Nine Perspectives on Living Orignalism

AND VAGUE LANGUAGE? Today most conservative originalists agree with me that we are not bound by original expected applications, only by original meaning. But the way they cash out original meaning often leads them to model it fairly closely on original expected applications—which includes not only how specific cases would be decided, but also how people at the time of adoption would have articulated the relevant constitutional principles. John McGinnis and Michael Rappaport’s article in this symposium is a good example of how this conflation occurs. McGinnis and Rappaport accuse me of committing what they call the “abstract meaning fallacy,” that is, “conclud[ing] that possibly abstract language has an abstract meaning without sufficiently considering and weighing the alternative possibilities.” When we look at their article more closely, however, we will discover that there is no fallacy in the sense of an error of logic. Rather, there is a disagreement between us about how to understand constitutional language. Behind that disagreement, in turn, is a deeper disagreement about the purposes of a constitution, and the sources of constitutional legitimacy. This becomes obvious when McGinnis and Rappaport restate the fallacy later in their article: “The fallacy involves an inference that a constitutional provision has an abstract original meaning that operates to delegate decision-making authority to future decision makers because the provision employs what seems to be abstract language.” That is, the fallacy involves adopting 9. See BALKIN, supra note 1, at 7. 10. John O. McGinnis & Michael B. Rappaport, The Abstract Meaning Fallacy, 2012 U. ILL. L. REV. 737, 739. 11. Id. at 741. BALKIN.DOCX (DO NOT DELETE) 4/25/2012 2:14 PM No. 3] NINE PERSPECTIVES ON LIVING ORIGINALISM 823 the very theory of constitutions and constitutional language that I argue for in Living Originalism! McGinnis and Rappaport agree with me that some terms in a constitution are generally recognized terms of art. An example would be “Letters of Marque and Reprisal” in Article I, Section 8. We also agree that historical research may be necessary to determine whether the Constitution refers to generally recognized terms of art, and whether language that might appear abstract to us actually embraces a narrower or more specific concept. Of course, historical inquiry might have the opposite effect. As discussed below, I argue that the original meaning of “commerce” is “intercourse,” which has a somewhat broader meaning than “trade,” because it includes transportation and communication networks. Finally, the three of us agree that historical research may also be necessary to determine whether a legal norm is a rule, standard, or principle. McGinnis and Rappaport point out that some texts that look like abstract principles might actually be abstract or general rules. Conversely, I have argued that some texts that look like rules, like the compact clause, might actually be principles or standards. We part company in how we understand abstract terms like “equal protection” and “freedom of speech.” I believe that such language must be worked out through constitutional constructions that may change over time. Although the original semantic meaning of the words “equal protection of the laws” remains the same as it was in 1868, how we apply the guarantee may change. By contrast, McGinnis and Rappaport believe that abstract terms, like “cruel and unusual punishments,” “freedom of speech,” and “equal protection of the laws,” have a legal meaning that is also their original meaning. They believe that we should treat these terms in much the same way as “Letters of Marque and Reprisal.” McGinnis and Rappaport’s concept of original legal meanings is really just another name for the initial constructions offered by people in the adopters’ generation. McGinnis and Rappaport, however, reject the distinction between interpretation and construction, and therefore believe that these contemporaneous constructions are part of original meaning. McGinnis and Rappaport argue that original meaning includes what they call original interpretive methods, and, as we will see shortly, that theory in turn rests on a still deeper theory about how constitutions work and why originalism is the best method of interpretation for a constitution. McGinnis and Rappaport are not always very clear about what these original interpretive methods are—indeed, their interpretive theory 12. U.S. CONST. art. I, § 8, cl. 11. 13. See infra text accompanying notes 114–38. 14. McGinnis & Rappaport, supra note 10, at 747–51 (offering examples). 15. BALKIN, supra note 1, at 47–48, 349–50 n.12. BALKIN.DOCX (DO NOT DELETE) 4/25/2012 2:14 PM 824 UNIVERSITY OF ILLINOIS LAW REVIEW [Vol. 2012 needs to be fleshed out in considerably more detail. Nevertheless, from the present article and from their previous writings, we can gather that original interpretive methods include the canons of statutory construction inherited from the common law as well as the common lawyer’s toolkit of arguments and methods as they existed at the time of adoption for each particular part of the Constitution. That would be 1789 for the original Constitution, 1791 for the Bill of Rights, 1868 for the Fourteenth Amendment, 1961 for the Twenty-Fourth Amendment, and so on. Now many of these methods are the same as the ones we have today. To be sure, there may be a few differences: for example, in 1787, arguments from legislative history or original intentions were not generally employed. Instead, lawyers generally looked to the intent or purposes of a statute through an inspection of its text. But the familiar categories of argument and the canons of construction are remarkably similar to the ones we use today. In Living Originalism, I argue that lawyers can and should use all of the traditional resources of lawyers both in ascertaining original meaning and in creating constitutional constructions that implement original meaning. So today, lawyers may make structural arguments, inspect dictionary definitions, invoke traditional rules of statutory construction, look to the history of previous readings and interpretations, build on previous precedents (both judicial and nonjudicial), make appeals to national ethos, and take into account prudence and consequences. All of these tools were available to common-law lawyers at the time of the Founding; indeed, as Philip Bobbitt has explained, the reason why we use these tools today is that they were inherited from the common law. When contemporary lawyers use these traditional tools, however, they may develop constructions very different from those the adopters would have expected or desired. One reason is that contemporary lawyers may draw on precedents and constructions throughout history, not merely those contemporaneous with ratification. Another reason is that history will often look different to us as we move forward in time. So lawyers in 16. See, e.g., John O. McGinnis & Michael B. Rappaport, Original Methods Originalism: A New Theory of Interpretation and the Case Against Construction, 103 NW. U. L. REV. 751 (2009). 17. H. Jefferson Powell, The Original Understanding of Original Intent, 98 HARV. L. REV. 885, 897–98 (1985). 18. McGinnis and Rappaport might disagree with this last statement and insist that we do not use similar interpretive methods today. I would respectfully disagree—in my view, American lawyers are still very much in the common-law tradition and, with only a few exceptions, still use the same modalities and the same canons. But this merely suggests that there can be good-faith disputes about how exactly to characterize these interpretive methods and the proper level of generality to describe them. As a result, disputes about original interpretive methods will tend to replicate disputes about constitutional meaning at a different level of inquiry. 19. See BALKIN, supra note 1, at 17, 46, 89, 129, 205, 256–57, 333, 341–42 (explaining that interpreters should use all of the traditional modalities of constitutional argument). 20. See PHILIP BOBBITT, CONSTITUTIONAL FATE: THEORY OF THE CONSTITUTION 3–8 (1982); PHILIP BOBBITT, CONSTITUTIONAL INTERPRETATION 8, 12–13, 24 (1991). BALKIN.DOCX (DO NOT DELETE) 4/25/2012 2:14 PM No. 3] NINE PERSPECTIVES ON LIVING ORIGINALISM 825 2003 might conclude that gays deserve legal protection under the Fourteenth Amendment even though lawyers in 1868 would not. McGinnis and Rappaport’s theory of original-methods originalism, however, makes a different and much stronger claim than this. By “original methods,” they do not simply mean that we should use the same methods and modalities of legal argument that well-trained lawyers at the time of the adoption of the text would have used. That in itself is not a very significant requirement, for these methods are part of our contemporary legal culture. What they mean is that we should use these methods in the same way that lawyers at the time of adoption would have used them to generate legal meanings. That is, the legal meaning of “Cruel and Unusual Punishments” is the construction that a hypothetical well-trained lawyer in 1791, using the interpretive methods of that time, would have produced. The goal is to retrace such a lawyer’s steps and consider the same evidence that well-trained lawyers would have employed at the time of adoption in 1791 (or 1868) to generate legal meaning. As a result, the contemporaneous legal opinions of well-trained lawyers at the time of adoption, while not necessarily conclusive in all respects, are very strong evidence of original legal meaning. That is why McGinnis and Rappaport can argue that the mere fact that constitutional language appears vague or abstract today is irrelevant. Vagueness arises because of uncertainty about “whether or not a term extends to a proposed application.” But lawyers in 1791 might not have been uncertain. Moreover, the interpretive rules at the time of the Founding required that where language is vague or uncertain interpreters should pick the meaning that is most likely given “the relevant originalist evidence—evidence based on text, structure, history, and intent—and select the interpretation that was supported more strongly by the evidence.” Thus, if lawyers at the time of adoption would have understood constitutional language as more likely to reach some applications but not others, “there is no legal ambiguity or vagueness, regardless of whether there is vagueness or ambiguity in the ordinary language.” 