Section 8
Omnipotence and Immutability Peter Suber, Paradox of Self-Amendment Table of Contents

• A. Omnipotence and immutability are inseparable concepts
• B. Entrenchment, self-entrenchment, disentrenchment
• C. Supposed limitations on the U.S. federal amending power
• Notes

A. Omnipotence and immutability are inseparable concepts

In Section 3 I outlined the dilemma of paradoxical self-amendment versus an absurd infinite regress of rules of change. The possibility that a supreme rule of change might be immutable was eliminated from the dilemma as empirically false and question begging. The ACs of 47 state constitutions have actually been amended (Appendix 2), showing their mutability and self-mutability. To suppose that any rules are legally immutable begs the question of the mutability of the rules of change: for a changeable rule of change might in some attainable new form reach all the rules of the system, even if its original form could not. Nevertheless the dilemma is at least a trilemma, and the possibility of immutable rules of change is a distinct third option. Now that the development of the dilemma has proceeded without the question begging assumption of immutability, the latter may be examined on its merits.

By "immutability" I will here mean only legal immutability. If a legal rule cannot be changed by any legal means whatsoever, then it is legally immutable. Whether it makes sense to say that a rule is legally immutable at a particular time and mutable at other times, will be explored; it will not be ruled out as self-contradictory. No legal rule is absolutely immutable, of course, for illegal and extra-legal means such as revolution are always available.[Note 1]

A legally immutable rule of any kind is a prima facie proof that the supreme rule of change is not omnipotent. For if the supreme rule of change were omnipotent, it could amend the rule that purports to be immutable. Conceivably a rule that is immutable at one moment may be "transmuted" in the next moment, if the AC can change itself into a form that can transmute immutable rules or convert them to mutable rules (see Appendix 3). This will be explored.

An omnipotent AC is one that can authorize any change of law at the constitutional level. Hart's distinction between self-embracing and continuing omnipotence[Note 2] is a distinction with a difference and should be followed (see Section 5.B). A self-embracing omnipotent AC can authorize any change of law, including changes in itself. A continuing omnipotent AC must continue to have continuing omnipotence; therefore it can authorize any change of law except a change that would limit its omnipotence or the continuity of its omnipotence.[Note 3] Therefore continuing omnipotence is actually limited in its power because it cannot limit or repeal itelf; if this is understood, it may still be called continuing "omnipotence".

The argument that the distinction between self-embracing and continuing omnipotence solves or dissolves the paradox of self-amendment is considered and rejected in Section 11.

Immutability and omnipotence are complementary concepts and must be discussed together. The continuing character of continuing omnipotence is immutable. Moreover, continuing omnipotence is incapable of making immutable rules, for an immutable rule is an irrevocable limit to its omnipotence. Self-embracing omnipotence, by contrast, may enact any immutable rule. Such rules will irrevocably limit its power, but that kind of self-limitation is possible for self-embracing omnipotence.

Initially, then, continuing omnipotence is limited by one immutable rule (that it be continuing omnipotence) and can enact none, while self-embracing omnipotence is limited by no immutable rules and can enact any. The existence of any immutable rule other than the continuity of continuing omnipotence is incompatible with both types of omnipotence; but in the case of self-embracing omnipotence, such rules may be made anyway, forsaking omnipotence in the act. Continuing omnipotence may be called immutable omnipotence, without loss of meaning, and self-embracing omnipotence may be called mutable omnipotence.

One conclusion we may draw immediately is that if the AC must have either continuing or self-embracing omnipotence, then the immutability of at least one legal rule is either always possible or always actual. If the supreme rule of change or AC has continuing omnipotence, then its continuing character is actually immutable. If the supreme AC has self-embracing omnipotence, then no rule may initially or presently be immutable, but immutable rules may be enacted or irrepealably entrenched. And of course if no rule of change is omnipotent in either sense, then some actual or possible rules are therefore immutable, namely, those that limit the amending power. Omnipotence either starts qualified by one immutable rule, or is capable of qualifying itself by immutable rules —or omnipotence itself is precluded due to the presence of other immutable rules.

(In Section 21.C I argue for a theory in which no rule is possibly or actually immutable. Hence, it denies that the AC either has continuing or self-embracing omnipotence.)

This analysis applies equally to legal sovereigns, metaphysical deities, and hypothetical powers.[Note 4] Instead of seeing the problem only as a conflict between an amending power and an arguably immutable rule that resists amendment, or between an amending power and its power over itself, it applies as well to the cognate problems of an irresistible force (omnipotent amending power) and an immovable object (immutable rule) and of an omnipotent deity in an immutable cosmic order that resists its power to put mercy before justice, square the circle, become mortal, or limit its own power irrevocably. But this is not the whole story: the intriguing possibility that "immutable" rules may finally be amended if self-amendment is permissible will be explored later in this section.

If an AC can amend any rule of the constitution, and if it can amend itself into any other form, then it has self-embracing omnipotence. If it can amend the rules of the constitution and itself only in ways that do not diminish or augment its power (see Section 5.B), then it has continuing omnipotence. If it cannot amend itself even in these ways, or if it cannot amend certain provisions of the constitution, then it is not omnipotent at all, unless it can make itself omnipotent, and at least one rule is immutable, unless the AC can transmute it.

B. Entrenchment, self-entrenchment, disentrenchment

Limitations on the amending power may be of two basic types. The first is typified by a clause in the constitution that says of clauses other than itself that they cannot be amended except perhaps in special ways. The federal AC "entrenches" the equal suffrage of the states in the Senate in just this way. The last clause of Article V, the AC of the federal constitution, holds that "no State, without its consent, shall be deprived of its equal suffrage in the Senate." This language seems to require a special amendment procedure if any state is to be disenfranchised from the senate, namely, the regular procedure in which the victim happens to vote aye.

The second type of limitation is typified by a clause that says of itself that it is immune to amendment. There is no explicit example of this type in the federal constitution, although there are several among the states and in the English tradition which will be examined (some in Section 9). Let us call the former type "entrenchment clauses", and the latter "self-entrenchment clauses".[Note 5]

There may be cases of implied self-entrenchment. Because an entrenchment clause is less useful for its intended purpose if it is not self-entrenched, there is a temptation to read every entrenchment clause as impliedly self-entrenched. For example, in the case of the entrenchment of the equal suffage of the states in the federal amending process, we seem to have a case of entrenchment without express self-entrenchment. The clause of Article V prohibits amendments of a certain kind. But since no clause protects that clause the way it protects the equal suffrage of the states, may it be repealed by ordinary amendment procedures as the first step toward making the prohibited sort of amendment? If so, it may be repealed without the consent of the states eventually to be disenfranchised from the senate. The language of the clause does not help us much, and good policy grounds could be cited on both sides. If it is read to prohibit its own repeal, then it is impliedly self-entrenched. (More on implied self-entrenchment in Section 8.C below.)

We can interpret entrenchment and self-entrenchment clauses equally well as procedural or substantive limitations on amendment. A procedural limitation bars amendments that do not conform to a certain procedure, while a substantive limitation bars amendments with a certain content. The clause impeding the disenfranchisement of a state from the senate is clearly both. The objectionable content is named in the clause, and a special procedure is laid down for amendments of that kind. Entrenchment and self-entrenchment clauses are often thought to be procedural limitations only; in England they are called "manner and form" limitations as if to highlight their indifference to content.[Note 6] But this overlooks the fact that the clauses always protect certain provisions of law from (ordinary) amendment. The protection they afford consists in making the amendment procedure more difficult, but the object of protection is invariably some specific content or substantive kind of law that the clause-writers did not wish to see enacted except with extraordinarily wide consent.

The very idea of procedural limitations on a legal power deserves a word of clarification. Procedures may be ordinary or extraordinary. For example, amendments to the federal constitution must be ratified by two-thirds of each house of Congress and by three-fourths of the states. These procedural requirements are ordinary; they are the "default" conditions to which any amendment must conform unless some special procedure is required. If a particular state must vote to ratify an amendment, because its consent is required for its own disenfranchisement, then that procedural requirement is extraordinary.

It sounds very odd to say that ordinary requirements are "procedural limitations" on the amending power because they are just as much the constitutive or defining conditions of that power. They do not seem to "limit" the power any more than DNA "limits" the nature of human beings. But a moment's reflection shows that speaking in this way is harmless and even helpful in many contexts. If it were physically impossible to transgress the limitation, then the oddity would remain. But in law it is possible, frequent, and even perilously easy for an intended act of amendment, legislation, adjudication, or divorce to violate the ordinary procedural limitations on acts of that kind and consequently to be a legal nullity. It often takes lawyers who specialize in that area of law to ensure that the ordinary procedure is followed in every particular. Nevertheless, extraordinary requirements feel more like limitations than ordinary requirements, for they restrict the valid application of a power that is already defined more broadly.

Self-entrenchment, like self-amendment, may depend on the stipulated scope of "self". If an AC has many sections and one says of another that it shall not be amended, then that is self-entrenchment broadly conceived. Narrow or strict self-entrenchment, however, must be restricted to the entrenchment of a unit of language by words within that unit.

Entrenchment affects the problem of self-amendment in many ways. When a constitutional rule is entrenched, we must ask whether the AC can repeal the entrenching language and amend the previously protected rule. This aspect of the omnipotence of the AC is the subject of this section. The entrenchment and self-entrenchment of the AC itself, or of parts of the AC, is the subject of the next, Section 9. This is a good place to warn the reader that the topics of Sections 8 and 9, and especially their detailed discussion, can cause dizziness in proportion to one's concentrated attention.

Entrenchment bars amendment, and entrenchment of the AC bars self-amendment. But to the extent that an AC can repeal entrenching language it can reassert its power over the protected provision of law. When it is itself the protected provision we may well wonder whether it can repeal the entrenching language in order to amend itself, an act I will call "self-disentrenchment". It appears to present the challenge of a prisoner who can only escape by opening the outer door of the prison before the inner cell door.

We must also distinguish complete and incomplete entrenchment. Complete entrenchment protects a passage from amendment per se, while incomplete entrenchment only protects it from ordinary or routine amendment. Incomplete entrenchment may suspend the usual amendment procedures for the entrenched passage and require a special, more difficult procedure.

Whether complete entrenchment protects rules from amendment immutably should not be prejudged from the word "complete". If an entrenchment clause said, "rule R can be amended only by convention" or "only by 90% supermajorities in each house and the concurrence of 80% of the states", then it would describe incomplete entrenchment. That, obviously, would do little good unless it entrenched itself to the same or greater degree of difficulty. For this reason entrenchment clauses are often thought to imply self-entrenchment. Hence in corporate charters, much more often than in constitutions, the articles requiring supermajority votes to amend certain basic rules are usually and prudently self-entrenched incompletely at the same supermajorities.

