This essay originally appeared in Christopher B. Gray (ed.), Philosophy of Law:  An Encyclopedia, Garland Pub. Co, 1999, II.790-792.   Copyright © 1999, Peter Suber.

Self-Reference in Law
Peter Suber, Philosophy Department, Earlham College

We know from more than two millenia of experience that self-referential statements, such as the liar's ("This very statement is false"), can be debated by philosophers and logicians for millenia without producing consensus on their solutions. We should not be surprised, then, if self-referential laws produce paradoxes which puzzle lawyers. What is surprising, though, is that some of these paradoxes bother only the logicians and philosophers who study law from outside, and do not bother lawyers at all. This fact should interest philosophers of law even more than the paradoxes themselves.

Alf Ross (1969) argued that a constitutional amending clause could not be used to amend itself; the act could be reduced to a formal self-contradiction. He did not know, or did not acknowledge in his essay, that self-amendment is commonplace in legal history (Suber 1990). If it is contradictory, that fact had never been noticed before and had never bothered either legal officials or citizens. What if Ross is right that self-amendment is contradictory? We may want to conclude that the legal practice of self-amendment is invalid, even if it has been accepted by citizens and courts wherever it has occurred, just as inconsistent theories are false even if widely accepted. Or we may want to conclude that in some circumstances law can harbor contradiction, just as essays and novels may harbor contradiction. Ignoring some qualifications, these are the positions of Ross and Suber. There may be other explanations better than these, but these are enough to show that self-reference in law is not a minor curiosity; in this case it raises profound questions about the boundaries of legal change and the nature of legal rationality.

While self-amendment has occurred in almost every state which has an amending clause, the act is rarely as important for practice as it is for theory. It can be seen as a variation on the theme of the paradox of omnipotence: can a deity create a stone so heavy she cannot lift it? If she can, there is a stone she cannot lift; and if she cannot, then there is a stone she cannot create. Either way, she seems to lack classical omnipotence. An amending clause seems to be legally omnipotent, because it can modify any law in its system, perhaps including itself. But we can ask of it what we asked of the deity: can it limit itself irrevocably? If it can, then there is a limit it cannot overcome; and if it cannot, then there is a limit it cannot enact. But if the amending power is not legally omnipotent, then no law-making power is legally omnipotent. Are we to conclude, then, that in a democracy the people cannot make any law at any time? Here is where the theoretical question can become quite practical.

In the contract tradition, can the founding generation give consent to a constitution which binds its successors forever? Or must we seek consent from each generation? If the omnipotence of one generation allows it to bind its successors irrevocably, then its consent can establish a constitution over the dissent of its descendants. But if the equal omnipotence of the succeeding generations means that they can overrule any decision made by their ancestors, then the consent of each generation will be needed. If legitimate government derives from the consent of the governed, then we must wade into the paradox of legal omnipotence in order to decide whose consent matters, or to decide legitimacy. Suber argues (1990) that a self-applicable amending clause can repeal any previous law, which makes all limits on the people's power revocable. It follows that the only irrevocable limit on the people's power is that they cannot enact other irrevocable limits on their power. This is consonant with the rule in England and the United States that one legislature cannot bind its successors irrevocably.

It also follows that John Austin's theory of sovereignty is false in holding that every sovereign is both unlimited and illimitable. If the supreme legal power (usually the amending power) is unlimited, then it can limit itself irrevocably; and if it is illimitable, then it cannot limit itself irrevocably and so is already limited. No sovereign can be both unlimited and illimitable, not even the people.

Legal self-reference can be desirable as the alternative to infinite regress. If all valid law must be validated by prior or higher law, as many formalists hold, then the existence of valid law commits us to an infinite regress of prior, or higher, laws. To avoid this absurdity, we seem forced to acknowledge that some powers can create law ex proprio vigore, or from their own strength. Historical candidates for such sources of law are custom, contract (or consent), revolution, conquest, and natural law as promulgated by human reason. Another example in which self-reference helps us prevent infinite regress is the so-called bootstrap doctrine, or a court's jurisdiction to determine the scope of its own jurisdiction. If a court lacked this self-applicable jurisdiction, then challenges to its jurisdiction could only be answered by another court, and a determined advocate could push back the regress indefinitely.