21. McGinnis & Rappaport, supra note 10, at 746 (“[T]he original meaning approach requires that [ambiguity or vagueness] be resolved based on the interpretive rules that existed at the time. This is how the reasonable and knowledgeable interpreter, employing the original meaning approach, would resolve the question.”) (footnote omitted); see id. at 747–48, 764 (arguing that the due process clause has the meaning that was historically settled at the time of adoption); id. at 750 (arguing for “pin[ning] down” meaning by appealing to “the meanings developed through previous legal traditions and processes”); id. at 747 n.36 (“[W]e believe that the original-methods approach best captures the actual meaning of the Constitution, because the enactors would have understood the Constitution against the background of the relevant interpretive rules of the time.”); see also id. at 748 nn.41–43 (listing examples of historically concretized legal meanings). 22. McGinnis & Rappaport, supra note 16, at 774. 23. Id. 24. Id. See also McGinnis & Rappaport, supra note 10, at 748 nn.41–43 (offering examples of historically informed interpretations that ask what the legal meaning of seemingly vague and abstract clauses would have been at the time of adoption). BALKIN.DOCX (DO NOT DELETE) 4/25/2012 2:14 PM 826 UNIVERSITY OF ILLINOIS LAW REVIEW [Vol. 2012 This approach leads McGinnis and Rappaport to equate original meaning to something very close to original expected application. Original expected application, after all, concerns how adopters expected the constitution’s language would be applied. And McGinnis and Rappaport are interested in whether a constitutional term extends to a proposed application. Lawyers at the time of adoption might have had a range of different views on these questions, but some were more likely than others, and the most likely interpretation is precisely what we should be looking for. Modern constructions outside this range are not consistent with legal meaning. Thus, although McGinnis and Rappaport agree with me that original meaning is not the same thing as original expected applications, their model of original interpretive methods sticks fairly closely to original expected application, because it sticks closely to the views and expectations of adoption-era lawyers. Under McGinnis and Rappaport’s model, future interpreters may not create new constructions that vary from the legal meaning that welltrained lawyers in 1791 (or 1868) would have understood and recognized. Although not all lawyers would have agreed on all questions, we may not use that limited range of disagreement to impose interpretations that no well-educated lawyer at the time of adoption would have thought reasonable. Thus, the set of reasonable constructions (or legal meanings, in McGinnis and Rappaport’s parlance) available at the time of adoption more or less fixes the scope of permissible interpretation today. To be sure, changes in technology or factual context might be relevant to how we apply original legal meaning—for example, in the case of railroads, airplanes, or electronic media. Nevertheless, McGinnis and Rappaport insist that changes in moral judgments from time of adoption to the present day should not affect the original legal meaning—for example, changed moral views should not affect whether women or gays are protected by the Fourteenth Amendment or whether the right to use contraceptives is a privilege or immunity of citizens of the United States. As McGinnis and Rappaport explain, we are trying to understand the adopters’ judgments about the proper boundaries of constitutional government, and if we treat too many of the adopters’ moral judgments as mistaken, we have likely failed to grasp their legal meaning. Thus, McGinnis and Rappaport reject Living Originalism’s central distinction between interpretation and construction. They do not believe that anybody in 1787 thought that vague or abstract language either invited or required construction, either by the contemporaneous generation or by future generations. Rather, well-trained lawyers and judges would simply decide what the most likely reading was; this practice was 25. See McGinnis & Rappaport, supra note 16, at 774 (advocating narrow definitions of vagueness and ambiguity). 26. See John O. McGinnis & Michael Rappaport, Original Interpretive Principles As the Core of Originalism, 24 CONST. COMMENT. 371, 378–81 (2007). BALKIN.DOCX (DO NOT DELETE) 4/25/2012 2:14 PM No. 3] NINE PERSPECTIVES ON LIVING ORIGINALISM 827 legal interpretation, and the legal meanings generated by this practice, as part of original meaning, are still binding on us today. For similar reasons, McGinnis and Rappaport reject my view that the use of standards and principles in a constitution delegates future construction to later generations. If well-trained lawyers and judges in 1791 (or 1868) using the generally accepted methods of the era would have generated narrower and more precise readings of the Constitution than we do today, those are the legal meanings—and hence the original public meanings—of this language. I do not accept this model of original-meaning originalism for reasons I describe at length in Living Originalism. Some of my reasons are historical, and some are theoretical. First, relying on work by Larry Kramer, Caleb Nelson, and Saul Cornell, I argue that there was no consensus about the correct way to interpret the Constitution at the time of the Founding, in part because the idea of a federal constitution was so new, and analogies to trusts, treaties, contracts, and statutes pointed in very different directions. There were also disputes about the role of professional legal knowledge versus popular constitutional or plain meaning approaches. The very assumptions that McGinnis and Rappaport make about the primacy of lawyers’ interpretations in fixing constitutional meaning were hotly contested at the Founding and were not resolved by ratification. Second and relatedly, original legal methods are not the same thing as original meanings; not everyone who ratified the Constitution was a lawyer, and not everyone who participated in the ratification debates assumed that the Constitution should be interpreted according to lawyers’ views. Some believed, to the contrary, that the Constitution belonged to the public as a whole, including and especially those untutored in law, and that “The People Themselves” would enforce the Constitution through politics. Third, and perhaps most important, even if there was consensus about interpretive methods, it does not follow that to accept these methods we must also accept how lawyers in 1791 would have employed them. That is, we should not confuse original methods with the original applications of original methods. For example, take a key Founding-era principle offered by McGinnis and Rappaport: when in doubt, we should 27. See McGinnis & Rappaport, supra note 16, at 774–76 (arguing that vagueness and ambiguity is rare because judges are required to choose the most probable meaning). 28. See id. 29. See Saul Cornell, The People’s Constitution vs. The Lawyers’ Constitution: Popular Constitutionalism and the Original Debate over Originalism, 23 YALE J.L. & HUMAN. 295 (2011); Larry Kramer, Two (More) Problems with Originalism, 31 HARV. J.L. & PUB. POL’Y 907, 912–13 (2008); Caleb Nelson, Originalism and Interpretive Conventions, 70 U. CHI. L. REV. 519, 555–56, 561, 571–73 (2003). 30. See Cornell, supra note 29, at 304–20. 31. LARRY D. KRAMER, THE PEOPLE THEMSELVES: POPULAR CONSTITUTIONALISM AND JUDICIAL REVIEW 6–7 (2004). BALKIN.DOCX (DO NOT DELETE) 4/25/2012 2:14 PM 828 UNIVERSITY OF ILLINOIS LAW REVIEW [Vol. 2012 choose the most probable meaning of abstract and vague language. We might employ the same principle today. But the evidence we would look to and the reasons we would offer to settle on the most probable meaning of a vague text like “freedom of speech” or “due process of law” today might be very different than in 1791 because we have lived through 220 more years of history and experience. Bernadette Meyler has pointed out that the incorporation of common-law concepts and common-law terms into a constitution might also include a common-law process of future development of those terms and concepts. The differences between McGinnis and Rappaport and me are not primarily disputes about the philosophy of language. Nor are they disputes about the dictionary definitions of “original” or “meaning.” Rather, we disagree about how to cash out the idea of “original meaning” in practice, and the most sensible way to read abstract language in an ancient constitution that we hope will endure for centuries. Our disagreements, in short, are disagreements about how written constitutions work and what gives constitutions their contemporary democratic legitimacy. These questions matter greatly in the case of the U.S. Constitution, which is not only one of the oldest written constitutions in the world, but is also one of the most difficult to amend. “Meaning” is a capacious concept, and indeed, it has many different meanings, including semantic content, purposes, intentions, practical entailments, and cultural associations. Conceived most broadly, “meaning” includes a vast array of cultural associations, traditions, conventions, and background assumptions. Any version of “original meaning” in legal interpretation must inevitably carve out a subset of these cultural meanings and treat this portion as remaining in legal force over time. Therefore, any version of “original meaning” will necessarily be anachronistic, because it will insert some portion of the vast array of past cultural meanings into a contemporary setting without bringing the other parts along with it. Inevitably, then, we face a choice in the present about what aspects of cultural meaning should constitute “original meaning” for purposes of constitutional interpretation. There is no natural and value-free way to make this selection. It cannot be settled by the meaning of “meaning,” much less the meaning of “original.” It is a choice that is informed by the purposes of a constitution and the promotion of the kind of legitimacy 32. McGinnis & Rappaport, supra note 15, at 773–75. 33. See Bernadette Meyler, Towards a Common Law Originalism, 59 STAN. L. REV. 551, 566–67, 593–95, 600 (2006). 