Complete entrenchment, by contrast, would be indicated by a simpler clause, such as, "rule R may not be amended" or, in the case of complete self-entrenchment, "this section may not be amended". The significance of the distinction is shown by Hart who argued (not in these terms) that in a nation whose legislature or parliament has continuing omnipotence, one parliament cannot make laws that its successors cannot repeal, but it could make laws that its successors could repeal only with special, difficult procedures.[Note 7] In the terms we are using here, continuing omnipotence is compatible with incomplete entrenchment and self-entrenchment. It is compatible with complete entrenchment and self-entrenchment only if the completely entrenching language is itself subject to amendment. Complete entrenchment clauses may be mutable if they are not themselves entrenched or if they are only incompletely entrenched. If they are completely self-entrenched, then they seem to be immutable; they may be mutable only if the AC is omnipotent in such a way that it can transmute immutable rules.

Hart argues that the question whether a parliament in a nation of parliamentary supremacy has continuing or self-embracing omnipotence is an empirical question.[Note 8] In England the rule that parliament cannot pass laws that its successors cannot repeal is a venerable legal axiom.[Note 9] It implies continuing power and, if parliament is otherwise omnipotent, then continuing omnipotence.[Note 10] That the federal and state legislatures in the United States are just as firmly barred from making laws that their successors cannot repeal is not as well known or commonplace, but it is the repeated holding of American courts facing the question.[Note 11] This implies continuing power, but because our legislatures are not supreme, it does not imply continuing omnipotence. Similarly, neither the English parliament nor an American legislature can irrevocably limit its own power or the power of its successors to make or unmake law.[Note 12] There are no comparable holdings, at the federal or state levels, that an AC cannot bind its successors. Hence the AC, which is supreme, might be omnipotent without being continually omnipotent.

These empirical tests are decisive when available. An AC is known to be self-embracing if it has actually been used to amend itself or to limit its future power, without interference from courts; it is not, however, thereby known to be omnipotent. An AC that has never amended itself and that has never been prevented from amending itself by a court may have the potential for either self-embracing or continuing power, depending on future history. It may determinately be one or the other, by the intent of the framers, dicta of past courts, or some other test, and become the other as future courts invoke new criteria.

Alf Ross is not content with these empirical tests, however. If an AC has been used to amend itself, and if courts have not struck down the attempt or the product, then Ross still wants to be able to say that the AC cannot have self-embracing power. Self-application of all kinds is logically objectionable, he says, legal history notwithstanding (see Section 5.D). If an AC is omnipotent, we know a priori that it must have continuing rather than self-embracing power, Ross says.

This position, in a word, is preposterous. It applies only to abstract, logical rules whose validity can be a matter of a priori reasoning. Legal rules are legally valid or invalid for legal reasons alone. Legal reasons often include logical reasons, but need not, and in all cases include fewer than all logical reasons and more than merely logical reasons. Ross offers no legal reason whatsoever why an AC could not be self-embracing or self-applicable, but concludes that the illogical is illegal. Any jurisprudence that declares that what is legal by legal tests may be illegal by logical tests is to that extent false and deficient. We might say that just as the laws of the United States make no exception for the conscience of Henry David Thoreau,[Note 13] they make no exception for the principle of non-contradiction or the logical scruples of Alf Ross. Ross's argument can only legitimately conclude that courts that have failed to prevent or nullify acts of self-amendment have erred. He cannot conclude that the correction to that error is already law. Now he is in the absurd position of saying that what the courts uphold as law is not law. For the same reason, as will become important below, logic, politics, and morality may suggest many "implied limitations" on the amending power, and these might become law if courts are persuaded to heed them. But if courts have not yet heeded them, then these limitations cannot be said to exist in law, even if their omission is illogical, immoral, or tragic.

Even the rich history of self-amendment among the states (Appendix 2) is inconclusive on the question of this section. It may show that self-amendment is legally permissible in the United States, but it does not show whether an AC can limit itself irrevocably through self-amendment, and hence whether it can be omnipotent in either of the two senses we have discussed. No attempted act of self-amendment, so far as I know, has ever been enjoined or nullified by an American court on grounds even remotely related to the logic of self-application.[Note 14] Therefore our history of permitted self-amendment cannot reveal whether some conceivable self-amendments would be impermissible, for example, for violating an irrevocable self-entrenchment clause, contradicting the continuity of continuing omnipotence, or transcending the power of a non-omnipotent amending power.

This raises a distinction that applies to all entrenchment clauses. They may be inserted by the framers of the original constitution or they may be inserted by amendment. Let us call the former "original" limitations on the AC, and the latter "self-imposed". The latter are self-imposed even if the AC is not directly amended by them, for the provision that is entrenched against future amendment stands as a limit on the power of the AC, put there by the AC.[Note 15] The distinction is important because original and self-imposed entrenchment clauses may differ in their mutability.

I will assume that all limitations on the AC that could make it less than omnipotent take the form of entrenchment and self-entrenchment clauses, some possibly tacit and implied. An AC could be described in language that clearly suggested limited power, but any such description could be interpreted as immunizing some provision (possibly indeterminate) from amendment, thereby acting as an entrenchment of that provision (to be determined). If no clause of the constitution is immunized or entrenched, by words or tacit rules qualifying the protected clause or the AC, then the AC is clearly unlimited. Whether it has self-embracing or continuing omnipotence, or some strange mix or third type, is a question for courts or philosophers looking at the history of amendment.

Even though all limitations on an AC may be construed as entrenchment clauses, there is still a useful distinction between limitations that make certain amendments inconsistent with protective language, and limitations that make certain amendments void. Amendments of the former kind are sometimes upheld by courts in deference to the people's will in ratifying them, while the latter seem impermissible or void ab initio. The former may overcome the protective language they violate by virtue of the rule of priority favoring the most recent voice of the sovereign in cases of irreconcilable conflicts. I will call these two types limitation through protection and limitation through incompetency.

Limitation through protection prohibits amendments that contradict valid rules of protection; limitation through incompetency prohibits those that are ultra vires, that is, unauthorized or beyond the scope of authority. The two are not always easy to distinguish, as when one argues that the Bill of Rights "cannot" be repealed. The advocates of that idea might mean that a tacit entrenchment clause protects the Bill of Rights, or that the amending power is incompetent to repeal certain basic values of the community or "natural rights".

One might argue that original limitations on the AC are more likely to be limitations through incompetency, and that self-imposed limitations —especially complete self-entrenchment clauses— are more likely to be limitations through protection. This theory might be based on the proposition that an AC cannot limit its own competency, but can protect a rule from amendment, especially from routine amendment. Of course both types would constitute the self-limitation of the competency of the AC.

I will also assume that an entrenchment clause that is not self-entrenched, like any other clause that is not entrenched, may be amended or repealed in the ordinary way. This must be true, regardless whether the entrenchment clause is original or self-imposed. The only questionable case is entrenchment of the AC itself (for which see Section 9), or the implied self-entrenchment of an entrenchment clause (which will be examined below). If Section A entrenches Section B without entrenching itself, then the framers of A may well have intended that it be as difficult to amend as B. But I will treat the problem of self-entrenchment independently, and therefore need not complicate matters by finding it tacit in an indefinite number of cases of mere entrenchment.

Hence, entrenchment clauses are not immutable and do not limit the power of the AC, except possibly when they entrench the AC itself (Section 9). Our present inquiry, then, reduces to the question whether self-entrenchment clauses can be immutable, or whether they can be amended or repealed. Implied limitations on the amending power that have been suggested by writers and plaintiffs will also be examined as possible limitations on the amending power.

First, notice that immutable self-entrenchment clauses are compatible with both self-embracing and continuing omnipotence in small and different ways. Self-embracing omnipotence may limit itself immutably, and therefore may be the source of an immutable self-entrenchment clause. After the self-entrenched amendment is adopted, of course, the self-embracing omnipotence has forsaken its omnipotence. Continuing omnipotence is already limited immutably; this limitation may take the form of a self-entrenchment clause that describes its immutable continuing character. Such a clause could be original or self-imposed by an AC that formerly had self-embracing omnipotence. After the amendment is adopted which immutably gives an AC continuing omnipotence, self-embracing omnipotence may be at an end, but continuing omnipotence is created and protected.[Note 16] In short, an immutable self-entrenchment clause may be the product and termination of self-embracing omnipotence, and the beginning and safeguard of continuing omnipotence. But an immutable self-entrenchment clause that does not entrench the AC excludes both types of omnipotence from the moment it becomes immutable.

Incomplete self-entrenchment is the easiest to cope with. If Section A contains a clause that says, "nothing in Section A, including this provision, may be amended except by a constitutional convention," then the clause can obviously be repealed by convention. One possible difficulty arises when the self-entrenchment clause names a method of amendment not included in the AC. In such a case a stubborn proponent of the inference model might deny that the constitution contains the authority to repeal the clause. But anyone else, not necessarily an acceptance theorist, could use the method outlined in the clause itself as permitted for the use of repealing the clause, either by appeal to the framers's intent, or the reasonably plain meaning of the clause —and of course despite any paradox of self-amendment.

Another possible difficulty arises when the incomplete self-entrenchment clause says it can only be repealed by a special method, and the amending body attempts to amend it by another method, perhaps the usual method. I do not believe this has ever happened in the United States, but it happened in New South Wales, Australia. The N.S.W. constitution created a body called the Legislative Council and provided that it could not be abolished by a simple amending act of parliament, but only by an amendment that was ratified in a popular referendum. Moreover, the requirement of a referendum could only be repealed by a referendum. Hence, the existence of the Legislative Council was incompletely self-entrenched. The N.S.W. parliament, however, wanted to abolish the Legislative Council without resort to a referendum, and believed that its sovereignty and omnipotence allowed it to do so through an ordinary amendment. The highest court in N.S.W. struck down the attempt, and thereby upheld the validity of "manner and form" limitations on the amending power, or incomplete self-entrenchment, even against a parliament supreme and omnipotent in the English tradition. Attorney General for New South Wales v. Trethowan, (1931) 44 C.L.R. 394; [1932] A.C. 526.[Note 17]

Another case, just as famous and more dramatic, occurred in South Africa in 1952. When the English parliament wrote the South African constitution in 1909 it included one Section (number 35) that prohibited racial discrimination, and one (number 137) that assured the equal status of the English and Dutch languages. A third section (number 152) incompletely entrenched both of these provisions and itself, requiring for their amendment a two-thirds vote of each house of the South African parliament sitting in joint session. In order to institute apartheid, the South African parliament tried to repeal all three of these sections by a statute, using only simple majorities. It knew that the 1909 clauses stood in the way of this plan, but argued that its intervening independence from Britain made it sovereign and capable of amending its constitution without British interference. The South African Supreme Court rejected the attempt, partly on the ground that South Africa was not completely sovereign so long as Britain could repeal the Statute of Westminster that granted former colonies their independence (see Section 20.I). Harris v. Dönges, (1952) 1 T.L.R. 1245. The South African government eventually established apartheid "legally" by packing the parliament until the two-thirds votes could be obtained.[Note 18]

The rule in the United Kingdom and its former colonies and dominions is that incomplete self-entrenchment clauses are binding and can only be amended or repealed by the procedures they lay down for themselves. In other words, a parliament cannot bind its successors irrevocably, but it can make things difficult. Continuing omnipotence, however, is not threatened by mere difficulty or by revocable limitations on power. Unfortunately no case has apparently arisen in which the "manner and form" of permissible amendment set out in the incomplete self-entrenchment clause is so difficult as to amount to virtually complete self-entrenchment. Such a case might reveal that complete and incomplete self-entrenchment differ only in degree, not in kind, and that an extremely difficult amendment procedure will be treated like a prohibition on amendment, not like a manner and form limitation.[Note 19]

If "revocability" is a matter of degree, matching the degrees of difficulty of amendment, then a parliament could approach the irrevocable binding of its successors by increasing the difficulty of the manner and form limitation. If courts strike down would-be irrevocable limitations on future power, should they strike down extremely difficult manner and form limitations? What about manner and form limitations that are not so difficult for some entrenched rules but that would be virtually impossible to overcome for certain popular rules by virtue of their popularity? For example, if an article permitting the institution of slavery and an article asserting the freedom of religion were each to be incompletely self-entrenched so that only 80% supermajorities of the national population in a referendum could repeal them, then the first would still be repealed while the second would become virtually immutable. Should courts take this into account when deciding what constitutes an irrevocable limitation on future generations?

