Fidelity and Constraint Citation Lawrence Lessig, Fidelity and Constraint , 65 Fordham L. Rev. 1365 (1997). Published Version http://ir.lawnet.fordham.edu/flr/vol65/iss4/8/ Permanent link http://nrs.harvard.edu/urn-3:HUL.InstRepos:12954336 Terms of Use This article was downloaded from Harvard University’s DASH repository, and is made available under the terms and conditions applicable to Other Posted Material, as set forth at http:// nrs.harvard.edu/urn-3:HUL.InstRepos:dash.current.terms-of-use#LAA Share Your Story The Harvard community has made this article openly available. Please share how this access benefits you. Submit a story . Accessibility Fordham Law Review Volume 65|Issue 4 Article 8 1997 Fidelity and Constraint Lawrence Lessig Recommended Citation Lawrence Lessig,Fidelity and Constraint, 65 FordhamL. Rev. 1365 (1997). Available at: http://ir.lawnet.fordham.edu/flr/vol65/iss4/8 This Article is brought to you for free and open access by FLASH: The Fordham Law Archive of Scholarship and History. It has been accepted for inclusion in Fordham Law Review by an authorized administrator of FLASH: The Fordham Law Archive of Scholarship and History. For more information, please [email protected]. FIDELITY AND CONSTRAINT Lawrence Lessig* TABF OF CoNrrrrs INTRODUCrION ................................................. 1365 I. FIRST STEPS ............................................. 1367 A. What Fidelity Is Not ................................ 1368 B. The Idea of Translation ............................. 1371 C. How Much Translation Explains .................... 1376 1. TVs and Airforces .............................. 1376 2. Commerce I .................................... 1377 3. Privacy .......................................... 1378 4. Federalism, Herein Commerce II ............... 1379 5. Takings ......................................... 1381 D. The Constraints on Forward Translation ............ 1384 II. Ti CONSTRAINTS OF CoNTExr ........................ 1386 A. Rendered Meanings. The Rate-Making Cases ....... 1388 B. How Contests Render Meaning ..................... 1393 C. Rendering Passive: Erie, et al ...................... 1400 D. Rendering Active: Equal Protection. ................ 1412 E. What the Erie-effect Explains ...................... 1430 CONCLUSION ................................................... 1432 INTRODUCTION HFERE are two ideas, apparently unrelated, that I want to argue are, from the perspective of a theory of interpretive fidelity-or fidelity theory-intimately connected. First: Judges live life subject to constraints. They judge subject to constraints. These constraints, for the most part, are the same sort of constraints that we all face. They are the constraints of how they, and how we, see the world. They are a function of the things we take for granted-the things we don't think much about; that we argue, subject to; that we contest, relative to. Judges judge subject to these con- straints. These constraints change. And judgments (and judges) are vulnerable to these changes. * Thanks to Curtis Bradley, Rosemary Coombe, Jack Goldsmith, Larry Kramer, Alan Meese, Richard Posner, Arti Rai, Anthony Sebok, Peter Strauss, and Ruti Teitel for guidance on much of this, and especially the University of Colorado at Boulder's Legal Theory Workshop, the Columbia Legal Theory Workshop, the University of Virginia Legal Workshop Series, and NYU's Sager/Eisgruber Colloquium, for helpful criticisms of the Erie-effect. I am indebted to David Strauss for help on the equal protection argument in particular. Funding for this project was provided by the Douglas Law and Government Fund and the Sarah Scaife Foundation. Excellent re- search assistance was provided by Ashley Parrish. 1365 1366 FORDHAM LAW REVIEW [Vol. 65 Second: Readings of the Constitution change. This is the brute fact of our constitutional past. The Constitution is read at one time to mean one thing; at another, to mean something quite different. These changes track no change in constitutional text; nor do they follow con- fessions of earlier mistake. How, we should ask, could these changes be consistent with a theory of interpretive fidelity? We have been debating this question of interpretive fidelity for a very long time.' It is late in this argument's day. So if there is some- thing new to be added to this debate, one has a duty to put the new up front. So here is the new that I want to argue: That any theory of interpretative fidelity must account for the first point; and that ac- counting for the first point is a way to understand the second. That understanding how changed readings can be readings of fidelity de- pends upon understanding how the constraints of context matter in a theory of interpretive fidelity. Most theories ignore the first point, and simply hide the second. Most proceed as if this stuff taken for granted can, as it were, be taken for granted. Others (within the