Paradox of Self-Amendment by Peter Suber Most people who hear the paradox of self-amendment described respond that the old and new amending clauses need not be valid at the same time, and hence do not "contradict" each other. Section 10 examines this solution in detail and finds it wanting. Section 11 criticizes the adequacy of two other responses to the paradox, one holding that self-embracing omnipotence in the amending clause would avoid the paradox, and the other holding that specific authorization for self-amendment in the amending clause would avoid the paradox. The result of Part One (Sections 1-11) is that the paradox of self-amendment is difficult to banish. It is not dissolved for any existing formal logic or for the inference model of law by any of the usual objections or stratagems. In that sense Alf Ross is vindicated. However, if we abandon (the exclusivity of) formal logics in the interpretation of law, or the inference model of legal change and validity, or both, then the paradox does not arise. Appeal to Hart's acceptance model is one way to do this, but not the only way. Insofar as the acceptance model is plausible, or more plausible than the inference model, Alf Ross is repudiated. Part Two, especially Sections 12-19, turns aside to examine the same issues as Part One with regard to other, "unofficial" methods of amending the U.S. constitution. While Sections 3-11 somewhat single-mindedly focus on the legal and logical problems of using an explicit amending clause against itself, Sections 12-19 are more broadly concerned with unofficial methods of amending the constitution and their reflexive applications. Section 12 shows why it is important to the discussion to bring in these unofficial methods of amendment. In order to negate some salient assumptions of Part One, it examines cases of self-amendment in which the amending clause is not supreme and when the new form of the clause is not inconsistent with the old form. To do so it introduces several models of negation and contradiction for law, without which the discussion of whether two laws "contradict" one another remains vague and imprecise. Section 13 explores the effectiveness of using one amending clause against a second, and the second against the first, and so on, until the desired content is reached. This "see-saw" method may or may not be able to reach any desired content from any initial position, but it does manage to avoid strict self-amendment. Section 14 treats the amendment and repeal of constitutional rules that expire according to a clause stipulating their expiration date. Section 15 covers the kind of amendment that occurs when a provision is interpreted and reinterpreted by courts. Section 16 deals with the implied repeal of old rules by newly enacted rules that contradict them. Section 17 explores the special case of this in which the new law is a treaty and the old law which it contradicts is some provision of the constitution. Section 18 asks how the right to revolt, which is curiously explicit in many state constitutions, differs from the amending power. Section 19 argues that amendment by desuetude, or obsolescence, seems to occur in the United States although it is officially denied. For each of the unofficial methods of amendment in Sections 13-19, the question of self-application is raised and historical examples (if any) are discussed. As a result, the book is a fairly complete compendium of the constitutional law relating to constitutional amendment, quite apart from its discussion of paradox. Section 20 is a collection of reflexivities and paradoxes in law apart from self-amendment. It is a haphazard and personal collection that belongs here only to broaden the scope of this, the first book-length study of reflexivity and paradox in law. I know it is not complete because I have had to restrain myself and omit many variations in order to keep the section from swamping the rest of the book. Section 21 is a conclusion. It looks back over Part One to summarize the argument, and looks ahead to the oddities and difficulties facing the position I have taken. The summary (21.A) can be read before or independently of the rest of the book, or afterwards to bring the many threads together. In the more exploratory sections I offer an "historicized" modal logic for law that can account for the peculiarities revealed in the text while dispensing with immutable rules. Three appendices round out the inquiry. The first discusses the unavailing attempts at direct self-amendment at the federal level. The second surveys the many successful cases of direct self-amendment among the states. This is the primary documentation for the claim that self- amendment is commonplace even if it is logically "impossible". The third appendix presents a game of self-amendment. Notes on the text Citations to cases, statutes, constitutions and other forms of law are in standard legal form. Citations to books and articles, however, are in the clearer and more complete format of the humanities. Notes are gathered at the end of each section. I am aware of the maxim that nothing important should be relegated to a note. Nevertheless, my notes are sometimes expansive and discursive, and often contain much more than a bibliographic citation. My maxim was to put any matter in a note that would interrupt the continuity of the main exposition and that would not be the topic of the main exposition elsewhere in the book. The research on which my inquiry and bibliography are based is thorough up to 1982 when the first draft of this book was finished. After 1982 it is sporadic, as I began to live and work far from a large law library. The only abbreviation I use with regularity is "AC", for "amending clause". vi