When an incomplete self-entrenchment clause is to be violated by a proposed amendment, as attempted in Australia and South Africa, then the violation may possibly be cured by acceptance. In such cases the perceived fairness and difficulty of the procedures specified in the entrenchment clause would have some bearing on the acceptability of circumventing them. Circumvention would almost always become more acceptable as the limitations on the amending power were perceived as impositions of a foreign power or former imperial sovereign. When the specified procedure is not intolerably difficult, however, or perceived as a remnant of colonial subservience, then the acceptance model will usually agree with the inference model in finding manner and form limitations valid against the amending power. The acceptance model provides, however, that when acceptance in the appropriately complex sense is present, then it suffices to override even the most "reasonable" entrenchment clauses to cure violation or circumvention.

Under the inference model complete self-entrenchment clauses are immutable, providing that we tolerate the self-reference involved. A clause that said, "this section may not be amended," when it is part of the entrenched section, plainly forbids its own amendment. If it is an original limitation, then it supersedes the general language of the AC as specific language customarily supersedes general language (the lex specialis principle). If it is a self-imposed limitation, then it supersedes both as the more specific rule and as the more recent voice of the sovereign (the lex posterior principle) which impliedly repeals all existing portions of the text not reconcilable with it (see Section 16).

However, note that self-imposed complete self-entrenchment will almost always impliedly amend the AC, again both as the more specific and as the more recent language of the constitution. As an act of the AC that impliedly amended (limited) the power of the AC, it would constitute self-amendment, which the inference model could not tolerate. The inference model might solve the problem by dropping the lex posterior principle that gives recent law priority over older law in cases of irreconcilable conflict, since it is by virtue of this principle that the AC "transmutes" the "immutable" and amends the complete self-entrenchment clause. But this solution requires a change of law —for the lex posterior principle is actual law already— that the inference model cannot accomplish from its merely logical standpoint. Nor can it say that because the lex posterior principle has logically regrettable consequences some of the time that it is already repealed or limited in its applicability.

But the mere idea of this solution points up an important ambivalence of the inference model about the status of temporal priority of law in its relation to logical priority. On the one hand, the inference model requires that new law be authorized by temporally and logically prior law. That seems to imply deference to the lex posterior principle. So does the method of avoiding contradiction in law by invalidating (impliedly amending or repealing) the older rule in cases of conflict. But on the other hand, the lex posterior principle is denied by Ross's preferred method of avoiding contradiction in law, namely, to invalidate every attempt to derive a conflicting norm from an existing one, never allowing a conflicting rule to come into a position to impliedly amend or repeal an older rule. Moreover, if a pre-existing entrenchment clause forbids certain amendments, then the inference model would probably read any attempt to ratify an amendment of the forbidden type as a violation, not a repeal, of the language of the constitution.

Most models of legal change would read such an attempt this way, which suggests that the old AC could neither transmute nor amend a complete self-entrenchment clause. It would be prevented at least by contradiction, if not also by incompetency. The acceptance model is free to read it otherwise, however, if the attempt to ratify a forbidden amendment is allowed to run its course. For the dynamics of acceptance by the people and officials of the system, as opposed to whatever might be happening in ideality according to logicians, may validate the amendment despite its inconsistency with antecedent authority. This would make the amendment what it purports to be, an amendment, not a violation, of the language of the constitution. It is the lex posterior principle in action.

The inference model, then, probably forbids the enactment of complete self-entrenchment clauses on the ground that through them the AC limits its own power and performs self-amendment. But if such a clause is an original limitation on the AC, not self-imposed, then the limitation on the AC is not an amendment and hence not a case of self-amendment. Therefore it may limit the AC immutably for the inference model. For the acceptance model, by contrast, an amendment could repeal, rather than violate, the language of the constitution, since its authority need not rest exclusively on existing provisions of the constitution, but might arise instead from present or future acceptance.

The acceptance of the people is a source of authority that could justify the repeal of a complete self-entrenchment clause by an AC that was general in its language and silent on this special application. Under the acceptance model the violation or contradiction involved in allowing the AC to amend a provision that it is forbidden to amend can be cured if it contingently meets with the approval of the people and officials. The act still violates written law, but under the acceptance model that is not decisive; acceptance overrules written law, both to validate and to invalidate changes of law, even if it does so only rarely. If the people want to repeal a complete self-entrenchment clause, in short, logic will not stop them, and law will not stop them in the name of logic. Law may stop them on other grounds, of course, and as noted, the very provisions of the constitution that the amendment violates will have their effect in determining what is accepted.

Under the inference model an original complete self-entrenchment clause must be immutable, and a self-imposed self-entrenchment clause cannot exist. If, however, a self-imposed self-entrenchment clause could be enacted under the inference model, one might argue that it would be mutable, even if the self-entrenchment were complete. To make the argument work one must appeal to a principle of customary or "natural" law, often found in dicta, that a legal power can unmake anything that it can make. Such a rule is not at all unlikely, for it is a variant of the rule that one generation cannot bind its successors except revocably or in manner and form. If a proponent of the inference model finds this rule available to her, then she might allow some self-imposed complete self-entrenchment clauses to be mutable, if only they could exist to begin with.

Note, however, that this principle that legal powers can unmake anything they can make may give the people a legal right to abolish the constitution (see Section 18). This could result in the repeal of the AC, although not by the procedure of the AC. Proponents of the inference model, if they pick up the principle at all, are likely to qualify it by holding that a legal power can unmake anything it has made, but only by following the procedures then in effect, including any procedures made by the power to be unmade. If this qualification is not to reinstate the immutability of complete self-entrenchment clauses it must give an AC continuing omnipotence, or at least sufficient power to transmute or amend completely self-entrenched rules, even if no other existing procedure would allow it. The qualification could not give the AC self-embracing omnipotence, for then it could make something that it could not unmake, namely, a genuinely immutable limitation on the AC.

Enacting an immutable limitation on the AC would be impossible under the principle that legal powers can unmake what they make. It would also be impossible under the inference model stripped of that principle, since it would amount to self-amendment. Note that the qualified principle directly gives the AC the power that was sought in intermediate rules; it makes complete self-entrenchment clauses mutable by stipulating that the AC can amend them, not by appeal to a rule independent of the AC that could justify the result. If the AC was never limited by an immutable rule, then its repeal of a complete self-entrenchment clause should not count as an augmentation of its power, even if it is self-amendment. However, if even the mutable complete self-entrenchment clause is considered a limitation on the AC, albeit revocable by the AC, then repealing it would take us from a temporary period of limited power to a succeeding period of unlimited power, which is certainly a kind of self-augmentation of the power of the AC. It is like a person unlocking a room in which she had locked herself, or resigning from a game she had agreed to play. That would be a kind of self-amendment, revoking revocable self-limitations, and so would disqualify it under the inference model.

The acceptance model may authorize the repeal of a complete self-entrenchment clause by virtue of the capacity of the people and officials to overlook or tolerate contradiction. But there is a neater way. Suppose the AC amended itself[Note 20] by addition, appending to itself the following clause:

Through the procedures described in this Article the people may amend or repeal any provision now or ever to be part of this Constitution, without exception, notwithstanding any protective language, any supposed inherent or implied limitations on the amending power, any prohibition against self-application, any conjectured or actual reservations intended by any maker of this Contitution, or any future amendment to this Constitution not made in convention.

Could such a clause empower an AC to repeal a complete self-entrenchment clause? By its words and evident intent, certainly. But those words squarely contradict the words of a complete self-entrenchment clause, which might say, "nothing in this section may be amended by any means whatsoever". (They also contradict themselves, for which see below.) The fact that the AC may be worded to make it an irresistible force does not at all prevent us from wording a subsequent complete self-entrenchment clause to make it an immovable object.

If the amended AC does not suffice by its words to negate the words of the self-entrenchment clause, because it does not conceptually outflank the entrenchment clause, pierce a loophole in it, or supersede it as the more specific, then it may still control by the lex posterior principle if it is the later in time (see Section 16). If the self-entrenchment clause is original or self-imposed, but older than the AC, then it will lose by the rule of change that allows amendments to amend and by the rule of priority that favors recent law.

If a self-entrenchment clause is self-imposed after the AC is amended to include the new provision, then the entrenchment is the most recent voice of the sovereign. But the AC still says of itself that its power cannot be limited even by future amendment, except those made in convention. If the new self-entrenchment clause were not made in convention, then its protective language would probably fail in a court test. The rule that the clause entrenches would be exposed to ordinary amendment. One could say that the self-entrenchment clause, despite its lawfulness or validity, would have only "revocable effectiveness" because it remains subject to repeal. For a self-entrenchment clause that is a much more serious qualification than for other provisions law, almost enough to constitute its ineffectiveness or nullity.

If the self-entrenchment clause were made in convention, however, it could validly shield itself and the rules within its scope from amendment. At least the amended AC could not nullify the protective language. In such a case an acceptance theorist would turn back to the first method of repeal, and allow the AC to repeal the newer self-entrenchment clause if the proper show of acceptance by the people were made.

Conceivably, a "reaffirmation" of the amended AC sometime after the complete self-entrenchment clause would have the effect of giving the AC priority in cases of conflict, thereby giving it the power to repeal the clause. If such a course were to be lawful, it seems that the reaffirmation must receive the same supermajorities from the same bodies as ordinary amendment, which reduces to repetitious self-amendment, or it must receive some less formal show of faith, which reduces to overriding acceptance of the power of the AC to repeal the self-entrenched rule.

If we omitted the words "not made in convention" from the proposed addition to the AC, then it would be self-contradictory. It would give the AC continuing self-embracing omnipotence. It would prevent any new limitation on the amending power from ever taking effect, which is continuing omnipotence, and it would authorize all self-amendments without exception, which is self-embracing omnipotence. This is the contradiction at the heart of the paradox of omnipotence. The contradiction is clear when we think about using the new clause to repeal itself. By its own terms, it could do so, because it is a clause of the constitution and therefore mutable under the AC, but it could not do so, because to repeal it would limit the power of the AC, which the clause prohibits.