34. See Donald S. Lutz, Toward a Theory of Constitutional Amendment, in RESPONDING TO IMPERFECTION: THE THEORY AND PRACTICE OF CONSTITUTIONAL AMENDMENT 237, 261 tbl.11, 265– 67 (Sanford Levinson ed., 1995) (noting that in comparison with other countries, “the U.S. Constitution is unusually, and probably excessively, difficult to amend,” and arguing that the strategies of judicial revision are the likely consequence of such a constitution). BALKIN.DOCX (DO NOT DELETE) 4/25/2012 2:14 PM No. 3] NINE PERSPECTIVES ON LIVING ORIGINALISM 829 (democratic, social, procedural, or moral) we want our government to have. Movement conservatives first turned to original-meaning originalism in the 1980s to avoid difficulties with theories of original intention and original understanding. They argued that what binds contemporary interpreters should be original public meaning, not psychological states and intentions. In fact, the interpretation of a legal text always involves an ascription of purpose or intention to an author. But the point is that this ascription of purpose or intention is invariably selective. People identify some aspects of what the adopters sought, intended, assumed, or desired as central to the goals of interpretation—that is, the interpretive enterprise they are currently engaged in—and treat other aspects as optional or even as irrelevant. When we engage in interpretation, we do not seek fidelity to all aspects of an author’s intentions or meanings. Rather, we seek fidelity to the relevant aspects of intentions or meanings, given how we understand the nature and purpose of the interpretive practice we are engaged in. Our practices of interpretation are therefore always anachronistic and selective because we are interpreting for a (present-day) purpose. In Living Originalism, I focus on original semantic meaning (including generally recognized terms of art) and the adopters’ choice of basic technologies of linguistic constraint: rules, standards, principles, and silences. The remaining aspects of cultural meaning I treat as resources for construction, but not as part of the framework. Why do I do this? My (selective) account of original meaning flows from my view about how written constitutions work over long periods of time and what makes them legitimate for generations long after their adoption. Constitutions are basic plans for politics that have to be carried out over time by many different generations, who may not share the adopters’ cultural presuppositions and worldview. Not everything can be settled at the outset; therefore adopters must put their trust in later generations to carry out the plan and adapt it to new circumstances. That is, adopters need the contributions of later generations to keep the plan going. Later generations are not simply slavish followers of the adopters’ will; they have a crucial role to play in building out the plan and helping it succeed. The intelligence and creativity of later generations is just as important to the success of the plan over time as the initial insight of the adopters. Constitutions simultaneously constrain and enable political action by participants. But no constitution can be designed that is so perfect that it will succeed without the judgment and wisdom of later generations. Adopters, recognizing this, choose different technologies of constraint based on how they wish to constrain or enable future generations. To use Scott Shapiro’s phrase, plans for politics involve an “economy of BALKIN.DOCX (DO NOT DELETE) 4/25/2012 2:14 PM 830 UNIVERSITY OF ILLINOIS LAW REVIEW [Vol. 2012 trust” (and distrust). To understand the contributions that different generations make to the plan, we must understand this economy. Therefore, if we accept the adopters’ plan as our plan and wish to carry it forward into the future, the adopters’ choice of rules, standards, principles, and silences should matter to us. Rules minimize practical judgment by later generations; standards and principles channel judgment but also require it. Silences leave matters to be determined later on. For this reason, we should take seriously the use of abstract or vague language in a constitution, especially when adopters could have used more precise language or could have specified a historical standard—for example, they might have protected “freedom of speech as understood at the time of adoption of this Constitution.” My account of original meaning assumes that adopters recognize that a great deal will have to be worked out through construction and that their choice of rules, standards, and principles is deliberate. It ascribes this understanding and these purposes to the adopters. They are creating a constitution not only for their time, but for a later time they cannot know much about. This economy of freedom and constraint, trust and distrust is important for a second reason. The democratic legitimacy of the Constitution is not established at the moment of adoption. Rather, it derives from multiple sources. The first source of democratic legitimacy is the act of adoption or amendment. The second arises from the processes of constitutional construction over time. Constitutions cannot maintain their democratic legitimacy without contributions from multiple generations. For the Constitution to have democratic legitimacy in each generation it must succeed as “our law”: people must understand it as their accomplishment and as something that belongs to them. We make the Constitution “our law” by struggling over its meaning in constitutional politics, producing new constitutional constructions, and building on or revising older ones. Framework originalism argues that fidelity to the basic framework creates a space for constitutional construction, and that the framework allows—and requires—each generation to do its part in making the Constitution its own Constitution. For many conservative originalists, my account of original meaning is too barebones, and my theory of framework originalism leaves too much to be built out later on. Nevertheless, conservative originalists must also make a choice about what aspects of cultural meaning to carry forward from past to present as binding on us today, what economy of trust and distrust to ascribe to the Constitution, and what degree of anachronism to accept. Their choice in these matters cannot simply be one of definitional requirement; it does not flow from the meaning of “original meaning.” Whether consciously or not, their choice must be 35. See SCOTT J. SHAPIRO, LEGALITY 331–52 (2011). BALKIN.DOCX (DO NOT DELETE) 4/25/2012 2:14 PM No. 3] NINE PERSPECTIVES ON LIVING ORIGINALISM 831 based on background assumptions about the purposes of a constitution, and how its legitimacy is maintained over time. These considerations apply equally to McGinnis and Rappaport’s theory of original legal methods. Although McGinnis and Rappaport urge lawyers to stick to the ordinary work of legal argument, their version of originalism is nevertheless backed by a distinctive constitutional and political theory. What is that theory? They argue that originalism is the best theory of interpretation because it is likely to produce the best consequences over time. Originalism produces the best consequences because the 1787 Constitution was adopted by a supermajority vote—nine of thirteen state conventions had to adopt the Constitution to ratify it—and because ever since then, amendments under Article V have required two-thirds votes of both houses of Congress and ratification by three-quarters of the state legislatures. Supermajority ratification rules produce rules with superior consequences because adopters are naturally risk averse. They will not vote to ratify a constitution or a constitutional amendment unless they are quite sure that the constitution or the amendment will produce good results in the future, as it will be difficult to change the provision later on. Moreover, adopters vote for laws that will affect their descendants in very different situations, so there is a sort of temporal “veil of ignorance” that leads them to choose optimal rules that will benefit future inhabitants even if adopters do not know what conditions will obtain later on. 36. John O. McGinnis & Michael B. Rappaport, Originalism and the Good Constitution, 98 GEO. L.J. 1693, 1695 (2010). 37. Id. at 1735; McGinnis & Rappaport, supra note 10, at 774–75 (“Even in the rare instances where groups disagree and nevertheless would like a provision to be put in the Constitution, a delegation to the future would not represent a likely rational compromise because it would increase risks for everyone involved.”). 38. McGinnis & Rappaport, supra note 36, at 1697, 1702, 1708; McGinnis & Rappaport, supra note 10 at 779 (noting that veil of ignorance also helps to protect minority rights). The central role that supermajority rules play in McGinnis and Rappaport’s theory of originalism leads to a number of puzzles. If a three-quarters majority of states makes for better rules than a simple majority, why not require unanimity, as John Rawls does in his original position? In fact, the Articles of Confederation actually required unanimity for both adoption and amendment. Does this mean that its rules were ex ante more likely to be optimal than the procedures used to adopt the 1787 Constitution? Presumably, McGinnis and Rappaport would agree that there is a “sweet spot” for supermajority rules: just enough difficulty to force optimal content, but not so much that the threat of holdouts wrecks the process or forces morally undesirable compromises—for example, rules protecting slavery. Because McGinnis and Rappaport do not focus on this particular question, however, they do not tell us how this sweet spot is properly measured or whether it might change over time. Perhaps equally important, given the many different kinds of supermajority rules that might have been chosen, they also do not explain why the Framers of the U.S. Constitution miraculously happened upon just the right balance in 1787 and why this optimal balance continues two centuries later. In fact, there is reason to doubt that the adopters succeeded in picking supermajority rules best calculated to produce good consequences over time. The U.S. Constitution is among the most difficult to amend of any constitution in the world; if its procedures are optimal, this would suggest that the rest of the world’s constitutions—and those of the fifty states—are suboptimal. More likely, given the experience of subsequent history, the U.S. Constitution is well beyond the magical sweet spot for optimal BALKIN.DOCX (DO NOT DELETE) 4/25/2012 2:14 PM 832 UNIVERSITY OF ILLINOIS LAW REVIEW [Vol. 2012 In order to enjoy the benefits of supermajority ratification procedures, it is important that constitutional provisions be relatively clear and determinate; first, so that people can know in advance what they are voting for; and second, so that later generations cannot vary them, except through subsequent amendments that require supermajority support. Although in theory adopters could delegate to future generations the role of working out the details later on—which is the central assumption of framework originalism—McGinnis and Rappaport believe that it is very unlikely they would do so because rational adopters are risk averse. Rational adopters would not vote for provisions that are