subject to oral qualifications not written into the contract. [Resume]

29. The medieval literature on this and similar paradoxes is well covered in Paul Vincent Spade, The Mediaeval Liar: A Catalogue of the Insolubilia-Literature, Pontifical Institute of Mediaeval Studies, 1975; and E.J. Ashworth, "Will Socrates Cross the Bridge? A Problem in Medieval Logic," Franciscan Studies, 36 (1976) 75-84. [Resume]

30. Part 2, Chapter 51. This paradox is discussed in Max Black, Language and Philosophy, Cornell University Press, 1949, pp. 91ff. [Resume]

31. See George P. Fletcher, "Paradoxes in Legal Thought," Columbia Law Review, 85 (1985) 1263-1292 at p. 1269, and his cited sources. [Resume]

32. For further discussion of these issues, see my "Logical Rudeness," in S.J. Bartlett and P. Suber (eds.), Self-Reference: Reflections on Reflexivity, Martinus Nijhoff, 1987, pp. 41-67, esp. pp. 62f. [Resume]

33. Some witnesses may be caught in a further reflexive dilemma, although it is not related to knowing before we know. If the judge asks contextual questions to determine whether the witness is justified in invoking the privilege against self-incrimination, the witness is bound to answer these contextual questions under oath. It may be that one kind of self-incrimination she wishes to avoid is the exposure of earlier perjury. Truthful answers to the judge's contextual questions may expose her perjury by contradicting her earlier testimony. Hence, to the witness, the privilege would have to be invoked again at the meta-level, against the judge's contextual questions. Courts facing this situation compel answers to the contextual questions, as if there were no privilege, but protect the answers from future criminal prosecution. In re Boiardo, 34 N.J. 599, 170 A.2d 816 (1961). [Resume]

34. The same logic arises in Fourth Amendment (search) cases. When police search a car on the basis of the driver's consent, as opposed to probable cause or a search warrant, then the driver can terminate the search at any time by revoking her consent. The state of Maine once argued that the revocation of consent in the middle of a search, just as the police approach the glove compartment or a parcel, is itself a suspicious act creating probable cause to continue the search without the driver's consent. This argument was rejected in State v. Walker, 341 A.2d 700, 704 (1975). Preserving the legally "formal" understanding of consent against psychologically more "realistic" interpretations of it is analogous to preserving the formal against the realistic understanding of the invocation of the Fifth Amendment privilege. [Resume]

35. For other cases in which the admissibility of evidence turned on the factum probandum or fact to be proved, see Stowe v. Querner, L.R. Exch. 155 (1870), and State v. Lee, 127 La.1077, 54 So. 356 (1911). [Resume]

36. For an analysis of tort liability so that it oscillates and pyramids like tax liability, see George P. Fletcher, op. cit. at pp. 1285ff. [Resume]

37. See Richard Wasserstrom discusses the Beamish case in his The Judicial Decision: Toward a Theory of Legal Justification, Stanford University Press, 1961, at pp. 58-60. [Resume]

38. Joseph Heller, Catch-22, Dell Publishing Co., 1962, at p. 47. [Resume]

This file is one section of the book, The Paradox of Self-Amendment. Return to the Table of Contents.

Peter Suber, Department of Philosophy, Earlham College, Richmond, Indiana, 47374, U.S.A.
peters@earlham.edu. Copyright © 1990, Peter Suber.