There are two ways to prevent this contradiction. First, word the addition so that continuing omnipotence is not created. This was done originally by allowing limitations on the AC to be adopted in convention. Second, word it so that self-embracing omnipotence is not created, for example by preventing some kinds of self-amendment through the use of a complete self-entrenchment clause. Perhaps the whole addition could except itself from the amending power. This path is the most precarious, for as we have seen self-amendment in violation of complete self-entrenchment clauses will be possible in actual legal settings, under the acceptance model. Therefore, if such an addition gives the AC continuing omnipotence, it will probably have self-embracing omnipotence as well, from the force of acceptance in practice, thereby bringing the contradiction back again. Of course, however, the contradiction is tolerable under the acceptance model, if it is actually accepted.

Self-amendment under the acceptance model can prevent, even annul, immutability. The annullment or transmutation of immutability is a contradiction in terms, but also an accurate description of what could legally occur if self-amendment is permitted. If an AC can transmute and amend a complete self-entrenchment clause simply by doing so, with the acceptance of the people and officials, and without first amending itself to give itself the power, then its superiority to legal immutability is plain. Legal immutability still exists only at the level of law, and acceptance transcends law. On the other hand, if self-amendment explicitly increased the power of the AC over complete self-entrenchment, then the new AC would have purchased two "tickets" to priority over a previously "immutable" complete self-entrenchment clause: more recent status and more specific language.

This also illustrates the claim in Section 3 that the postulate of any immutable rules would beg the question of the mutability of the AC. If the AC can amend itself, then perhaps it can arrange to reach and amend any rule of the system. But the self-applicability of an AC is neither necessary nor sufficient to achieve transmutation of immutable rules. If an AC can simply amend a complete self-entrenchment clause with the direct authorization of acceptance, then strict self-amendment does not occur and is unnecessary. Authority from acceptance is necessary and sufficient for the conquest of immutability. It is necessary because, without it, the contradiction in transmuting immutability would forever bar the act.[Note 21] It is sufficient because we need nothing more than the good grace to survive contradiction in order to amend an immutable rule.

Omnipotence of the AC is also necessary and sufficient to conquer immutability. When acceptance directly authorizes transmutation, it is making the AC omnipotent within the sphere of law. When immutability is transmuted, the amending power is accordingly empowered. In short, if acceptance and the omnipotence of the AC are both necessary and sufficient for the same condition, then they are the same. This is not really a new or interesting proposition, however, until we decide what kind of omnipotence in the AC we can mean.

Of the three powers or properties, then, self-applicability is less important than acceptance and omnipotence in conquering immutability. Omnipotence and immutability may collide like an irresistible force and an immovable object, but the object moves when the "authority" behind the force can tolerate contradiction.[Note 22] If the irresistible force can enhance itself by self-application, then it may or may not reach a state where it can move the immovable object. In the legal analogue of these forces and objects, this is an empirical question on which history or experiment have more to say than a priori analysis.

The transmutation of immutable rules is not merely academic. In 1965, citizens of Georgia challenged a state constitutional amendment that would consolidate several school districts and allow the officials of the school district to declare the effective date of the amendment. The delay in the effective date beyond the ratification date and the deference to school administrators in "realizing" constitutional law were not specifically prohibited by any language in the Georgia constitution, although there was general language to the effect that amendments went into effect when ratified. The Georgia Supreme Court held that

if there was [a prohibition of such delay and deference], this amendment is the latest expression of the sovereign will of the people upon the subject and 'will prevail as an implied modification pro tanto of the former provision.'

McCullen v. Williamson, 221 Ga. 358, 144 S.E.2d 911, 916 (1965).[Note 23] This suggests that even explicit language prohibiting such delay and deference could be overcome by an ordinary amendment through an AC not specifically self-amended for the job. The lex posterior principle is the universal solvent that helps new law overcome old prohibitions.

Amendments that are prohibited by the constitution, the case seems to say, may be adopted anyway, since the act of amendment is the most recent, and therefore overriding, act of the sovereign. The prohibited amendments are only prohibited until adopted, since adoption occurs under a power that always supersedes the power of the prohibition. The prohibition may as well have been completely self-entrenched, since the sovereign apparently has the power to revoke all past acts and, more than that, to revoke them in the very act of violating them. But of course the outcome might have been different if the court had to face explicit prohibitory language, especially explicit self-entrenchment of the prohibition.

The theory that the lex posterior principle allows certain violations to become valid law, or converts violations to amendments, is further discussed in Section 21.C.

Can self-amendment allow an AC of continuing omnipotence to reach and repeal the continuing character of its omnipotence? Under the inference model, even allowing self-amendment, the answer is certainly not, for if anything is ultra vires for an AC of continuing omnipotence it is to undo its continuing omnipotence. Any attempt would be a nullity, not a valid amendment that might take priority over pre-existing rules under the lex posterior principle.

But under the acceptance model, why not? There are reasons why not, but the final answer must be that the acceptance theory would allow it if properly accepted, which is an empirical question. One reason why this is not an easy question, even for the acceptance theory, is that while one generation may fully accept the abrogation of the continuing omnipotence of the AC and thereby allow itself to bind its successors (to bind them, it thinks, irrevocably), nevertheless future generations may refuse to recognize the abrogation or to recognize that they are irrevocably bound by prior generations in anything. One generation might accept a law which it believes it has made irrepealable, while the next generation might with equal authority accept its repeal. There is no reason not to call the law temporarily immutable, because when made, immutable rules were possible, not ultra vires; continuing omnipotence had been overthrown. All abrogations of continuing omnipotence carry this possibility of future reversal, and under the acceptance theory the abrogation may always be abrogated or the transmutation transmuted. The immutability of continuing omnipotence can only be transmuted by the adoption of a different immutable rule, and if the first can be abrogated by acceptance, then so can the second.

The result is that the only immutable rule in a system of continuing omnipotence may be "repealed" by one generation, and in the next declared simply "violated" —under the acceptance model. Any supposed irrepealable law that one generation uses to abrogate the immutability of continuing omnipotence may be repealed by the next generation, proving to the repealing generation that the law was never immutable and the continuing omnipotence therefore never abrogated, or never abrogated irrevocably. No objective standpoint exists that could allow us to call the act of the second generation a violation or a repeal of an immutable rule; we will always be subject to the relativity of the two generations. What seems to one generation like the transmutation of immutability, may seem to the next like failure, or vice versa.

Finally, it should be obvious that under the acceptance theory, acceptance itself has a kind of continuing omnipotence. Strictly speaking, acceptance is not a rule of change and therefore cannot be omnipotent in any way. But it is a source of authority for legal change that can tolerate or authorize any change of law whatsoever. No exception need be made for itself, for acceptance is not law and cannot be changed by changes of law. It is a social and historical phenomenon, and one of the virtues of Hart's acceptance theory is that it refuses to leave law magically divorced from social experience and grounds it firmly in the dynamic, contingent, historical reality of human decision, acquiescence, and practice.

Acceptance may be considered sufficient to authorize any change of law without considering it to embrace a moral or legal right to revolt. Indeed, once the authority to change law is condensed into a right, all the questions of self-amendment arise. Can a moral or legal right to revolt to used to justify its own repeal? (See Section 18.)

If the immutably continuing omnipotence of acceptance attempts to transmute itself, then it will fail in a sense that acceptance theorists can call objective and non-relative. If one generation attempts to disenfranchise acceptance by establishing a legal order that claims to derive legitimacy from the whim of the new autocrat or the ascertained whim of Jehovah, or anything but acceptance, then it may seem to that generation that acceptance is no longer authorized to define law. We certainly can imagine, or know cases of, shifts from popular sovereignty to more traditional military, religious, or elite foundations of legality. It would then seem that the legitimacy of the regime would not be subverted if acceptance waned or attached to an alternative. But this is just what is denied by the present theory. Even if the new autocracy cannot be said to be established through public and official acceptance, future generations may always overthrow the regime and establish an alternative that becomes law (as opposed to long-lasting, effective hooliganism) through public and official acceptance. The overthrow may be revolutionary in its violence or in its discontinuity with prior law, or it may be a constitutional amendment that replaces autocracy with democracy. For the acceptance theory, no generation can lose its power to define the ultimate premises of law through its acceptance. The people hold this power inalienably, though they may alienate it revocably in times of weakness, panic, or duress. Therefore, while acceptance may acceptably lie dormant for a time, it can always awaken to its own power; it cannot transmute its immutably continuing omnipotence, or irrevocably shift its powers to any other sources of law and authority.

Further wrinkles in the theory of immutably continuing omnipotence and the possibility of revocable self-transmutation are explored in Section 21.C.

C. Supposed limitations on the U.S. federal amending power

The case for the omnipotence of the amending power is usually heedless of the paradox of omnipotence. A case in point is Raymond Uhl's defense of amendatory omnipotence.[Note 24] Uhl follows Austin's theory of sovereignty, which holds that the supreme authority is also omnipotent and autonomous. So when Uhl finds the federal amending power to be the supreme power in the American legal system, he automatically attributes to it all the other attributes of classical sovereignty. "It is legally omnipotent, unlimited, and illimitable....There is no legal restriction on the extent to which this power may go."[Note 25] Uhl ignores the fact that if the omnipotence he attributes to the amending power is continuing (illimitable), then it is already limited, and if self-embracing (unlimited), then it is limitable.[Note 26]

Uhl claims that "express limitations" on the amending power, such as the requirement in the federal AC that no state be deprived of its equal suffrage in the Senate without its consent, "are but formal and at most can remain in force only so long as the sovereign permits them."[Note 27] He does not defend or elaborate this view, or consider difficulties such as entrenchment and self-entrenchment clauses. But his position seems to be that sovereigns cannot bind themselves, not even by their own laws, which suggests continuing omnipotence; but of course continuing omnipotence requires one immutable limitation on sovereignty. Uhl's only real argument for amendatory omnipotence is the following:[Note 28]

If the power to amend is limited we must assume that there is some authority capable of establishing such limitations. The only power legally able to do this is that power which can change the Constitution of the United States. Any limitations, express or implied, on the exercise of this power are at most self-imposed and can be removed by the authority which created them.

Here Uhl wins this conclusion too easily. First, he does not consider the possibility that self-imposed limitations may be immutable, for example through entrenchment or self-entrenchment. Second, the AC may be limited by original, not self-imposed, limitations. He seems to be saying that the sovereign can always free itself from its own laws (by repeal if not by privilege), which may be the case, but he shifts from regarding the AC as sovereign to regarding the people as sovereign.[Note 29] The people can undoubtedly overthrow a constitution guarded by any number of any kind of limitations on the amending power and even by limitations on the people's power to overthrow it. But whether an AC can do so is not at all an easy question, and not addressed by Uhl. Self-imposed limitations are not equally or obviously repealable. They may be entrenched or self-entrenched, or they may possibly limit or destroy the amending power irrevocably.