vague and abstract because they would not know how these provisions will be applied in the future. Moreover, delegation to future generations would not appeal to risk averse adopters because constructions by judges and the political branches will not be disciplined by supermajority procedures, and so their work will likely be inferior to that of adopters constrained by supermajority rules. In particular, McGinnis and Rappaport believe that risk averse adopters would clearly reject the idea of “judicial updating” by unelected judges. Judges do not even represent majorities; therefore there is no reason to think that the constructions they produce would have good consequences. Viewed from an ex ante perspective, the legal meaning of “equal protection” in 1868 is likely to be superior to anything the Warren Court or Congress in the 1960s could come up with. Similarly, the lawyerly consensus about the meaning of the First Amendment in 1791 is likely to be superior to judicial development of free speech jurisprudence a century and a half later, because the former required supermajority consent, and the latter required only five votes on the U.S. Supreme Court. content. It is probably far too difficult to amend, which means that its rules are likely to be suboptimal. Equally likely, its amendment procedures are likely to lead to compromises that kick difficult and contentious issues down the road for others to solve. This is precisely what McGinnis and Rappaport deny. Moreover, the methods for adoption and amendment of the U.S. Constitution have varied. The 1787 Constitution was adopted by simple majority votes of three-quarters of state ratifying conventions, while subsequent amendments have been adopted by two-thirds vote of Congress plus threequarters of the states (which, except for Nebraska, has generally required ratification by two houses of the state legislature). Thus, the rules for amendment are considerably more onerous than the rules for initial ratification. Does this mean that rules appearing in the 1787 Constitution are ex ante more likely to be inferior to those in the Bill of Rights, or vice versa? Moreover, the Fourteenth Amendment was adopted by a rump Congress—Republicans refused to seat returning southern congressmen and senators—and southern states were not readmitted until they agreed to ratify the Fourteenth Amendment. Does this mean that judged ex ante the content of the Fourteenth Amendment is likely to be inferior to the content of the 1787 Constitution? 39. McGinnis & Rappaport, supra note 36, at 1735; John O. McGinnis & Michael Rappaport, The Condorcet Case for Supermajority Rules, 16 SUP. CT. ECON. REV. 67, 94, 112 (2008). 40. McGinnis & Rappaport, supra note 36, at 1737–43; McGinnis & Rappaport, supra note 10, at 771–74; see also McGinnis & Rappaport, supra note 16, at 775–76 (arguing that rational adopters would require resolution of ambiguity and vagueness to prevent discretion to future interpreters). BALKIN.DOCX (DO NOT DELETE) 4/25/2012 2:14 PM No. 3] NINE PERSPECTIVES ON LIVING ORIGINALISM 833 Moreover, McGinnis and Rappaport’s point is that supermajority procedures that produce better consequences in 1791 will continue to provide these benefits in 2012, because supermajority rules force adopters to adopt rules that will be optimal for the future, and if this proves not to be the case, a contemporary supermajority can use an Article V amendment to make corrections. Judicial “updating” by contemporary judges, by contrast, loses the discipline—and therefore the benefits—of supermajority procedures. Accordingly, McGinnis and Rappaport do not believe that the existence of vague or abstract principles in the Constitution delegates to future generations the job of engaging in constitutional construction. Instead, they believe that supermajority provisions produce “determinate principles about which there is a broad consensus” at the time of ratification. They argue that political or social groups would rarely agree to place provisions in a constitution when there is significant disagreement about their content. Doing so is simply not rational, because “[t]here is an alternative when groups cannot agree: leave out controversial principles and entrench only consensus ones.” It follows that the consensus view as of 1787 (or 1791 or 1868) is what matters for interpretation, because that is what opposed groups will converge on—otherwise the provision would not even make it into the Constitution—and this consensus view is likely to produce the best results not only for the current generation, but for future generations as well. Now we can understand why McGinnis and Rappaport believe that what they call “legal meaning” must be part of original meaning. Their background theory of supermajority rules explains and justifies their choices about what aspects of the past count as “original meaning.” What makes a constitution valuable for them is its ability to generate good consequences over time. Rules created by supermajorities must be relatively determinate because only relatively determinate rules can be assured of producing good consequences; vague and abstract texts will be too easy for later generations (and later judges) to shirk, thus debilitating the Constitution’s beneficial effects. Because adopters are rational and would seek to achieve good consequences over time, we must presume that they are risk averse and that they will agree only to relatively determinate rules. Although some constitutional language may seem abstract or vague to us today, this is an illusion. The language was not unclear or vague to risk averse adopters because they relied on the interpretations of welltrained lawyers at the time of ratification, as well as the canon that when language is vague or unclear we should always choose the interpretation that was most likely the one intended. 41. McGinnis & Rappaport, supra note 36, at 1722–23, 1737–48. 42. McGinnis & Rappaport, supra note 10, at 772. 43. Id. at 774. BALKIN.DOCX (DO NOT DELETE) 4/25/2012 2:14 PM 834 UNIVERSITY OF ILLINOIS LAW REVIEW [Vol. 2012 McGinnis and Rappaport, in short, argue that the Constitution’s legitimacy comes from the fact that supermajority rules have produced a “good Constitution” in which ordinary politics can operate. Ordinary politics, however, cannot change the Constitution’s legal meaning; nor can “judicial updating” that is responsive to politics. Legitimate constitutional development can only occur through supermajority procedures that are calculated to produce superior results. This model of originalism is a species of “skyscraper originalism.” Skyscraper originalism sees the Constitution as a more or less finished project for democratic politics, although it is always subject to later Article V amendment. The Constitution provides a space in which democratic politics occurs. Democratic lawmaking within this space, however, is not constitutional construction. It is ordinary law that is constitutionally permissible given the boundaries of the Constitution as adopted. It cannot change the Constitution, any more than everyday activity within a building can change the building’s structure. Thus, no matter how much social mobilizations like the civil rights movement or the women’s movement may have affected American political culture, they cannot legitimately change the meaning of the Constitution except through Article V amendment. I do not accept this account of constitutional language for two reasons. First, I do not agree that because adopters are risk averse, they will use general or abstract language to articulate relatively concrete and consensus commitments. Especially where fundamental rights are concerned, constitutional adopters use vague and abstract language because they want to state basic commitments that they expect others to carry out in the future. They wish to channel and guide politics by placing basic values in the Constitution so that people can refer to them in subsequent political struggles. Written declarations of rights, even if phrased in abstract language, shape political struggles, and this, in turn, affects the law that is produced by politics. James Madison offered a similar explanation when he introduced the initial draft of the Bill of Rights before Congress. Previously in the Federalist Papers, and in correspondence with Thomas Jefferson, Madison suggested that bills of rights were mere “parchment barriers” that might have little effect. Declarations of vague and abstract rights would be easy to manipulate and would offer little protection against a determined majority. In his address to the First Congress Madison repeated this concern, but ascribed it to others: “It may be thought that all paper barriers against the power of the community are too weak to be worthy 44. McGinnis & Rappaport, supra note 36, at 1720–22. 45. BALKIN, supra note 1, at 21–23. 46. THE FEDERALIST NO. 48, at 308 (James Madison) (Henry Cabot Lodge ed., 1888); Letter from James Madison to Thomas Jefferson (Oct. 17, 1788), in 5 THE WRITINGS OF JAMES MADISON 269, 272 (Gaillard Hunt ed., 1904) (“Repeated violations of these parchment barriers have been committed by overbearing majorities in every State.”). BALKIN.DOCX (DO NOT DELETE) 4/25/2012 2:14 PM No. 3] NINE PERSPECTIVES ON LIVING ORIGINALISM 835 of attention. I am sensible they are not so strong as to satisfy gentlemen of every description who have seen and examined thoroughly the texture of such a defence.” Despite this objection, Madison argued, written declarations of rights “have a tendency to impress some degree of respect for them, to establish the public opinion in their favor, and rouse the attention of the whole community.” These social and cultural effects “may be one means to control the majority from those acts to which they might be otherwise inclined.” Similarly, in his correspondence with Jefferson, Madison noted that stating basic rights in a constitution might have a salutary influence on political culture: “The political truths declared in that solemn manner acquire by degrees the character of fundamental maxims of free Government, and as they become incorporated with the national sentiment, counteract the impulses of interest and passion.” Madison surely understood, as we do today, that abstract rights language invites multiple and contrasting interpretations among constitutional adopters. But using more concrete language might undermine the possibility of agreement, or it might lead to narrow protections of specific examples of a freedom that become irrelevant as time passes. Using abstract and vague language naturally produces disputes about the true meaning of rights guarantees as new policy questions emerge. It thrusts discussion of constitutional rights into politics; in doing so, it begins the processes of constitutional construction that actually protect these rights in practice. The First Amendment provides a good example. When the First Amendment was adopted in 1791, the standard legal view—the legal meaning, in McGinnis and Rappaport’s terms—was that the guarantee of freedom of the press banned prior restraints on publication but did not prevent subsequent punishments for libel or seditious advocacy. Nevertheless, the actual practices of freedom of expression during the period leading up to the adoption of the Bill of Rights were considerably more libertarian, and developing ideas of popular sovereignty—in contrast to parliamentary sovereignty—made it crucial for ordinary individuals to be able to criticize their government. In fact, as Akhil Amar points out, the 47. 