cruder forms of the "it's all politics" school) presume such constraints are all that there is-as if constitu- tional law were simply epiphenomenal upon social and political con- text. But constraints of context are neither nothing, nor everything. They are something more interesting than either extreme allows. My claim is that we can say something useful about how and where they matter; and that saying something useful about how and where they matter is essential if a theory of fidelity is to be complete. Put most crudely, my account will track readings as a function of three states (as in the "states" that water may take-solid, liquid, gas) that contexts may take. The first is the state when things are taken for granted-when matters, in a way that I will describe more fully below, stand relatively uncontested in the present interpretive context. The second is when things stand contested-when, again in a way that I will describe more fully below, they are understood, in the present interpretive context to be the subject of dispute. And the third is when things change from contested to uncontested, or uncontested to contested. These states, and the constraints that they yield, will matter, I sug- gest, to a range of changed readings in our constitutional past; they will reveal a pattern in these changed readings; and they will suggest something about how we might expect readings to change in the fu- 1. The best of the early vollies is in Paul Brest, The Misconceived Quest for the Original Understanding,6 0 B.U. L. Rev. 204 (1980). Compare Edwin Meese III, Ad- dress Before the D.C. Chapter of the FederalistS ociety Lawyers Division, reprinted in Interpreting Law and Literature: A Hermeneutic Reader 25 (Sanford Levinson & Steven Mailloux eds., 1988) with Robert H. Bork, Neutral Principlesa nd Some First Amendment Problems, 47 Ind. L.J. 1 (1971). 1997] FIDELITY AS TRANSLATION 1367 ture. In what follows, I sketch this pattern and how it will matter to a theory of interpretive fidelity. The theory of fidelity that I describe is an account grounded in a practice of translation. Again, my broader claim is that any theory of fidelity must describe the constraints I discuss here. But a theory of fidelity as translation makes sense of just why such an account is needed. In the first section below, I describe the theory briefly, and incompletely, for its idea is quickly grasped (and I have described it in tedious detail elsewhere). In the second section, I then explore the place that these constraints of context have in the account of fidelity as translation, and the place they must hold in any account of interpre- tive fidelity. I. FIRST STEPS Readings of the Constitution have changed. A theory of fidelity must explain at least this. It must explain, that is, why readings change, whether such changes are changes of fidelity, and more gener- ally, how we could know whether such changes are changes of fidelity. A theory modeled on translation is one such account. It uses "translation" as a heuristic for suggesting just how changed readings could be changes of fidelity. It uses this heuristic for guiding this change to assure that they are. The heuristic is modeled upon linguis- tic translation, or what we ordinarily mean by "translation," but the pattern it invokes is much broader than that. As George Steiner puts it, "[w]hen we read or hear any language-statement from the past... we translate."' Translation, in this sense, is a practice to understand a contextually distant text, whether a text written in a different lan- guage, or a text written at a different time. It is translation in this broader sense that I mean to model, and the ethic that this practice invokes that I mean to draw upon. But we must be clear about the role that a theory of translation is to fill. It is not the only player in an account of interpretive fidelity or history. It is not an elixir of constitutional understanding. I don't be- lieve that a theory of translation explains all changed readings in our constitutional past: Some readings change for reasons unrelated to the understanding of translation. Some change, for example, because the Constitution has changed-because it has, for example, been amended.3 Some change because the interpreter is trying to synthe- 2. See George Steiner, After Babel: Aspects of Language and Translation 28 (1975). 3. Brushaber v. Union Pac. RR, 240 U.S. 1 (1916), is a simple example. Brusha- ber changed the reading of the taxing power offered in Pollock v. Farmers' Loan & Trust Co., 157 U.S. 429 (1895). Brushaber, 240 U.S. at 18-19. It did so because be- tween the reading offered in Pollack and the reading offered in Brushaber, the Six- teenth Amendment was ratified. i at 20. 