While Uhl has a naive sense of the omnipotence of the amending power, this turns out to be rare in the literature. Most scholars have a naive sense of its limitation and non-omnipotence.

The debate among jurists on the existence of implied limitations on the amending power turns for some on the question whether that power is delegated by the people. Delegated powers are often said to be limited or limitable per se.[Note 30] The opposite of a delegated power may be one inherent in government, not delegated by the people, or inherent in and retained by the people, never given up by delegation. This ambiguity plagues much of the discussion. Uhl, for example, denies that the amending power is a delegated power,[Note 31] but, as noted, his arguments shift back and forth between finding the sovereign amending power vested in the people and in the AC and the government. Edward Corwin, by contrast, believes the amending power is delegated and limited:[Note 32]

Probably the amending power, like all other powers organized in the Constitution, should be regarded as a delegated, and hence a limited power, although this does not necessarily infer [sic] that the Supreme Court is vested with authority to determine its limits.

The problem with most delegation theories is that they are vague or silent on the questions (1) why delegated powers must be limited, (2) whether the limits are revocable, and (3) how "delegation" should be traced back to intelligible social realities.

The arguments in favor of particular implied limitations are easier to evaluate than arguments based on the concept of sovereignty or delegation. One of the most serious arguments, which has led to much scholarly comment and even litigation, is the view that the federal AC has been impliedly amended, and limited, by the Tenth and Fourteenth Amendments. This theory asserts that the federal AC is limited in that proposed amendments that diminish the people's rights must be ratified, if at all, by state conventions rather than state legislatures. Congressional discretion to choose the method of ratification may have been unlimited in 1789, under Article V, but was limited (this theory goes) by the Tenth Amendment in 1791 and again by the Fourteenth Amendment in 1869. This argument is considered and rejected in Section 16 below. The theory that the AC may be limited or amended by judicial review is considered in Section 15.

William Marbury[Note 33] argued in 1919 that the amending power is limited by the intent of the framers. To him it followed that amendments must be improvements. Marbury cites the opinion in Livermore v. Waite, 102 Cal. 113, 119, 36 P. 424 (1894) (emphases in original):

The term Amendment implied such an addition or change within the lines of the original instrument as will effect an improvement, or better carry out the purposes for which it was framed.

This is an ingenious way to multiply the implied limitations on the amending power. Unfortunately it founders on the contestability of anyone's notion of what an improvement is. The only test compatible with the spirit of the constitution is whether the alleged improvement was ratified in the way specified for amendments.

Marbury is not daunted by this, and is quite sure that the framers did not intend that the amending power should be used to infringe on state powers or to enfranchise women. He argues by analogy to the taxing power, which is not expressly limited in the constitution but which was "wisely" held to implied limitations by the Supreme Court when it threatened to destroy the states. Marbury cites Collector v. Day, 11 U.S. 117 (1870) and Lane County v. Oregon, 7 U.S. 71 (1868). The taxing and amending powers are equally delegated powers, he argues, and therefore are equally subject to such implied limitations.[Note 34]

Marbury even worries that amendments should not be imposed on states against their will, implying that all ratifications should be unanimous,[Note 35] which contradicts both the explicit language of the constitution and the intent of the framers. He argues that the prohibition against depriving a state of its equal suffrage in the Senate without its consent implies a prohibition against any indirect interference with the functions of a state.[Note 36] He makes clever but sophistical use of the principle that Congress cannot make irrepealable laws by arguing that "legislative" amendments are impermissible because (as provisions of the constitution) they cannot be repealed by Congress.[Note 37] An amendment is "legislative" rather than "constitutional" when it deals so specifically with any subject that it is more appropriate for Congressional action and statutory regulation than constitutional amendment.

Marbury believes that all these arguments invalidate the Nineteenth Amendment (Women's Suffrage), and would invalidate the three Civil War Amendments if they had not been cured by long acquiescence.[Note 38] In short, he believes the implied limitations on the amending power protect state prerogatives and prohibit the amendments that abolished slavery and enfranchised blacks and women.[Note 39]

Aside from the argument against legislative amendments and the analogy to the taxing power, Marbury's arguments are clearly, almost comically, ideological rather than legal. He well illustrates how debates on the amending power raise the principles and passions surrounding our most basic rights and institutions: for an omnipotent amending power can sweep them all away.

One may always read clauses of the constitution in light of the framers' intent, when known. But the AC will always be a special case. To say that the framers never envisioned that the amending power might be used to enfranchise blacks or prohibit the sale of liquor may be true, but to cite such facts as if they limited the amending power is to forget that the clause we are interpreting authorizes change. Despite the framers' own substantive policies and preferences, in the AC they gave future generations the power to make whatever changes the people of those generations liked well enough to garner the wide consent stipulated in the AC. The evidence lies in Article V itself. Aside from the one substantive obstacle preserving the equal suffrage of the states in the Senate, the framers put only procedural obstacles in the way of amendment.

Marbury's analogy of the amending power to the taxing power breaks down and does not prove his point. First, the amending power might not be "delegated" as the taxing power is —although I will not enter the lists on that controversy. Second, the taxing power was not included in the constitution to make possible the peaceful adaptability to changing circumstances, and is not intended to change basic governmental structures. The amending power was and is these things, and to allow it to change some basic structures and not others is an arbitrary division not justified by the language of the constitution or even the intent of the framers. Marbury fails to show why states cannot consent through the amendment procedures to give up traditional prerogatives.[Note 40]

The argument that legislative amendments are prohibited usually rests on the declaration in Article I, Section 1, that "all legislative powers" are vested in Congress. The argument that a legislative amendment is impermissible because Congress cannot repeal it is specious, for the rule against binding one's successors irrevocably is satisfied as long as Congress makes no immutable statutes or contracts, and the amending body makes no immutable amendments. Neither Congress nor the amending body binds itself irrevocably when Congress and three-fourths of the states together adopt a revocable amendment. As for the former argument that all legislative power is in Article I, and none in Article V, the answer is that "legislative" and "constitutional" amendments differ only in degree of specificity. In making an amendment the people can draw the line wherever they please; if the courts accept it, then it it is a valid part of the constitution.[Note 41] The fact that the courts themselves may draw the line wherever they please, or refuse to draw it at all, suggests that we are not dealing with an inherent limitation on the amending power. Prohibition was probably the most "legislative" amendment in our history, and it survived four challenges to its constitutionality, including arguments that it was impermissibly legislative (see Section 16).

Marbury is also opposed here by a fellow delegation theorist. William Frierson has argued that even amendments that are concededly legislative are permissible and need not be defended as borderline "non-legislative" or "constitutional" amendments. His basic premise is that the constitution delegates legislative power to Congress in the first place, and the delegator of a power must be able to exercise the power it delegates.[Note 42]

The argument against legislative amendments, however, is most often made on policy grounds. One of the most thorough treatments of the policy questions is by Margaret Center Klinglesmith,[Note 43] who argued that the greater difficulty of amending the constitution, as opposed to statutes, made it imprudent to set the "ephemeral and the local" so far above responsive procedures. She was certainly right as a matter of policy. She was also right to note that the question whether there are any implied limitations on the amending power "is a momentous question,"[Note 44] for it will tell us whether our AC is omnipotent and whether our basic rights may in principle be repealed by a valid law (see Section 16).

Such policy arguments at best can show that legislative amendments, or other allegedly ultra vires amendments, are unwise, not illegal. Lester Bernhardt Orfield generally believes that no legal arguments are sound that might establish implied limitations to the federal amending power.[Note 45] Express limitations, however, might be valid.[Note 46] He is unsure and suggests that an express limitation such as that protecting the equal suffrage of the states in the Senate, "could probably be repealed by a unanimous vote of the states."[Note 47] Limitations not so closely tied to individual state consent could probably be repealed by the normal supermajorities. Walter F. Dodd also rejects the idea of implied limitations on the federal amendment power.[Note 48] (State ACs are all impliedly limited by a requirement that state constitutions conform to the federal constitution.) A more radical and confident statement is to be found in McCullen v. Williams, 221 Ga. 358, 144 S.E.2d 911, 916 (1965):

[A]ny attempt to limit the will of the people with respect to the substantive content of an amendment would amount to a gross infringement upon their sovereign right to establish, modify, and alter their organic law as they see fit....

If an AC is not absolutely valid, supreme, and omnipotent, but only valid, supreme, and omnipotent until amended, then is it limited? The paradox of self-amendment may inspire one to think of the validity of an AC as existing under a time limit (see Sections 10 and 14). Does that time limit comprise a limit on the power of the AC? The answer must be no, for any amendment that is authorized at all is authorized by an AC that is valid until amended, provided it is made while the AC is valid. If the time limit is not set for any particular moment, but is implied by the possibility of an amendment undoing its validity, supremacy, or omnipotence, then only self-amendment could impose the contemplated limitation. An AC that is unlimited but for self-imposed self-limitations has self-embracing omnipotence, and can enact any amendment whatsoever until it takes away its own power to do so. If the time limit is set for a particular time, say 12:01 AM, 2000 A.D., then the AC is under a sunset clause (see Section 14), but could still enact any amendment before the expiration date which an AC not under a sunset clause could enact. Similarly, a time limit contradicts continuing omnipotence, but would not prevent the adoption of any amendment that continuing omnipotence could adopt.

The only possible exception is a repeal of the time limit itself, which might be entrenched, at least impliedly. If the validity of entrenchment is upheld, such self-amendment would be prevented. Alternately, repealing the time limit might be considered a way in which the AC could entrench itself against future amendment. If self-amendment is possible, then self-imposed self-entrenchment of the AC is possible. If the time limit is put into words that forbid extensions of the expiration date, then the question is converted to the permissibility of the self-disentrenchment of the AC. For all these problems, see Section 9.