1 ANNALS OF CONG. 455 (1789). 48. Id. 49. Id. 50. Letter from James Madison to Thomas Jefferson, supra note 46, at 273. 51. See 4 WILLIAM BLACKSTONE, COMMENTARIES *151–52. As Blackstone explained: The liberty of the press is indeed essential to the nature of a free state; but this consists in laying no previous restraints upon publications, and not in freedom from censure for criminal matter when published. Every freeman has an undoubted right to lay what sentiments he pleases before the public: to forbid this, is to destroy the freedom of the press: but if he publishes what is improper, mischievous, or illegal, he must take the consequence of his own temerity. Id. (emphasis omitted). Leonard Levy famously argued that this limited, Blackstonian interpretation was the official legal meaning of free expression at the time of the Founding, while acknowledging the views of critics that the “law in action” was considerably freer. See LEONARD W. LEVY, EMERGENCE OF A FREE PRESS (1985). BALKIN.DOCX (DO NOT DELETE) 4/25/2012 2:14 PM 836 UNIVERSITY OF ILLINOIS LAW REVIEW [Vol. 2012 very process of ratifying the Constitution required that ordinary citizens be able to criticize government officials and the existing form of government—the Articles of Confederation. The dispute over the Sedition Act of 1798 brought the traditional lawyer’s reading—defended by the dominant Federalist Party—into conflict with the views of an emerging political mobilization, the Jeffersonian Republican Party, which argued for a broader legal conception of free expression. Jeffersonians had criticized the government for its approach in the developing conflict with France. They thought Blackstone’s doctrine, in James Madison’s words, to be a “mockery” of free speech. Nevertheless, Federalist judges, relying on traditional concepts of free expression, shut down five of the six most influential Republican papers. McGinnis and Rappaport’s original legal methods approach argues that the legal meaning of the First Amendment today should rest on “determinate principles about which there is a broad consensus” in 1791. The problem is that there was no broad consensus for the Republican position in 1791, and certainly not in 1798: the two parties were at each other’s throats over this very issue. Given that the Federalists dominated national politics in 1791 when the amendment was adopted, the broad consensus required by McGinnis and Rappaport’s theory would have to give strong weight to their position. A demand for consensus about rights provisions in 1791 would seem to suggest a lowest-common denominator approach, something closer to Blackstone’s view. Because 52. See AKHIL REED AMAR, AMERICA’S UNWRITTEN CONSTITUTION (2012). 53. James Madison, Report on the Resolutions (Feb. 7, 1799), in 6 THE WRITINGS OF JAMES MADISON 341, 386 (Gaillard Hunt ed., 1906). 54. RON CHERNOW, ALEXANDER HAMILTON 575 (2004); Jeffery A. Smith, Prior Restraint: Original Intentions and Modern Interpretations, 28 WM. & MARY L. REV. 439, 459–60 (1987). On Federalist prosecutions of Republican speech, see JAMES MORTON SMITH, FREEDOM’S FETTERS: THE ALIEN AND SEDITION LAWS AND AMERICAN CIVIL LIBERTIES 256–270 (1966). The Federalist report on the bill in the House of Representatives likewise insisted that there was nothing unconstitutional about the proposed bill. It argued that the Sedition Act was consistent with the First Amendment because it did not impose a prior restraint, because it was merely declaratory of the common law (in fact, it was more speech protective than the common law because it allowed a defense of truth), and because states with free speech and press guarantees in their constitutions already punished seditious and malicious publications. See 5 ANNALS OF CONG. 2988–90 (1799). John Marshall, who by the late 1790s was one of the Federalist Party’s leading figures, believed that the Sedition Act was bad policy and unwise; nevertheless, he defended its constitutionality in an anonymous report for the Virginia Legislature that responded to Madison’s famous report. See JOHN MARSHALL, REPORT OF THE MINORITY ON THE VIRGINIA RESOLUTIONS (1798), reprinted in 5 THE FOUNDERS’ CONSTITUTION 136–39 (Philip B. Kurland & Ralph Lerner eds., 1987) (attributed to Marshall); Kurt T. Lash & Alicia Harrison, Minority Report: John Marshall and the Defense of the Alien and Sedition Acts, 68 OHIO ST. L.J. 435 (2007) (describing Marshall’s participation in the Sedition Act controversy). 55. McGinnis & Rappaport, supra note 10, at 772. 56. Indeed, this interpretation would satisfy two important criteria of McGinnis and Rappaport’s theory of original legal methods. It would have been the most likely meaning that would have generated a broad determinate consensus among both Federalists and their opponents, and it would have been the legal meaning that assumed the constitutionality of the relevant legislation and left disputes up to the political branches to resolve. See McGinnis & Rappaport, supra note 16, at 775. BALKIN.DOCX (DO NOT DELETE) 4/25/2012 2:14 PM No. 3] NINE PERSPECTIVES ON LIVING ORIGINALISM 837 McGinnis and Rappaport appear to treat the Sedition Act as an antiprecedent, rather than as an honored source of original legal meaning, this is probably not what they had in mind. Yet, I must confess that, given this example, I am not sure how to operationalize their theory of determinate principles that enjoy broad consensus at the time of adoption. In fact, it is very unlikely that even the outer boundaries of the consensus on freedom of expression in 1791 would be consistent with our modern conceptions of free speech, free press, and freedom of association. If McGinnis and Rappaport are really serious about deriving original meaning in the way they say they are, they have a lot of explaining to do if they wish to justify modern civil rights and civil liberties protections. It is worth noting that, like many conservative originalists, McGinnis and Rappaport have had to devise an elaborate theory of nonoriginalist precedent to make their theory compatible with the modern state. Because I do not share McGinnis and Rappaport’s theory, I draw a very different conclusion from this evidence. The abstract language of the First Amendment left unresolved differing views about the meaning of freedom of speech and press; these disputes would break out into the open later on in the 1790s, just as unresolved disputes about the scope of federal power would lead to political struggles over a national bank. Positions on free speech divided along party lines because they were wrapped up in the French foreign policy crisis, leading the Jeffersonians to champion libertarian ideas that were connected to the right of the people, as popular sovereigns, to criticize their elected representatives. The same crisis, it should be noted, also led Jefferson to offer creative views about state interposition that are out of the mainstream today. Here we can see Madison’s point about the Bill of Rights in practice: having a text in the Constitution that guaranteed the freedoms of speech, press, and petition created additional respect for these principles in politics. It gave Republicans something to rally around in their dispute with the Federalists, even if the Jeffersonian position on free speech was controversial. The Republican victory in 1800 and the Jeffersonian revolution led to new constructions of the First Amendment that were confirmed by executive and legislative practices, including Jefferson’s pardon of persons convicted under the Sedition Act. The fight over the Sedition Act is the beginning of the processes of constitutional construction that we know as living constitutionalism. When the Supreme Court decided New York Times v. Sullivan in 1964, it looked to the later Jeffersonian construction of free speech, and not to the views of Federalist lawyers in 1791, as the appropriate resource for 57. See McGinnis & Rappaport, supra note 10, at 767 n.123. 58. See John O. McGinnis & Michael B. Rappaport, Reconciling Originalism and Precedent, 103 NW. U. L. REV. 803 (2009). BALKIN.DOCX (DO NOT DELETE) 4/25/2012 2:14 PM 838 UNIVERSITY OF ILLINOIS LAW REVIEW [Vol. 2012 constitutional construction. At the same time, the Warren Court did not accept the Jeffersonian theory of interposition, a theory seized on by Southern critics of Brown v. Board of Education. The Sedition Act controversy shows how constitutional construction grows naturally out of ordinary politics. Political controversies of the moment lead to disputes about vague or abstract constitutional language; the resolution of these disputes becomes part of the Constitution-inpractice, and a new set of resources for future generations in future controversies. My second objection to McGinnis and Rappaport’s theory of constitutional language flows from the first. McGinnis and Rappaport’s theory argues that rational constitutional adopters are risk averse; therefore seemingly abstract or vague language must refer to a relatively determinate consensus position at the time of adoption. There is a natural experiment to test this claim. Many countries have adopted constitutions following World War II that have abstract rights guarantees. The Canadian Charter of Rights and Freedoms and the South African Constitution are two prominent examples, but there are many others. If constitutional adopters are naturally risk averse and want their constitutions to deliver the best consequences, we should expect to see widespread adoption of something like original-methods originalism around the world. In fact the opposite is the case, even in the United States. Among the nations of the world, American political culture is perhaps the most focused on originalism, but devotion to the Framers’ wishes is honored more in the breach than the observance. Originalist conclusions are often rejected in practice, and American courts do not regularly employ anything like McGinnis and Rappaport’s original-methods originalism to resolve most constitutional controversies. Moreover, American ideas of originalism are not widely adopted outside the United States (although a few constitutional cultures employ originalist arguments in conjunction with other modalities). Looking 59. 