Some legal circles are problems, not solutions. If A and B make a contract in Illinois, and B violates it in Indiana, A will consult the contract to see whether Illinois or Indiana law applies to the breach. Suppose the contract specifies Illinois law. Illinois law may in turn require the parties to use the law of the state where the breach occurred. But Indiana law may require them to use the law of the state where the contract was made. Such a circle is called renvoi. Similarly, the accidental side-effect of many rules made over many years may be that an estate must be divided among claimants when A has priority over B, B over C, and C over A. If laws were simply rules, like software rules, then these situations would precipitate the legal equivalent of infinite loops. But because renvoi and circular lien problems are solved in finite time, in principled ways, lawfully, by human decisions which face the need to escape absurd literalism, they provide important clues to the sense in which laws are not simply rules, and law itself more a human enterprise than a formal system.

There are other cases in which legal powers or institutions act on themselves in a way which raises the specter of paradox. Can Article VII of the U.S constitution establish the conditions of its own establishment? Can the English House of Lords abolish itself? Can a judicial decision (London Street Tramways, 1898) declare, with nothing but the force of precedent, that precedents ought to be followed? Can such a decision later be overturned (Hansard Report, 1966)? Can a court constituted under the laws of a post-revolutionary regime decide the legality of the revolution, the regime, and itself? Can a court declare a constitutional amendment unconstitutional (U.S. v. Sprague, 1930)? Can a will forbid anyone to challenge its validity, on pain of being excluded from the estate (in terroram clauses)? Can a written contract declare that the parties have no oral amendments or qualifications to the written terms (integration clauses)? Can a treaty bind a nation to ratify it (Article III of the Kellogg-Briand Pact of 1928)? Can a "sunset clause" in a statute trigger the invalidation of the statute at a certain time, including the sunset clause? Can the effective date provision of a statute authoritatively declare, before the effective date, that the statute is not yet authoritative? Can a doctrine of desuetude (invalidation through non-use or obsolescence) become invalid over time through desuetude?

Fletcher, George. "Paradoxes of Legal Thought." Columbia Law Review, 85 (1985), 1263-1292.

Gilmore, Grant. "Circular Priority Systems." Yale Law Journal, 71 (1961), 53-74.

Goldstein, Laurence. "Four Alleged Paradoxes in Legal Reasoning." Cambridge Law Journal, 38 (1979), 373-391.

Hart, H.L.A. "Self-Referring Laws." In his Essays in Jurisprudence and Philosophy. Oxford: Oxford University Press, 1983, pp. 145-158.

Hicks, J.C. "The Liar Paradox in Legal Reasoning." Cambridge Law Journal, 29 (1971), 275-91.

Rinaldi, Fiori. "Dilemmas and Circles in the Law." Archiv für Rechts- und Sozialphilosophie, 51 (1965) 319-335.

Ross, Alf. "On Self-Reference and a Puzzle in Constitutional Law." Mind, 78 (1969) 1-24.

Suber, Peter. The Paradox of Self-Amendment: A Study of Logic, Law, Omnipotence, and Change. New York: Peter Lang Publishing, 1990. (Contains a bibliography on self-reference in law.)

Tammelo, Ilmar. "The Antinomy of Parliamentary Sovereignty." Archiv für Rechts- und Sozialphilosophie, 44 (1958) 495-513.

Teubner, Gunther. Reflexive Law: Autonomy and Regulation. Berlin: Walter de Gruyter, 1986.

[Blue Ribbon] Peter Suber, Department of Philosophy, Earlham College, Richmond, Indiana, 47374, U.S.A.
peters@earlham.edu. Copyright © 1999, Peter Suber.