Some limitations that have been alleged to be implied in the federal AC have been litigated and dismissed by courts. For example, some have argued that all amendments must be "germane" to the existing text, that is, specifically addressed to an existing clause and amending it. Such a limitation would preclude all amendments by addition and, in some versions, all germane amendments that fail to state which clauses they are to amend. This theory was rejected in State of Ohio v. Cox, 257 F. 334, 342 (S.D. Ohio 1919).[Note 49]

Because Article V says that Congress may propose amendment "whenever two thirds of both Houses shall deem it necessary," some have argued that no amendment is valid unless Congress makes a specific finding of "necessity". This argument was levelled against the extremely unpopular Eighteenth Amendment (Prohibition), as was the "germaneness" argument and many others, as part of a series of desperate but clever attempts to invalidate it (see Section 16). The Supreme Court rejected the "necessity" theory, and found a sufficient declaration of necessity implicit in a two-thirds vote. State of Rhode Island v. Palmer, 253 U.S. 350 (1920). Klinglesmith argued that the framers' use of the phrase "whenever two thirds of both Houses shall deem it necessary" is evidence of their intent to prohibit legislative amendments.[Note 50]

One of the most relentlessly ideological arguments for implied limitations is Arthur Machen's against the validity of the Fifteenth Amendment (Black Suffrage).[Note 51] Machen bases his argument on the only express limitation on the federal AC still in effect: "that no State, without its consent, shall be deprived of its equal suffrage in the Senate" (Article V). From this express limitation Machen infers many implicit limitations, including the necessity of preserving the Senate,[Note 52] preserving the states, preserving the states as entities capable of giving consent, and preventing any alteration of "the composition of a state" without its consent.[Note 53] He argues that "a state" is its people, particularly its citizens and voters, which immediately after the Civil War meant the white people of a state. The right of the white people of a state to elect two senators (by way of the state legislature) is taken away by the enfranchisement of blacks, especially in states like South Carolina where blacks outnumbered whites in 1870 when the Fifteenth Amendment was ratified. This would be legally irrelevant if it were not that Machen finds this dilution of white voice a diminution of their equal suffrage in the Senate. The states, as their white voters, have lost their equal voice in the Senate if their voice is diluted by the enfranchisement of blacks.[Note 54] Machen uses the one-person, one-vote principle, derived from the Fourteenth Amendment, before it was articulated by any court and which partisans of his camp were to oppose bitterly, to challenge the deprivation of equal suffrage, or equally weighted voices in the Senate, of a state considered as its white voters. He compares the sudden enfranchisement of blacks to the annexation of Puerto Rico to Rhode Island, depriving the Anglo Rhode Islanders of their equal voice in the Senate. This argument makes clever use of the Fourteenth Amendment to oppose the Fifteenth Amendment.

In many places Machen reveals the tenacious racism that drives his legal arguments. The annexation to Rhode Island of Puerto Rico, when Vermont would have worked as well, is just one example. He claims that states are their people, but that "[t]he white people and they alone constituted the State of South Carolina,"[Note 55] because they alone could vote. The fact that blacks could vote after the Fifteenth Amendment did not make them part of their states; it violated the rights of whites. "The Fifteenth Amendment amounts to a compulsory annexation to each state that refused to ratify it of a black San Domingo within its borders."[Note 56] He finally falls back on an "abuse of power" argument, and asserts that the greatest abuse of power is "to take a state government out of the hands of its citizens and give it over to a mass of persons who by its laws were mere chattels."[Note 57]

The Supreme Court has never ruled on the existence of Machen's implied limitations within the express limitation of the equal suffrage clause of Article V, nor has it ever directly ruled on the validity of the Fifteenth Amendment. Only once has the Supreme Court ruled on a theory that an implied limitation on the federal AC lay inside the equal suffrage clause. It held that the Senate is not barred by that clause from refusing to seat a newly elected member presenting proper credentials pending investigation of the legality of his election. Barry v. United States, 279 U.S. 597 (1929).

The equal suffrage limitation on the AC is an entrenchment clause protecting the equal suffrage of the states in the Senate. It is incomplete entrenchment because it can be repealed by a difficult procedure —getting the consent of the states to be deprived of their equal suffrage. It is not self-entrenched, at least not expressly. Therefore, it might be repealed by an ordinary amendment, provided the losers by it voted with the majority. To prevent this one might want to find an implied limitation on the AC forbidding repeal of the equal suffrage limitation except by, say, unanimous consent of the states. The rationale is that the consent of each state is necessary because each state might be a victim if the clause were repealed (see Section 9.A). In general one might find in any AC an implied limitation self-entrenching any incomplete entrenchment clause at the same level of difficulty it requires for the amendment of the rule it protects. Such an implied limitation is certainly a reasonable reading of the intent of the framers: this can be said without reference to any particular clause or framers because an incomplete entrenchment clause that is not self-entrenched is virtually pointless.

Such an implied limitation, however, would not interfere with the omnipotence of the AC, for at most a merely incomplete self-entrenchment is read into a clause, which could always be repealed by its own difficult procedure. If a complete entrenchment clause is impliedly self-entrenched, for the same reasons, then a limit on the omnipotence of the AC may be implied with it. If the AC can transmute immutable rules, its omnipotence will survive the implied limitation; otherwise not. Whether an AC can repeal a complete self-entrenchment of itself will be explored in Section 9.

Alf Ross has argued in effect for another implied limitation on an AC, namely, that it never violate the logical rules of inference, such as modus ponens, in the derivation of a new amendment from the old AC. This is to say that no amendment can be adopted that is inconsistent with, or not deducible from, a set of premises that includes the existing AC. This position has, I think, been sufficiently dealt with. It is an application of a priori rules to an empirical subject; it requires outcomes that no court has recognized and forbids outcomes that are commonly, indeed, invariably upheld by courts. It purports to overrule law in the name of logic and thereby betrays its alegal character. The presumption that law must be logical is very strong in many other areas, for example in the rule of priority that prefers the more recent rule when two or more rules conflict irreconcilably (see Section 16). We have such rules because we would rather reconcile or eliminate inconsistent laws than live with them. But of course this is a policy decision that we can reverse or suspend. If there is a contradiction in self-amendment, no legal rule forbids it.[Note 58] (The variable value of consistency in law is further discussed in Section 21.B.)

Because Ross believes that self-amendment is impossible, he allows the amendment of an AC only by means of a tacit rule transcending the constitution which authorizes just such changes. As noted in Section 6.B, Ross must and does regard this rule as immutable. If it were mutable, then the higher rule that authorized its amendment would be immutable, and so on. Such rules stand as limitations to the actual AC in the constitution; because they are tacit or implied, they are implied limitations. Their immutability protects them from the AC, as does as their hierarchical superiority to it. The permissibility of direct self-amendment eliminates the need to posit such rules, but they might nevertheless one day be posited or "discovered" and cited by a judge. If so, they would be judge-made limitations on the AC, which could always, when accepted, undo the omnipotence of the AC (see Section 15). That they would undo it revocably was argued above (Section 8.B, above; see also Sections 9 and 21.C).

This leads to a more interesting problem. If the AC can be amended by rules of change other than itself, it is potentially limited to that extent. In Part Two I present six methods of amending the federal constitution other than direct amendment through the AC. All are "implied" in the sense that the "official" view is that the methods of the AC are exclusive and, for example, that judicial review does not amend but only interprets and applies the constitution. The availability of such methods may limit the omnipotence of the AC, but may also, by the same token, assure its mutability even if self-amendment is one day (revocably) ruled impermissible.

Any immutable rule would be a limitation on the power of the AC, or at least a prima facie limitation until we determined which had to give way, the immutability of the rule or the omnipotence of the AC. One rule that appears to be immutable is the rule that what the constitution says is law, is law.[Note 59] In Hart's terms, that rule is a rule of recognition. It may be changed if acceptance shifts and authorizes a different rule of recognition, such as a rule declaring that only the utterances of a dictator will be considered law. But can the supreme rule of change, the AC of the constitution, change the rule that what the constitution says is law is law? Any amendment that purported to repeal that rule would rely on its validity and thereby fail to repeal it —unless it effectively declared its own validity independently of the rule. An amendment that purported to modify, without repealing, the rule might have better success. The rule could apparently be changed to read, "what the constitution says is law is law and what treaties say is law is law on a par with the constitution" (amendment by addition), or "what the constitution says is law is law except for what it says about the powers of constitutional conventions, which is not binding on such conventions so long as they are lawfully convened under their own rules" (amendment by exception or subtraction). Amendments to the rule that simply reversed the proposition, and declared that what the constitution said is law is henceforth not law, would not only be unlikely in fact, but even if passed they would not comprise limitations on the AC. An AC cannot make a donkey into a horse, but it may be unlimited in power as an AC, that is, as a rule of change for constitutional rules. If we realize that even an omnipotent AC can only make constitutional rules, we see that it is not limited by its inability to make anything else, such as constitutional rules that were not laws.

Similarly, in the theological version of the paradox, James Cargile has argued that the omnipotence of a god is not limited by its inability to make something not made by a god.[Note 60]

Let us now shift from the new theories of "natural law" that make legal validity depend on logic to the older theories that made it depend on morality. One of the most fundamental of the alleged limitations on an AC is an implied prohibition, implied by due process or principles of fairness, against retroactively abrogating vested rights by a change of law. The states agree that no one has a vested right in the law as it is that might preclude its amendment or repeal, that any reliance on existing law may be upset by proper use of the procedures of legal change, and that states make no implied promise to protect their citizens from incidental injury caused by changes in the law. See e.g. Tom & Jerry, Inc. v. Nebraska Liquor Commission, 183 Neb. 410, 160 N.W.2d 232 (1968), Cirelli v. Ohio Cas. Ins. Co., 133 N.J.Super. 492, 337 A.2d 405 (1975). But often the cases limit the states' discretion to change law by a standard of reasonableness, or a duty to avoid arbitrariness and comply with due process. Id. Some cases are more specific in prohibiting unreasonable impairments of contracts by statute or constitutional amendment. City of New York v. Town of Colchester, N.Y.S.2d 156, 66 Misc.2d 83 (1971). And most agree that a retroactive abrogation of vested rights cannot be effected by statute, Rosenberg v. Town of North Bergen, 61 N.J. 190, 213 A.2d 662 (1972), or in some places even by constitutional amendment, Permann v. Knife River Coal Min. Co., 180 N.W.2d 146 (N.Dakota, 1970).

Is the amending power limited by justice, fairness, or equity? Here we reach a fundamental question of legality. An old Virginia statute set out a right of free exercise of religion, and then added:

And though we well know, that this assembly, elected by the people for the ordinary purposes of legislation only, have no power to restrain the acts of succeeding generations, constituted with powers equal to our own, and that, therefore, to declare this act irrevocable, would be of no effect of law, yet we are free to declare, that the rights hereby asserted, are of the natural rights of mankind, and that if any act shall be hereafter passed, to repeal the present, or to narrow its operation, such act will be an infringement of natural right.

See Currie's Administrators v. the Mutual Assurance Society, 4 Hen. & M. (14 Va.) 315 (1809). This statute recognizes the continuing omnipotence of the legislature, much like a British parliament. It recognizes that one assembly cannot bind its successors irrevocably, and yet tries to protect its assertion of fundamental rights from amendment or repeal. While it denies the existence of any legal duty on future generations to let the statute stand, it imposes a moral duty to let it stand. While repeal may be lawful, it will violate the fundamental rights outlined in the statute itself. The legislature, reaching the limit of its power, placed a curse on its posterity in place of an injunction.

Can natural rights, insofar as they are expressed in positive law, lawfully be repealed? Is an act that repeals positive law but violates natural rights thereby a violation of positive law? Might it be within the powers of a legislature or amending body and still be a wrong, such as a tort? If the statute adds something to the common law, then repeal seems permissible and effective. If the statute simply restates the common law or customary morality, which was the medieval view of statutes, then perhaps a repeal will be either impermissible, as a violation, or ineffective, leaving intact the unreachable moral environment.