376 U.S. 254, 274–77 (1964) (arguing that “[a]lthough the Sedition Act was never tested in this Court, the attack upon its validity has carried the day in the court of history” and citing Jefferson’s and Madison’s views with approval). 60. 347 U.S. 483 (1953); see LUCAS A. POWE, JR., THE WARREN COURT AND AMERICAN POLITICS 58–60 (2000) (describing the role of James J. Kilpatrick and the Richmond News Leader in creating a theory to justify Southern resistance to Brown); Joseph J. Thorndike, “The Sometimes Sordid Level of Race and Segregation”: James J. Kilpatrick and the Virginia Campaign Against Brown, in THE MODERATES’ DILEMMA: MASSIVE RESISTANCE TO SCHOOL DESEGREGATION IN VIRGINIA 51, 51–71 (Matthew D. Lassiter & Andrew B. Lewis eds., 1998) (describing the evolution of interposition arguments as justifications for resistance to Brown). 61. Canadian Charter of Rights and Freedoms, Part I of the Constitution Act, 1982, being Schedule B to the Canada Act, 1982, c. 11 (U.K.); S. AFR. CONST., 1996. 62. See, e.g., Michel Rosenfeld, Constitutional Adjudication in Europe and the United States: Paradoxes and Contrasts, 2 INT’L J. CONST. L. 633, 656, 656 n.83 (2004) (“In Europe . . . recourse to originalism is virtually nonexistent” and “even implicit references to originalism in substance are quite rare.”). David Fontana points out that originalist arguments or originalist rhetoric may sometimes appear in other countries, especially in those with a revolutionary tradition. See David Fontana, Response, Comparative Originalism, 88 TEX. L. REV. 189, 189–90 (2010). But Fontana does not argue BALKIN.DOCX (DO NOT DELETE) 4/25/2012 2:14 PM No. 3] NINE PERSPECTIVES ON LIVING ORIGINALISM 839 immediately to the north, there is no reason to think that Canadians are especially more daring or risk-loving when it comes to constitutions than Americans; yet Canada is famous for its “living tree” doctrine and, as Peter Hogg has put it succinctly, “Originalism has never enjoyed any significant support in Canada.” Yet if McGinnis and Rappaport are correct about what rational adopters would demand for their constitutions, Canada should be just as much a hotbed of originalism as the United States. In fact, scholars and jurists in other countries are often puzzled by the American debate over originalism and America’s originalist obsessions. They do not understand why vague and abstract language requires originalist methods. The dominant strategy for dealing with vague and abstract language in constitutions around the world is proportionality review, and not original-methods originalism. It is possible that—and certainly many American politicians would like us to believe that— everyone else around the world is simply crazy, but perhaps there is a more charitable explanation. The reason why a constitutional culture finds originalism attractive probably has little to do either with rationality or risk aversion. David Fontana has suggested that originalist rhetoric is most likely to appear in countries with a revolutionary tradition in which a constitution is strongly identified with the creation of the nation itself. Ozan Varol, in his study of the secularism provisions of the Turkish Constitution, has suggested that “originalism blossoms when a political leader associated with the creation or revision of the nation’s constitution develops a cult of personality within the nation.” More generally, it seems reasonable to think that a predilection for originalism stems from features of cultural memory characteristic of a particular political tradition, and does not follow from theories of language or rational choice. Perhaps only “real Americans” possess the risk aversion that McGinnis and Rappaport insist is shared by all rational constitutional that any country—including the United States—has adopted something like McGinnis and Rappaport’s model of original methods originalism. See id. 63. See Peter W. Hogg, Canada: From Privy Council to Supreme Court, in INTERPRETING CONSTITUTIONS: A COMPARATIVE STUDY 55, 83 (Jeffrey Goldsworthy ed., 2006). 64. See Jud Mathews & Alec Stone Sweet, All Things in Proportion? American Rights Review and the Problem of Balancing, 60 EMORY L.J. 797, 799 (2011) (“[C]onstitutional courts in legal systems around the world have converged on a method for adjudicating rights claims—proportionality analysis . . . .”). 65. Fontana, supra note 62, at 190 (arguing that “the most relevant” factor explaining a country’s view on originalism “is whether or not its constitution created the nation that lives under the constitution, or whether the constitution merely reorganized the institutions of the country but did not create the nation that lives under the constitution”). 66. Ozan O. Varol, The Origins and Limits of Originalism: A Comparative Study, 44 VAND. J. TRANSNAT’L L. 1239, 1246 (2011). If Varol and Fontana are correct, then originalism would not develop in revolutionary traditions like France, first because the French nation predated the French Revolution and the creation of the First Republic, and second, because the French Revolution itself, and not individual revolutionaries like Robespierre, is the object of civic devotion. BALKIN.DOCX (DO NOT DELETE) 4/25/2012 2:14 PM 840 UNIVERSITY OF ILLINOIS LAW REVIEW [Vol. 2012 adopters. But there is a natural experiment even here. There are fifty American state constitutions, and they are constantly being updated and rewritten. If McGinnis and Rappaport are correct, original-methods originalism, because it produces the best consequences, should have become the standard method of interpretation for state constitutions as well as the federal constitution, because risk averse adopters would demand it. But this does not seem to have occurred. Although state supreme courts, like federal courts, often make originalist arguments, they do not seem to be rigorously originalist in the way that McGinnis and Rappaport demand. Indeed, state courts often seem to outdo federal courts in their embrace of living constitutionalism, or as McGinnis and Rappaport call it, “judicial updating.” For example, many living constitutionalist claims about race relations, sex equality, and gay rights have originated in state courts before gradually being accepted by federal courts. The practice of state courts makes perfect sense in my model of how living constitutionalism operates. Mobilizations and countermobilizations naturally turn first to local venues to win converts, and many of the most important transformations in constitutional doctrine have occurred first in the state courts. But these practices are not consistent with McGinnis and Rappaport’s account. Of course, what courts do is not necessarily the same thing as what rational adopters expect or demand. It is theoretically possible that courts around the world—and in the several states—have repeatedly and systematically acted in defiance of risk averse adopters. Adopters repeatedly insist on original-methods originalism, and the courts repeatedly ignore them. One would think that adopters, being both rational and risk averse, would eventually have figured this out by watching what happened in other jurisdictions, and would stop using such vague and abstract language, or else require constitutions to include qualifiers such as “as understood at the time of adoption.” Apparently, however, they never learn. A more parsimonious explanation, however, is that constitutional adopters are not risk averse in the way that McGinnis and Rappaport imagine they are, and that they adopt vague and abstract language for reasons that are closer to the ones I offer: people use vague and abstract language in constitutions to state general commitments that they hope 67. See SANFORD LEVINSON, FRAMED: AMERICA’S FIFTY-ONE CONSTITUTIONS AND THE CRISIS OF GOVERNANCE 331–35 (2012) (describing the frequency of amendment and the multiple constitutions that have existed in the various states). 68. See, e.g., Jeffrey Omar Usman, Good Enough for Government Work: The Interpretation of Positive Constitutional Rights in State Constitutions, 73 ALB. L. REV. 1459, 1485 (2010) (noting that state constitutional courts generally dispense with elaborate canons of construction and that state uses of originalist argument may be quite different from those found in federal courts). 69. See, e.g., Goodridge v. Dep’t of Pub. Health, 798 N.E.2d 941 (Mass. 2003) (same-sex marriage); Baker v. State, 744 A.2d 864 (Vt. 1999) (same). BALKIN.DOCX (DO NOT DELETE) 4/25/2012 2:14 PM No. 3] NINE PERSPECTIVES ON LIVING ORIGINALISM 841 will guide and channel future politics but whose precise contours will have to be worked out by others over time. III. DEMOCRATIC LEGITIMACY AND DEMOCRATIC PROCESSUALISM Living Originalism argues that the processes of living constitutionalism that have developed in the United States contribute to democratic legitimacy over long periods of time. Fidelity and legitimacy, however, are distinct concepts, and neither guarantees the other. Moreover, as Mariah Zeisberg’s article points out, neither concept is identical with political success or political dominance. In fact, political success can undermine a constitution’s democratic functioning and democratic legitimacy. Fidelity means adherence to a plan for politics. Fidelity concerns (1) whether people are following the plan and (2) whether they are trying to work out the details of the plan in a way that best furthers the plan and its goals. One can speak of fidelity to the entire plan or fidelity to various aspects of the plan—for example, fidelity to the guarantee of the equal protection of the laws. Participants in the constitutional system will often disagree about constitutional fidelity, especially as it relates to constitutional construction by the political branches and by the judiciary. Especially where issues of state-building construction are concerned (for example, how to construct cabinet departments or create institutions that deliver social services), a wide variety of constructions might be faithful to the plan. Nevertheless, participants will argue that some constructions are more faithful to the Constitution than others. Libertarians might argue that the creation of the modern post-New Deal state is less faithful to the constitutional plan than many pre-New Deal constructions; conservatives might insist that cases like Roe v. Wade and Lawrence v. Texas are not faithful interpretations but that robust presidential war powers are faithful to the plan; liberals might disagree, and so on. Fidelity is a judgment made within the constitutional system by the participants; it is often a comparative judgment (more or less faithful), and it is often highly controversial. As a participant in the system, I have my own views on what constructions are most faithful to original meaning and underlying principles, and I articulate some of them in Living Originalism. In some cases, my views are mainstream, in others I am a constitutional dissenter. Moreover, people employing similar methods to mine may reach different conclusions; for example, Randy Barnett and I disagree about the proper interpretation and construction of national power. 