1368 FORDHAM LAW REVIEW [Vol. 65 size an amendment into the unchanged Constitution.4 Some change because an earlier reading turns out to be a mistake.5 Or some change because of simple politics.6 If these are the five reasons that readings may change, then translation is one of four justified reasons why read- ings change. It is an account that seeks to explain a residual-to ex- plain changes that amendment, or synthesis, or mistake, don't. This is the positive use of a theory of translation. There is a nega- tive use as well-one which perhaps motivates the positive theory, and best introduces the role that the positive theory must play. The negative use is in response to what might seem an intuitively attractive account of interpretive fidelity. Its aim is to dislodge this initial intui- tion. So let me begin with this negative use, by arguing what fidelity is not. A. What Fidelity Is Not Fidelity on any account implies a certain constancy; the question is what sort of constancy this is. One might think that the faithful reader is the one who does just what the author would have done-or in our case, that the faithful interpreter of the Constitution is the one who does just what the Framers would have done. This is fidelity at the level of results: We can imagine the Framers confronted with the questions we must answer; we can imagine what they would answer; fidelity (so understood) demands that we give the same answers that they would have given. We can call this one-step originalism, and distinguish two different forms. One is historical: it concerns questions that the Framers did confront-paradigm cases, we might say, that expressed their under- standing of the Constitution's original meaning.7 The other is counterfactual: it concerns questions that we imagine the Framers confronting-questions like, What would Madison say to this ques- tion? In either form, we are returning, in a sense, to a place or person long gone, and putting a question there, or to them, to see what their answer was, or would be. That answer, this form of fidelity says, is the answer that we too must adopt. It is true that no theorist of fidelity, whether judge or academic, has ever adopted such a theory for every kind of constitutional question. But it is not true that no one is a one-step originalist, at least for a 4. This is my view of Boiling v. Sharpe, 347 U.S. 497 (1954). See Lawrence Lessig, Understanding Changed Readings: Fidelity and Theory, 47 Stan. L. Rev. 395, 409-10 (1995) [hereinafter Lessig, UnderstandingC hanged Readings]. 5. The recent reversal of the question whether Victim Impact Statements are ad- missible is defended as a change grounded in a mistake. See Payne v. Tennessee, 501 U.S. 808, 830 (1991). 6. A better understanding of the same case may be just politics. See id. at 844, 850 (Marshall, J., dissenting). 7. See Jed Rubenfeld, Reading the Constitution as Spoken, 104 Yale L.J. 1119, 1169-71 (1995). 1997] FIDELITY AS TRANSLATION 1369 broad range of important cases." One-step originalism is alive and well in our legal culture. But whether general or not doesn't matter for my purposes here. The point of the category is to make plain one conceptually possible theory of fidelity, and then to make plain just why such a conception can't be fidelity in any meaningful sense. This we can see either from a reflection in the form of an example, or in the form of a theory. The examples are many, but one should suggest the point. The one- step says that we look to what the Framers would have said to deter- mine what the Constitution means. So does the Constitution include the power to regulate commerce in meat? If one answered that ques- tion by returning to the Framing, and putting the question to a select few of the Framers, the answer would have been "no." At the time of the Framing, the slaughter business was local: Meat could not be shipped any great distance without great spoilage; its market was re- stricted; and because its market was restricted, a clause that gave Con- gress the power to regulate commerce among the several states did not give Congress the power to regulate slaughtered meat. Early in the nineteenth century, cold-storage transport was in- vented; by the mid-nineteenth century, it was the dominant mode of transporting slaughtered meat.9 This transport carried slaughtered meat far beyond the limits of its local area of production. Indeed, after the establishment of cold-car transport, the market for meat be- came quite national. Meat was shipped in interstate commerce across the country, and the quantity of meat produced increased dramatically. All this is obvious. But the obvious should remind us about what is within the practice of a faithful reading of the commerce power. Under any plausible modem reading of the commerce power, it in- cludes the power to regulate the sale of this meat,10 because readings of the commerce power depend upon something about the world the commerce power functions in. Readings must be made with reference to that world. But on the method of the one-step, they are not. On the method of the one-step, the question is answered in 1791, cold-car transport notwithstanding. Now again, no one takes this extreme one-step position--consist- ently at least with respect to commerce, even though the rhetoric 8. While Justice Scalia's views are more subtle than this, his writing does at times suggest one-stepism. See Harmelin v. Michigan, 501 U.S. 957 (1991); Oilman v. Evans, 750 F.2d 970, 1038 n.2 (D.C. Cir. 1984) (en banc) (Scalia, J., dissenting); Antonin Scalia, Originalism: The Lesser Evil, 57 U. Cin. L. Rev. 849 (1989) [hereinafter Scalia, Originalism]. 