Ronald Dworkin has argued convincingly that principles may legally supersede and qualify legal rules, and therefore that the line between legal and alegal standards cannot be fixed as rigidly as legal positivists would like.[Note 61] If justice limits the amending power, then it seems that an unjust constitution erected by revolution could never be ratified by acceptance or acquiescence, even if other revolutionary constitutions could be so validated.[Note 62] I made a similar argument in Section 2. If government is legitimate only by virtue of the consent of the governed, and if the people's use and contentment not to use the AC is a measure of their consent to be governed by the constitution, then the difficulty and fairness of the amending process must be taken into account. A procedure could be so difficult or unfair that the legitimacy of the constitution would be diminished. Similarly, acceptance could be indefinitely withheld from an unjust amendment or constitution. That would deprive them of legal validity under the acceptance theory, something different from legitimacy under the consent theory.[Note 63]

My use of the acceptance model in this essay, as noted, departs from Hart's in some particulars that are relevant here. I have assumed that the outcomes of procedures validated by the rule of recognition may be invalid if not accepted and, on the other side, may be valid so long as they are accepted even if the rule of recognition was violated, the procedures were defective, or the outcome contradicts its authority. I have assumed that acceptance does not merely validate the rule of recognition, and some otherwise unauthorized judicial decisions, while leaving the rest of the legal system to operate under a Rossian inference model. I have assumed that the curative powers of acceptance may intervene at every stage and provide legal authority for self-contradictory entities and procedures, defective operations, skipped steps, and an infinity of other anomalies that strict compliance with rules would not permit.

From this basic model of law we must conclude that the effect of injustice on the validity of an amendment or new regime is a matter of contingent history. They may be accepted or not depending on a host of variables such as the perceived need for order over justice, the rhetorical effectiveness of those promising to deliver justice, the packaging of the unjust law, false consciousness in the population, the power of the beneficiaries of the unjust law over its victims, the ratio of beneficiaries to victims, and so on. But I insist that, while unjust laws and changes in law may be legal or valid by virtue of both acceptance and preexisting rules of change,[Note 64] legitimacy may be lost or diminished by an unjust change of law that lessens the people's ability to change the law in the future.[Note 65] In this sense the people's acceptance is a source of authority with continuing omnipotence both descriptively and normatively. Descriptively, the people and the laws are powerless to limit irrevocably the power of acceptance to validate law. Normatively, an unjust legal impediment to the people's continuing power to change law weakens the legitimacy of the system of law to that extent.[Note 66]

Notes

1. One might argue that actual legal systems depend on, or inevitably embrace, at least a few principles of "natural law" or morality, and that the latter are absolutely immutable. For example, one might argue that not even revolution can change the rule that the people have a right to revolt (see Section 18), that promises should be kept, that self-defense excuses homicide, and so on. For the view that natural law is absolutely immutable, see Gaius, Digest, 7.5.2.1; Justinian, Constitutio Dedoken 18, Constitutio Tanto 18, Codex 1.17.2.18; Thomas Aquinas, Summa Theologica, I-II, 97, 1 and 1. See also E.K. Williams, "The Unalterable Law," Canadian Bar Review, 17 (1939) 551-57; Giorgio Del Vecchio, "Change and Permanence in Law," The Jurist, 18 (1958) 18-38; Charles E. Whittaker, "Immutable Moral Values," Texas Bar Journal, 27 (1964) 597-604; and David Daube, "Greek and Roman Reflections on Impossible Laws," Natural Law Forum, 12 (1967) 1-84 at p. 23. For the opposite view see Karl Kreilkamp, "Dean Pound and the Immutable Natural Law," Fordham Law Review, 18 (1949) 173-203; and Richard B. Hall, "The Alterability of Natural Law," The New Scholasticism, 55 (1981). For the view that legal principles in Dworkin's sense may survive revolution, see J.M. Eekelaar, "Principles of Revolutionary Legality," in A.W.B. Simpson (ed.), Oxford Essays in Jurisprudence, Second Series, Oxford University Press, 1973, pp. 22-43. [Resume]

2. H.L.A. Hart, The Concept of Law, Oxford University Press, 1961, at p. 146. [Resume]

3. These formulations are not synonymous with the claim that omnipotence is "infinite" power. The power to authorize an infinity of rules —one plausible reading of "infinite power"— may be incompetent to authorize a particular rule, or even a particular infinity of rules. As Cantor proved, not all infinities are equal, and even among equal infinities some are proper subsets of others. The formulation I prefer is that omnipotence is unlimited rather than infinite power, except that continuing omnipotence is limited in its power of self-limitation. The negation of omnipotence, however, is captured equally well for most purposes by the phrases "finite" and "limited" power. [Resume]

4. J.L. Mackie, "Evil and Omnipotence," Mind, 64 (1955) 200-12. [Resume]

5. These terms are somewhat rare in the United States but common in Britain and its other former colonies and dominions. See e.g. Hart, op. cit. at pp. 146ff, and D.V. Cowen, Parliamentary Sovereignty and the Entrenched Sections of the South Africa Act, 1951. See also George Winterton, "Can the Commonwealth Parliament Enact 'Manner and Form' Legislation?" Federal Law Review, 11 (1980) 167-202. [Resume]

6. Hart, op. cit. at p. 70; R.D. Lumb, "Fundamental Law and the Processes of Constitutional Change in Australia," Federal Law Review, 9 (1978) 148-84 at p. 180. [Resume]

7. Hart, op. cit. at pp. 146-47. Hart denies that parliament binds its successors equally in the two cases. "Parliament has not "bound" or "fettered" Parliament or diminished its continuing omnipotence, but has "redefined" Parliament and what must be done to legislate." Ibid. at p. 147. In the passage following this quotation Hart admits that "binding" and "redefining" one's successors cannot be kept rigidly distinct. [Resume]

8. Hart, ibid. at p. 146. [Resume]

9. The classical statements and sources of this axiom are well collected by Ilmar Tammelo, "The Antinomy of Parliamentary Sovereignty," Archiv für Rechts- und Sozialphilosophie, 44 (1958) 495-513. See also O. Hood Phillips, "Self-Limitation by the United Kingdom Parliament," Hastings Constitutional Law Quarterly, 2 (1975) 443-78. [Resume]

10. Note that binding one's successors or oneself implies self-embracing power but not necessarily self-embracing omnipotence. This distinction will become important when we ask whether a self-embracing omnipotent entity can repeal or destroy itself (Section 9) or change itself into any other entity (Section 11). [Resume]

11. See the many cases collected in Corpus Juris Secundum, "Statutes", Section 279.n.80. [Resume]

12. Ibid., Section 279.n.81. [Resume]

13. This was a favorite observation of William Earle, professor of philosophy at Northwestern University. [Resume]

14. However, for cases of self-amendment invalidated on other grounds, see the comments on self-amendment in Arkansas and Nevada in Appendix 2. [Resume]

15. Therefore, anyone who objects to self-limiting self-amendment must also object to self-imposed limitations, at least to self-imposed complete self-entrenchment clauses, even if they do not directly entrench or limit the AC. This just one way to say that adding an immutable provision to the constitution would limit the AC irrevocably just as much as a direct attempt at irrevocable self-limitation. [Resume]

16. While self-embracing omnipotence may be converted into continuing omnipotence by the simple means of adopting a completely self-entrenched self-amendment that protected the continuing character of the AC from subsequent limitation, it turns out to be impossible to convert self-embracing omnipotence into a continuing self-embracing omnipotence, or one that is immutably self-embracing. See Section 11. [Resume]

17. See W. Friedmann, "Trethowan's Case, Parliamentary Sovereignty and the Limits of Legal Change," Australian Law Review, 24 (1950) 103-08; Zelman Cowen, "Parliamentary Sovereignty and the Limits of Legal Change," Australian Law Review, 26 (1952) 237-40; H.W.R. Wade, "The Basis of Legal Sovereignty," Cambridge Law Journal, no vol. (1955) 172-97. [Resume]

18. Geoffrey Marshall, Parliamentary Sovereignty and the Commonwealth, Oxford University Press, 1957, Chapter XI; H.L.A. Hart, "Self-Referring Laws," Festskrift Tillagnäd Karl Olivecrona, Stockholm: Kungl. Boktryckeriet, P.A. Norstedt & Söner, 1964, pp. 307-16, at pp. 311-12, and The Concept of Law, Oxford University Press, 1961, at pp. 70-72; D.V. Cowen, op. cit. note 5 above, passim; H.W.R. Wade, op. cit. note 17 above, passim. [Resume]

19. An AC that seems to be omnipotent is in practice very seriously limited in its power when incomplete entrenchment clauses require very difficult amendment procedures. But in theory, if I may use an analogy from the theory of computation, continuing omnipotence is no more overthrown by severe difficulties of procedure than the computability of a function is denied by NP-completeness. But this latter fact is of merely theoretical importance. While it might affect the theological version of the paradox of omnipotence, where practical difficulties can never impede the kind of power at the center of the paradox, this theoretical point is as uninteresting in law as the computability of NP-complete functions is to working programmers. [Resume]

20. Self-amendment in general, and not only the special case of the repeal of complete self-entrenchment clauses, may involve the capacity of the people and officials to overlook and tolerate contradiction. But the method to be outlined relies only on self-amendment, not directly on the repeal of completely self-entrenched rules. [Resume]

21. Strictly speaking, acceptance is not necessary if some other authority for law, e.g. the might of an army or the will of a deity, can tolerate contradiction as acceptance can. [Resume]

22. If contradiction can be tolerated, it is just as likely that the "object" will resist the "force". When the authority behind the "force" is acceptance, then the "force" moves the "object" if that is what the people and officials have accepted. Since it is the people and officials who apply the "force" by using the amending procedure, the "object" or old law gives way. If the authority behind the "force" were the will of a deity, for example, and if the deity's priests reported that the old rule had the deity's favor but the new amendment movement did not, then the "object" would resist the "force". This is the question whether temporally posterior rules (the new amendment) have logical priority over temporally prior rules (the old rule to be amended) —or whether the living generation is sovereign and can overrule the sovereign acts of prior generations. The answer is contingent on how the people interpret the source of authority in their legal system. Acceptance understood as acceptance always gives the living generation sovereignty, that is, at least the kind of omnipotence required to amend any law it has inherited. [Resume]

23. Emphases added; see Corpus Juris Secundum, Constitutional Law, section 39.n.42. [Resume]

24. Raymond Uhl, "Sovereignty and the Fifth Article," Southwestern Social Science Quarterly, 16 (1936) 1-20. [Resume]

25. Uhl, ibid. at p. 15. Uhl cites Austin and Dicey at p. 19.n.40. For a good collection of Austinian statements that every legal system must have an unlimited law-maker, see Lester Bernhardt Orfield, The Amending of the Federal Constitution, University of Michigan Press, 1942, at p. 137.n.20. Orfield himself denies this proposition, ibid., p. 138. Orfield's own discussion of the sovereignty of the amending power, ibid., pp. 130ff, is more sophisticated than Uhl's, but also too loose on the question of the omnipotence of the AC and the power of the sovereign to determine the limits of its own competency; see ibid., pp. 130, 133. [Resume]