70. Mariah Zeisberg, The Politics of Constitutional Fidelity, 2012 U. ILL. L. REV. 801, 804–07. BALKIN.DOCX (DO NOT DELETE) 4/25/2012 2:14 PM 842 UNIVERSITY OF ILLINOIS LAW REVIEW [Vol. 2012 In contrast to fidelity, legitimacy is not a property of individual constructions or interpretations but rather of the constitutional and political system as a whole. Legitimacy concerns whether the constitutional and political system is sufficiently worthy of respect by its participants, that it is reasonable for them to consent to the state’s power over them and others. If the system is not sufficiently worthy of respect at present, then legitimacy requires faith that it will eventually become so. Legitimacy has several dimensions. Sociological legitimacy concerns the extent to which people accept the system as having the right and the authority to rule them. Procedural (or legal) legitimacy concerns the extent to which people clothed with state power (including, for example, government officials, jurors, and voters) make decisions according to official legal rules and procedures. Moral legitimacy concerns the extent to which the system is just or morally admirable, for example, because it protects human rights and makes just decisions about matters of social welfare and foreign policy. Living Originalism is primarily concerned with democratic legitimacy. A system is democratically legitimate to the extent that it allows the members of the political community to govern themselves and makes government action accountable to public will, public values, and public opinion. Democratic legitimacy draws on aspects of the three other forms of legitimacy, although it is not identical with any of them. For example, democratic legitimacy may require that the state treat the members of the relevant political community as equal citizens or with equal and appropriate respect, which, in turn, will require the protection of basic human rights and regular legal procedures. The Constitution and its amendments are acts of popular sovereignty that create a basic plan for a democratic politics that must be filled out over time. Constitutional construction is either directly produced by the political branches (through state building constructions) or is in the long run responsive to them because of institutional constraints on the federal judiciary. Thus, popular sovereignty is exercised directly through the creation of the constitutional framework (including amendments) and either directly or indirectly through constitutional construction. The processes of living constitutionalism are consistent with and promote democratic legitimacy in the medium to long run because they allow political and social mobilizations to change the Constitution-in-practice through constitutional construction as well as constitutional amendment. The American model of constitutional development evolved over a long period of time, partly by design but mostly through successive political struggles. It does not guarantee justice or good outcomes. Indeed, it may produce very unjust results. That is one reason that constitutional faith is required by constitutional dissenters who see its defects all too clearly. BALKIN.DOCX (DO NOT DELETE) 4/25/2012 2:14 PM No. 3] NINE PERSPECTIVES ON LIVING ORIGINALISM 843 Although I believe that the system of living constitutionalism promotes democratic legitimacy, it does not guarantee it. The system has many flaws, and depending on one’s substantive conception of democracy, its flaws may be quite significant. Moreover, there is no guarantee that people working within the system will not, over time, make it less democratically responsive. This is Zeisberg’s point: the mere fact that certain constitutional constructions succeed in winning widespread public acceptance or public support does not mean that they are the most faithful constructions or that they make the constitutional system democratically legitimate. Political success is not the same thing as constitutional fidelity. It is also not the same thing as democratic legitimacy. The first point should be obvious. Political success produced the Jim Crow Republic following the 1896 election. A series of constitutional constructions rationalized racial apartheid in the United States, and the constructions, in turn, were supported by powerful political elites and either accepted or actively pursued by key players in the two major political parties. Jim Crow was a political success; it lasted for seventy years. But this does not make it the most faithful interpretation of the Constitution. Perhaps many people at the time thought that it was, but many constitutional dissenters disagreed. For them, the Constitution was being betrayed. Even a democratically responsive Constitution-in-practice does not guarantee that people will further the constitutional plan in the way that best serves moral legitimacy or democratic legitimacy. A democratically responsive constitution can lead to antidemocratic practices, to great evils, or degrade into a tyranny or a dictatorship. Indeed, this is precisely why the Framers feared democracy in its simplest form. The second point may seem more counterintuitive, but Zeisberg’s basic claim seems correct. Political success does not by itself guarantee democratic legitimacy if certain features of constitutional development undermine democratic legitimacy. For example, the Senate’s rules have evolved so that the filibuster is invoked on even routine legislation. This is the result of successful politics, but, in my view, it undermines democratic responsiveness. The Jim Crow Republic may have been established and reinforced by multiple electoral victories, but I believe it violated a key precept of democratic legitimacy: that the state must treat all of the members of the political community as equal citizens and with the equal respect that comes with that status. In our own day, perceived threats to national security have led the political branches to create an ever more powerful presidency with proliferating institutions and budgets that are shrouded in secrecy, and that are increasingly difficult to subject to either democratic or judicial scrutiny. BALKIN.DOCX (DO NOT DELETE) 4/25/2012 2:14 PM 844 UNIVERSITY OF ILLINOIS LAW REVIEW [Vol. 2012 The famous Carolene Products footnote and John Hart Ely’s theory of judicial review are premised on the idea that some constitutional constructions better serve the goals of democratic legitimacy than others, and that some constitutional developments can cripple democracy or cumulatively undermine self-government. Zeisberg generalizes this familiar structural idea. She argues that we should judge actors within a constitutional system by how well they employ—and build out—their institutional capacities to further the purposes of their particular institution in the larger system of constitutional democracy. She calls this the criterion of democratic “processualism.” Nevertheless, Zeisberg recognizes, the idea of processualism faces an endemic problem. Perhaps the different branches of government have a duty to stay true to the constitutional values that define their proper role in the larger system. But institutions are not always the best judges of their proper capacities and responsibilities. Zeisberg argues that the political branches have a duty of “institutional reflexity”: “The branches are charged with, at a minimum, maintaining their own capacities to govern in ways that are disciplined by their animating constitutional functions.” The difficulty is that although political actors should be self-reflective, their self-analysis is often skewed in their own favor. For example, the more they succeed in political struggles, they more they will tend to assume that they are helping the entire system work better as well. Conversely, when they lose in particular political struggles, they tend to see their failure as due to the failure of opposing institutions to behave according to their proper role. Disputes between the President and Congress, or between federal courts and state political officials, are familiar examples. We may want the President continually to inquire whether the construction of the National Surveillance State is the best way to harness the presidency’s distinctive capacities and obligations to democracy. But the President is likely to think that protecting the country from harm is not only one of his primary duties, but an obligation that he will be judged most harshly for if he fails. In the long run, therefore, presidents will not be institutionally reflexive in the way that Zeisberg would want; they will err on the side of more secrecy and less democratic accountability. It is no accident that President Barack Obama has continued many of President George W. Bush’s policies on detention and surveillance, and is now the primary beneficiary of the FISA Amendments Act of 71. United States v. Carolene Prods. Co., 304 U.S. 144, 152 n.4 (1938). 72. JOHN HART ELY, DEMOCRACY AND DISTRUST: A THEORY OF JUDICIAL REVIEW (1980). 73. Zeisberg, supra note 70, at 808. 74. Id. at 811. 