9. See Oscar Edward Anderson Jr., Refrigeration in America: A History of a New Technology and Its Impact 65 (1953). 10. At least as to its transport across state lines. As for the bucking of the old Court in what might be thought to be related areas, see Hammer v. Dagenhart, 247 U.S. 251 (1918). 1370 FORDHAM LAW REVIEW [Vol. 65 among originalists often suggests something quite close." But my point is not to defeat a movement with a single example. It is instead to establish what seems to me an inescapable conclusion: that once one rejects this clean, one-step originalist view, one is launched on a very difficult task: to separate out changes in the interpretive context that will matter to interpretive fidelity from changes in the interpre- tive context that won't. How one does this is an extremely important question.'2 However one does that, my point so far is just this: That fidelity must include something more than just asking what they would have done. That is the example of why one-step originalism can't be a theory of fidelity. The theory follows just about as neatly. The Framers said what they said in order to say something; what their words said was set in part by the context within which they said them. Had that con- text been different, then, to say the same thing, they would have had to speak differently. Their words, their meaning, hung upon this context. In interpreting their words, then, no doubt the original context sets an anchor. It sets the orientation, or the direction, or the meaning, of what they wrote. It determines the measure against which all that happens afterwards must be set. It is the baseline that should guide later readings. What should this guidance be? For the one-step originalist, the gui- dance is the result; the result is the beginning and the end; it sets the original and current answer. But if meaning hangs on context, then this method creates an odd result. For if the context within which this original answer is applied changes, then the meaning of the original answer may change. If we do the same thing as they did, then the meaning of what we do may be different. And if fidelity is to preserve meaning, then this means that the one-step originalist's method is a method that changes the Constitution's meaning. For the translator, in contrast, this first moment is just a beginning. With it as a baseline, the translator then constructs a text-a read- ing-that in the current context has the same meaning as the original text in its original context. Contexts change, so readings must change. The aim of the translator is to find a reading that neutralizes the change in context. 11. McIntyre v. Ohio Elections Comm'n, 115 S. Ct. 1511, 1525 (1995) (Thomas, J., concurring). 12. Michael Klarman suggests that any such division will be "arbitrary." See Michael J. Klarman, Anti-Fidelity, 70 S. Cal. L. Rev. (forthcoming 1997). I don't be- lieve so if the division is supported by reason. If Klarman's point is that reasons never matter, because reasons are mere pretext, then perhaps, but so what? Ours is an enterprise that presupposes the negation of just that view. Global skepticism can't touch local arguments. 1997] FIDELITY AS TRANSLATION 1371 How is a difficult question. But the point so far is just that some- times it must. It must, that is, if meaning is contextual, and if meaning is what the fidelitist must preserve. This is the negative use of a theory of translation. It should get us going on a search for a theory a bit more thick. B. The Idea of Translation A positive theory of fidelity is such a theory. It uses the model of translation to describe just why changes in readings can be changes of fidelity. Begin then with this practice (translation) that the theory (fi- delity as translation) models. In its simplest, perhaps crudest form, the translator's task is this: She is presented with a text that is written in one language. This is the source text; the language is the source language. Her aim is to write another text in a second language-the target language, and her text, the target text. If the translation succeeds-if it is a good transla- tion-then there is an important relation between the two texts, in these two contexts: naively put, their "meaning" is to be "the same." Different texts; different contexts; same meaning. My claim is that judges face a similar task. Like the linguistic