26. Orfield falls into the same trap with this formulation, ibid., p. 157:

Sovereignty, viewed in the broadest sense, may be regarded as the power to make a law on any subject binding for all time in all places. This capacity theoretically exists in the amending body. [Resume]

27. Uhl, op. cit., at p. 18. [Resume]

28. Uhl, ibid., at p. 19. [Resume]

29. Orfield's discussion is superior to Uhl's chiefly in its clarity on this question. Orfield unambiguously believes that the sovereign amending power is the power of the people's elected agents, not that of the people themselves or of the government. See Orfield, op. cit., at pp. 102, 124, 138, 141f, 143, 148. [Resume]

30. Why this should be so is not often stated clearly. The Tenth Amendment does not assume that the delegated powers are limited or limitable, only that the undelegated powers will not be exercised or usurped. The ultimate rationale is probably that the federal government is limited and has only those powers expressly and impliedly given to it, unlike the state governments which have all the powers not expressly or impliedly denied to them. But even this rationale cannot explain why a power expressly or impliedly delegated to the federal government could not be unlimited or illimitable, unless we imagine that one unlimited power could make all the other powers unlimited. Another possible theory is that the people alegally retain the sources of all legal authority, which prevents any power delegated by them from being unlimited or illimitable, even state power. That is, the acceptance and usage, or the consent, of the people has continuing omnipotence. [Resume]

31. Uhl, op. cit., at p. 17. [Resume]

32. Edward Samuel Corwin, The Constitution and What It Means Today, Princeton University Press, third ed., 1924, p. 88. [Resume]

33. William Marbury, "The Limitations Upon the Amending Power," Harvard Law Review, 33 (1919) 223-35. [Resume]

34. Marbury, op. cit., at pp. 226-28. [Resume]

35. Marbury, ibib., at pp. 224, 228. At p. 228:

Congress has no inherent power to adopt, and the legislatures of three fourths of the States have no inherent power to ratify, any amendment to the Constitution of the United States, and make it binding upon nonassenting States.

Article V is explicit in making unanimity unnecessary for amendment, and makes no provision for dissenting states to be exempted from the effect of amendments ratified by others. Here, then, Marbury is no longer interpreting the constitution, but only complaining from the standpoint of his political philosophy or what he might call natural law. However, even from his own standpoint, one might find that state assent to all properly ratified amendments lies in each state's consent to the constitution upon its admission to the union. [Resume]

36. Marbury, ibid., at p. 229. [Resume]

37. Marbury, ibid., at p. 229. This argument is repeated, including the conclusion that the Nineteenth Amendment is "legislative" and therefore void, by George Stewart Brown, "Irresponsible Government by Constitutional Amendment," Virginia Law Review, 8 (1922) 157-66. [Resume]

38. Marbury, op. cit. at pp. 230f. He particularly notes that if the Civil War Amendments had been challenged closer to the times of their adoption, a right-thinking Supreme Court would have nullified them. Ibid. at p. 231. [Resume]

39. Marbury does not say why state prerogatives supersede the federal amending power except to cite the framers' intent without documentation. He is probably vulnerable to this objection by Westell Woodbury Willoughby, The Constitutional Law of the United States, Baker, Voorhis and Co., 3 vols., second ed. 1929, I:599:

To the author of the present treatise, the fundamental error of all those who have sought to place inherent limitations upon the [federal] amending power...is that they necessarily start with the assumption that the Constitution is in the nature of an agreement or compact between the States, or that it implies an understanding between them, or between them and the National Government, that the allocation of powers as provided for in the original instrument shall not be changed in any of its more important or essential features. It is surprising to this writer that this theory, which since the Civil War, has been so decisively rejected by the American people and by the courts, should again be brought to support a constitutional argument [against the validity of the Eighteenth and Nineteenth Amendments].

Note that the only court to strike down a ratified amendment to the federal constitution appealed to the view of the constitution as a pact or treaty among the states. U.S. v. Sprague, 44 F.2d 967 (1930) (striking down the Eighteenth (Prohibition) Amendment), reversed 282 U.S. 716 (1931). See Section 16 for more discussion of the Sprague case. [Resume]

40. George B. Skinner also believes that some state powers, particularly the police powers, are inalienable and cannot be abolished or shifted to the federal government even with state consent. George B. Skinner, "Intrinsic Limitations on the Power of Constitutional Amendment," Michigan Law Review, 18 (1919) 213. Skinner bases his argument on the Ninth and Tenth Amendments, not the intent of the framers of Article V. He concludes that the Eighteenth (Prohibition) Amendment is void. See Section 16. [Resume]

41. For an example of a state amendment struck down for its specificity and self-mutability, see Section 14 and State ex re. Halliburton v. Roach, 230 Mo. 408, 130 S.W. 689 (1910). [Resume]

42. William L. Frierson, "Amending the Constitution of the United States: A Reply to Mr. Marbury," Harvard Law Review, 33 (1920) 659-66 at p. 664. [Resume]

43. Margaret Center Klinglesmith, "Amending the Constitution of the United States," University of Pennsylvania Law Review, 73 (1925) 355-79. [Resume]

44. Klinglesmith, ibid., at p. 373. [Resume]

45. Orfield, op. cit., at pp. 24, 45, 84ff, 87-88, 92, 96ff, 115-26, 156f. [Resume]

46. Orfield, ibid., at p. 205. [Resume]

47. As to the equal suffrage in the Senate limitation, Charles K. Burdick agrees that repeal by unanimous consent of the states would suffice. Charles K. Burdick, Law of the American Constitution: Its Origin and Development, G.P. Putnam's Sons, 1922, at p. 45. However, on this point Uhl is certainly right that the usual three-fourths supermajority could suffice if the states to be deprived of their equal suffrage by the amendment voted with the majority. Uhl, op. cit., at p. 16. This implies that if a state is to receive three senators, then unanimous ratification (possibly minus the state to benefit) would be necessary. On repeal of the limiting clause outright, Uhl would agree with Orfield and Burdick on the need for unanimity. [Resume]

48. Walter F. Dodd, "Amending the Federal Constitution," Yale Law Journal, 30 (1920) 321-54, at p. 338. [Resume]

49. Cf. Willoughby, op. cit., at pp. 598-600. [Resume]

50. Klinglesmith, op. cit., at p. 375. [Resume]

51. Arthur Machen, Jr., "Is the Fifteenth Amendment Void?" Harvard Law Review, 23 (1910) 169-93. [Resume]

52. The necessity to preserve the Senate under the equal suffrage limitation is denied by Blewett Lee, "Abolishing the Senate by Amendment," Virginia Law Review, 16 (1930) 354-69. He argues that the equal suffrage limitation is not an impediment to the abolition of the Senate, but would merely be ineffective if the Senate were abolished, ibid. at p. 365. Lee represents a school of democrat that advocated the abolition of the Senate as a vestige of aristocratic rule smacking of the English House of Lords.

The analogous British problem of abolishing the House of Lords is complicated by the legal omnipotence of the parliament of which it is a part. Hence, only self-embracing omnipotence could do the job. See the dispute between George Winterton and Peter Mirfield: George Winterton, "Is the House of Lords Immortal?" Law Quarterly Review, 95 (1979) 386-92; Peter Mirfield, "Can the House of Lords Lawfully Be Abolished?" Law Quarterly Review, 95 (1979) 36-58. [Resume]

53. Machen, op. cit., at pp. 172-73. [Resume]

54. A similar argument was made against the Nineteenth (Women's Suffrage) Amendment by William Marbury, "The Nineteenth Amendment and After," Virginia Law Review, 7 (1920) 1-29. [Resume]

55. Machen, op. cit., at p. 178. [Resume]

56. Machen, ibid., at p. 178. [Resume]

57. Machen, ibid., at p. 192. [Resume]

58. A more detailed discussion of what sorts of inconsistency can and cannot be tolerated in law may be found in Section 21.B. [Resume]

59. I owe this idea to H.W.R. Wade, "The Legal Basis of Sovereignty," Cambridge Law Journal, [no vol. number] (1955) 172-97, at pp. 187-88, where he argues that the only common law rule that the English parliament cannot repeal is that what parliament enacts is law. He quotes Salmond on Jurisprudence, Sweet & Maxwell, 11th ed. by Glanville Williams, 1957, at p. 155:

But whence comes this rule that Acts of Parliament have the force of law? This is legally ultimate; its source is historical only, not legal....It is the law because it is the law, and for no other reason that it is possible for the law itself to take notice of. No statute can confer this power upon Parliament, for this would be to assume or act on the very power that is to be conferred.

Wade calls the rule "the ultimate political fact upon which the whole system of legislation depends," op. cit. at 188 (emphasis added). It was this idea of Wade's of which Hart was thinking when he denied that the ultimate rule of recognition was "just 'political fact'". H.L.A. Hart, The Concept of Law, Oxford University Press, 1961, at pp. 108, 247. See Section 7.

Note that Salmond (Glanville Williams) and Wade presuppose that legal authorities cannot become effective by a self-justifying enactment. This is supposed to be as self-evident as the fallaciousness of begging the question in argument. In this sense it is another version of Ross's presumption that what is logically impossible must be legally impossible as well. It is no better supported as a tenet of law than Ross's own version of the presumption. [Resume]

60. James Cargile, "On Omnipotence," Nous, 1 (1967) 210-05, at pp. 202-03. On other acts beyond the power of a deity that should not count against its omnipotence, see Richard Swineburne, "Omnipotence," American Philosphical Quarterly, 10 (1973) 231-37, passim. [Resume]

61. Ronald M. Dworkin, "The Model of Rules," University of Chicago Law Review, 35 (1967) 14-46. However I disagree with Dworkin that Hart is one who has tried to draw this line rigidly. See Section 6.A, and Section 7.A. The alegal, social nature of acceptance precludes any complete separation of legal and moral standards. Hart, op. cit., at pp. 165ff. [Resume]

62. See Orfield, op. cit., at pp. 9, 53, 78f, 156. [Resume]

63. Indeed, Hart may be criticized for obliterating the distinction between legitimate and illegitimate government to the extent that he allows equal validity to all outcomes of all procedures validated by the rule of recognition. The distinction is restored if acceptance can overrule the rule of recognition at every stage. This is the chief modification of Hart's theory that I employ. [Resume]

64. Hart argues that unjust laws are not for that reason alone invalid, op. cit., at pp. 201-07. [Resume]

65. I do not mean to imply that the difficulty and fairness of amendment and the legitimacy of a legal system are inversely proportional all the way up and all the way down each scale, although I deliberately conceive legitimacy as a matter of degree. Nor do I mean to imply that the difficulty and fairness of amendment is the only, or even the primary, determinant of legitimacy. However a more complete account of legitimacy is considerably beyond the present topic. [Resume]

66. The issues of the continuing omnipotence of acceptance and the contingency of validation will be taken up again and in more detail in Section 21. [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.