75. Jack M. Balkin, The Constitution in the National Surveillance State, 93 MINN. L. REV. 1, 15, 24–25 (2008); Jack M. Balkin & Sanford Levinson, The Processes of Constitutional Change: From Partisan Entrenchment to the National Surveillance State, 75 FORDHAM L. REV. 489, 489–90 (2006). BALKIN.DOCX (DO NOT DELETE) 4/25/2012 2:14 PM No. 3] NINE PERSPECTIVES ON LIVING ORIGINALISM 845 2008, which ratified several of the Bush Administration’s practices and gave authority to expand surveillance programs. As the Framers recognized, participants in a self-enforcing constitutional system must struggle with each other to keep the system functioning appropriately over time. Arguments about proper democratic functioning in such a system are in politics, not outside it. Therefore, Zeisberg concludes, “Since the constitutional politics of the branches is conducted in partisan ways, we can expect that applying the standard of processualism will lead to conclusions that will register as partisan in the broader political landscape.” I agree. This is one of the consequences of a self-enforcing constitutional system. To paraphrase what is often said in First Amendment law, in such a system of government, the only cure to the kind of politics we do not like is more politics. IV. CONSTITUTIONAL CULTURES AND CONSTITUTIONAL IMPLICATIONS Jeffrey Goldsworthy’s article is so rich that I cannot even begin to do it justice in this reply. He has at least five claims, each of which would merit (at least) a full article in response. In particular, I will have to leave unanswered his succinct and powerful critique of the American system of judicial review; his argument that a parliamentary system without judicially enforceable constitutional rights guarantees—like Australia’s— is superior in terms of democratic legitimacy to the system of living constitutionalism that I describe in Living Originalism; and his doubts about why Americans should even have faith in their constitutional system, given its obvious imperfections. These are serious matters, but to give them the responses they deserve would turn this article into a second book. And, in many ways, Goldworthy’s criticisms are as much directed to American constitutional theory in general as they are to Living Originalism. Even so, I was unable to limit myself to a single part of his interesting article to comment on. Instead, I chose two. The first concerns how the analysis in Living Originalism translates to other constitutional cultures. The second concerns the role that intentions and implications play in the interpretation of constitutional texts. 76. Foreign Intelligence Surveillance Act of 1978 Amendments Act of 2008, Pub. L. No. 110-261, 122 Stat. 2436. 77. Zeisberg, supra note 70, at 812. 78. Cf. Whitney v. California, 274 U.S. 357, 375, 377 (1927) (Brandeis, J., concurring) (stating that “the fitting remedy for evil counsels is good ones” and “the remedy to be applied [to evil speech] is more speech”). 79. Jeffrey Goldsworthy, Constitutional Cultures, Democracy, Implications, and Underlying Principles, 2012 U. ILL. L. REV. 683. BALKIN.DOCX (DO NOT DELETE) 4/25/2012 2:14 PM 846 UNIVERSITY OF ILLINOIS LAW REVIEW [Vol. 2012 A. The Importance of Constitutional Culture Living Originalism argues that in order to be successful, a constitution like America’s must succeed simultaneously as basic law, higher law, and our law. By basic law, I mean that the Constitution serves as a basic framework for governance that allocates powers and responsibilities, sets up a plan for ordering political life, and offers ways of implementing, expanding, or modifying the plan over time. By higher law, I mean that the Constitution serves as a source of aspiration and a reflection of values that stand above ordinary law and hold it to account. By our law, I mean that people are attached to the Constitution, feel a sense of ownership in it, and see it as the product of their collective efforts as a people. Viewing the Constitution as “our law” helps constitute Americans as a single people, dedicated to a political project that extends over time, in which past, present, and future generations participate. This collective identification, in turn, is a constitutional story that allows Americans to regard the Constitution as their own even if they never officially consented to it. But what if a constitution is not like America’s? Goldsworthy’s article points out that constitutions in different political cultures may not serve these functions or serve them in the same way. Goldsworthy offers the United Kingdom, Australia, and Canada as examples. The British Constitution of precedents, customs, and ancient statutes may be basic law, but it can be changed by ordinary statute. Goldsworthy does not believe that Australia’s Constitution serves an aspirational or higher law function because it does not contain rights guarantees (although the Australian Supreme Court has implied a few). Canada’s Constitution is not “our law” in the same way as America’s because it began as the British North America Act, created and imposed on Canadians by a colonial power. Nor, in Goldsworthy’s view, do Australia’s or Canada’s constitutions form important—much less central—parts of their national narratives in the same way that the American Constitution does in the United States. I agree with Goldsworthy’s basic insight. Living Originalism is a book about the American Constitution, and about the distinctive features of American political culture that make the Constitution central to American political identity. To apply the arguments of Living Originalism to other political cultures, we must first ask what role a constitution plays in a country’s political culture, and we must know something about that country’s history of constitution making. Most constitutions probably do serve as basic law in my sense; as the example of Britain’s unwritten constitution suggests, however, they 80. BALKIN, supra note 1, at 59. 81. Goldsworthy, supra note 79, at 684–90, 694. 82. Id. at 694. BALKIN.DOCX (DO NOT DELETE) 4/25/2012 2:14 PM No. 3] NINE PERSPECTIVES ON LIVING ORIGINALISM 847 may do so in importantly different ways. Many constitutions may not serve as higher law in the same way that America’s Constitution does; the idea of political aspiration may come from different features of a political culture and its history. People in other countries may not point to their constitutions as the symbol or embodiment of their political ideals and aspirations. They may point to national traditions, religious institutions, or past political struggles. The U.S. Constitution and the history of its development serve a crucial role in the country’s national narrative and in Americans’ sense of themselves as Americans. But other constitutions—indeed, perhaps most—may not serve as “our law” in the same strong sense as the American Constitution. Much depends on the circumstances in which a constitution was created—imposed by a former colonial power, drafted under Soviet domination, promulgated by a former or current strongman or autocrat, or developed through a popular referendum or democratic process widely assumed to be fair and legitimate. Even in the latter case, a constitution may not have a significant role in national identity; for example, because the nation is much older than its current constitution, or because national identity is primarily ethnically or religiously defined. Moreover, few written constitutions are as durable as the American example; some are changed fairly often. Many never manage to obtain the sense of pride or reverence that Americans spontaneously show toward their ancient Constitution. To apply the argument of Living Originalism to other constitutional systems, then, one must first look to differences before one can profitably examine similarities. It may be useful to ask at the outset whether a particular constitution serves, or even needs to serve, the functions I have described as basic law, higher law, and our law; and if not, why not. We may find that some of these aspects are distributed and realized elsewhere in the politics and culture of a nation. To be sure, many of the arguments in Living Originalism about interpretation and construction can be applied to other political systems. But in doing so, one must always take into account the circumstances of American constitutionalism and the many ways that American political history and political culture have shaped its Constitution, and vice versa. B. Constitutional Intentions and Implications The final part of Goldsworthy’s article discusses the role of intentions in original-meaning originalism. Even though advocates of original-meaning originalism eschew reference to the intentions of the adopters, Goldsworthy points out that one cannot avoid ascribing purposes and intentions to constitutional adopters. I agree. In Living Originalism, I argue that we ascribe purposes or intentions to the adopters when we attempt to ascertain original meaning. For example, we ascribe purposes or intentions to constitutional BALKIN.DOCX (DO NOT DELETE) 4/25/2012 2:14 PM 848 UNIVERSITY OF ILLINOIS LAW REVIEW [Vol. 2012 adopters when we resolve ambiguities in the text in the following situations: 1. When we decide whether a certain word or phrase refers to one concept rather than another. 2. When we decide whether the text states a rule, a standard, or a principle. 3. When we decide whether the text employs a nonliteral usage or a generally recognized term of art. (Examples include the word “writings” in the progress clause or the words “Congress” or “speech” in the First Amendment.) 4. When we identify certain very basic assumptions in the text, for example, that the numerals in the Constitution are stated in base ten, not base two, and that the calendar employed is the Gregorian calendar, not the Jewish lunar calendar or the Mayan calendar. We ascribe purposes or intentions in each of these cases because we are trying to identify the basic plan for politics on which we are trying to build, and we cannot do this without ascribing purposes or intentions to the drafters and adopters of the plan. In my discussion of McGinnis and Rappaport’s work, I point out that appeals to original meaning are always partly anachronistic. In constitutional interpretation, we do not include all aspects of the past as part of original meaning, only those parts that we think are relevant to the interpretive enterprise we are engaged in. In the same way, we are not interested in all aspects of the adopters’ purposes or intentions, only those that are relevant to our enterprise. Goldsworthy is particularly concerned with the fourth category listed above. Constitutional texts, like other texts, contain implicit presumptions; these are tacit assumptions that are taken for granted and therefore are not explicitly expressed to the audience. Sometimes they are not even consciously understood by the speaker or the audience. When we identify presuppositions we are ascribing intentions or purposes to the authors of a document. But this ascription of purpose or intention is not the same thing as a description of psychological states that people held, for three reasons. First, people in the adopting generation might not have thought about whether the numbers they wrote down were in base two or base ten, or whether they were using the Gregorian calendar as opposed to an alternative calendar. Second, not everyone in the relevant pool of adopters might have had the same psychological states (purposes, beliefs, or intentions)—or indeed, any at all—with respect to a particular question. Nevertheless, we ascribe purposes or intentions to the group as a whole.