trans- lator, the judge is faced with a text (say, the Constitution), written in an original or source context (America, late eighteenth century); she too must write a text (a decision, or an opinion) in a different context (America, today); this decision, in its context, is to have the same meaning as the original text in its context. Of course, unlike the lin- guistic translator, the judge can't simply rewrite the original text. That text must remain the same. But the judge does change readings of that original text: The judge gives a reading of the original text that in the current context yields the same meaning as the original text in its original context. For the legal translator, the reading is to the original text as, to the linguistic translator, the target text is to the source text. This process is not automatic. But neither is it automatic for the linguistic translator. There is no obvious, or complete, mapping of one language onto another; no simple formula will carry meaning from one to another. Always the linguistic translator, like the legal translator, makes a certain judgment about how best to carry the meaning of one world into second. And always this judgment requires choice. The choices are of two kinds. Fist there is a choice about the kind of fidelity that a particular type of translation requires. No single practice of translation governs the full range of texts that can be trans- lated. Bibles, instructions to children's toys, plays, political slogans- while there is a practice of translation that can apply to each of these, the ends of these practices are very different. These differences in ends suggest differences in means. Thus the first choice that a transla- tor must make is a choice about the end that his translation is to serve. 1372 FORDHAM LAW REVIEW [Vol. 65 A second choice is about the constraints on the practice, whatever the end that has been selected. And here, the links with linguistic translation are quite rich. For translators struggle with a tension that defines the tension confronted by the judge: If translation requires creativity-if there is no such thing as "mechanical" translation-then some counsel the translator to a kind of humility.'3 Humility means this: to avoid translations that the translator believes make the text a better text; to choose instead translations that will carry over a text's flaws as well as its virtues. This counsel to humility is offered as a virtue in the translator's practice. It is an integrity-to be faithful to the strengths of a text as well as the vice; to exercise the power of the translator in a sense not to change the text translated, while in a sense changing the text translated fundamentally.'4 On this view of the translator's task, fidelity requires a certain restraint-the restraint to minimize the voice of the translator in the text being translated. This is the integrity of the linguistic translator' 5-an integrity that guides and constrains the translator in the practice of translation. Both dimensions of this integrity-setting the end of the translation, and practicing humility within that end-link the practice of the lin- guistic translator with the practice of the judge. For the judge as well must select the kind of fidelity that her reading will preserve, and she must pursue that fidelity constrained by an analogous form of 16 humility. But a nagging difference remains. The translator moves between two texts written in two languages; the judge moves between two texts written in one. The constitutional interpreter is reading a text in Eng- lish that was written as a text in English. What gain could there be in modeling a process of reading a text in the same language on a prac- tice of reading a text from a different language? The gain in part is the gain from being able to draw on a body of related interpretive theory, to guide analogous problems in these two interpretive practices. But in the main, the gain is a certain discipline, or analytic clarity, that comes from separating out two steps where a less careful method might see just one. The two steps are these:17 The first is to locate a meaning in an original context; the second is to ask how that meaning is to be carried 13. This, for example, was Samuel Johnson's view. See Bayard Quincy Morgan, Bibliography: 46 B.C.-1958, in On Translation 275 (Reuben A. Brower ed., 1966). 14. See Lawrence Lessig, Fidelity in Translation,7 1 Tex. L. Rev. 1165, 1189-1211 (1993) [hereinafter Lessig, Fidelity in Translation] for a discussion of the parallels be- tween translation and legal interpretation. 15. Not to say all do this. See Morgan, supra note 13, at 277 ("[T]he live Dog better than the dead Lion."). 16. See Lessig, Fidelity in Translation,s upra note 14, at 1206-11. 17. By distinct, I do not mean that the translator does not alternate between the two steps. The claim is not that one step is complete before the other is taken. The