0521652502 cambridge university press the american language of rights aug 1999

The American Language of Rights is a major contribution to contemporary political theory, of interest to scholars and students in politics and government, constitutional law, and America

THE AMERICAN LANGUAGE OF RIGHTS

Richard A Primus examines three crucial periods in American history (the late eighteenth century, the Civil War and the 1950s and 1960s) in order to demonstrate how the conceptions of rights prevailing at each of these times grew out of reactions to contemporary social and political crises His innovative approach sees rights language as grounded more in opposition

to concrete social and political practices than in the listic paradigms presented by many political philosophers This study demonstrates the potency of the language of rights throughout American history and looks for the ®rst time at the impact of modern totalitarianism (in Nazi Germany and the Soviet Union) on American conceptions of rights The American Language of Rights is a major contribution to contemporary political theory, of interest to scholars and students in politics and government, constitutional law, and American history richard a primus studied political theory at Harvard and Oxford and law at Yale.

universa-ideas in context

Edited by QUENTIN SKINNER (General Editor), LORRAINE DASTON

and JAMES TULLY The books in this series will discuss the emergence of intellectual traditions and of related new disciplines The procedures, aims and vocabularies that were generated will be set in the context of the alternatives available within the contemporary frameworks of ideas and institutions Through detailed studies of the evolution of such traditions, and their modi®cation by different audiences, it is hoped that a new picture will form of the development of ideas in their concrete contexts By this means, arti®cial distinctions between the history of philosophy, of the various sciences, of society and politics, and of literature may be seen to dissolve.

The series is published with the support of the Exxon Foundation.

A list of books in the series will be found at the end of the volume.

THE AMERICAN LANGUAGE

OF RIGHTS

RICHARD A PRIMUS

The Pitt Building, Trumpington Street, Cambridge, United Kingdom

  

The Edinburgh Building, Cambridge CB2 2RU, UK

40 West 20th Street, New York, NY 10011-4211, USA

477 Williamstown Road, Port Melbourne, VIC 3207, Australia

Ruiz de Alarcón 13, 28014 Madrid, Spain

Dock House, The Waterfront, Cape Town 8001, South Africa

©

for Sarahlike everything else

``All of us, grave or light, get our thoughts entangled in metaphors, and act fatally on the strength of them.''

George Eliot, Middlemarch

1 Rights theory and rights practice 10

2 History and the development of rights 45

4 Rights and Reconstruction: syntheses and shell games 127

xi

My foremost thanks for help with the project that became this book

go to Michael Freeden, who patiently inducted me into the discipline

of political theory I am also grateful for the assistance of a host ofother ®ne teachers, including Bruce Ackerman, Bernard Bailyn,Peter Berkowitz, Eric Foner, Paul Kahn, Sanford Levinson, MarkPhilp, and Adam Swift Without the bene®t of their skepticism andsuggestions, the ideas in this book would be even less developed thanthey now are I also thank three others who have touched the life ofthis book in more diffuse but not less important ways: GuidoCalabresi, Owen Fiss, and Bob Gordon

There are two other teachers without whom this book could neverhave been written One is H W Perry I have studied constitutionallaw with several distinguished scholars since the September morningwhen he required me to state the facts of Marbury v Madison, but hiscourse remains my image of what a course in that subject should be.Finally, it is my greatest regret in these acknowledgments that Icannot thank the teacher to whom, in some ways, I owe more thanany other: Judith Shklar I believe that she would have disapproved

of much of this project, but I ®nd some consolation in believing thatshe would have enjoyed ®ghting with me about it

The friends who have helped me with and through this book, inone way or another, are so many that I would risk offending byomission if I tried to name them all Nevertheless, the pleasure ofnaming names is so great that I accept the risk, hoping to do moregood than ill I thank Gary Bass, Michael Bloom, Jennifer Bradley,Niko Canner, Marilyn Cooper, Ariela Dubler, Jonah Edelman, ElieFishman, Jesse Furman, Robert Gordon, Eric Gregory, Jay Koh,Niko Kolodny, Rachel Harmon, Dan Libenson, Sarah Levine,Meira Levinson, Jack Levy, Laura McGrane, Nader Mousavizadeh,Brian Reed, Kevin Stack, Jody Seim Timmins, and Rebecca

xiii

Thomas Tingle I also thank three others who followed by progress

so closely and participated in my thinking so intimately, that Isometimes wondered how they could have any time or mentalenergy left for their own work They are Noah Feldman, Nien-heHsieh, and Kurt Strovink For their friendship and their intellectualcomeradeship, I am deeply grateful

Richard Fisher and Elizabeth Howard of the Cambridge sity Press provided patient hand-holding as a rough manuscriptbecame a presentable book Alan Lord tirelessly accommodated mydemands on his library I also thank the editors of the Duke LawJournal and the Yale Law Journal for permission to publish here, albeit

Univer-in expanded form, some of the same ideas that I have previouslypresented in their pages

Finally, I thank Aryeh, Lisa, Roni, Ida, Dima, my parents, and mygrandparents They know why

R.A.P.New Haven, 1999

This is a book about rights It is, to be more speci®c, a book aboutthe place of rights in American political debate The language ofrights has been central to American political culture for centuries,and nearly every major issue in American political history has beenargued as an issue of rights In some circles, however, rights haverecently fallen upon hard times It has become a familiar refrain inpolitics that America is a land of too many rights and too fewresponsibilities In the academy, the concept of rights today attractscriticism that is both serious and diverse The critical legal studiesmovement charges that a rights-based political order is intellectuallyincoherent and morally pernicious, the former because rights con-

¯ict with one another and the latter because a system of rightsentrenches the power of the privileged classes Communitarianscontend that framing political debate in terms of rights leads toexcessive individualism, unwillingness to compromise, and thedecline of community spirit These two schools of thought differ inmany important ways, but their different critiques of rights share theidea that the substance of American politics is conditioned by thefact that it speaks the language of rights To some extent, thatobservation must be valid Because language is often a constituentelement of thought, belief, and action, political activity from legisla-tive drafting to electioneering to international diplomacy relies uponand is shaped by the language that mediates and helps constitute thepolitical world When communitarian and critical legal scholarsargue that speaking the language of rights pushes a political culturetoward some substantive outcomes and away from others, the realitythat language plays a constituent role in politics lends intuitiveappeal to their claim

That claim, however, underplays two other elements of therelationship between politics and language First, the constituent

1

relationship between them is reciprocal Precisely because languageand politics are closely interconnected and therefore dif®cult toanalyze sophisticatedly in isolation from each other, it is hard toimagine language as a separate and antecedent sphere that in¯u-ences politics while remaining itself untouched.1 Language doesshape politics, but politics shapes language as well Second, thein¯uence of language upon politics is not always narrowly deter-mined Even if speaking the language of rights does distinctivelycondition political discourse in America, it does not follow that thelanguage of rights shapes politics in the speci®c ways that prominentcritics contend The tendency of rights discourse to favor one set ofpolitical outcomes over another, I suggest, has been widely over-stated As a conceptual matter, people with all sorts of differingpolitical views can use the language of rights to their advantage.Such diverse usage is a historical fact, not merely an abstractpossibility: the history of American rights discourse shows thatAmericans of all political stripes have in fact used the language ofrights to support their various causes, and all of the greatest politicalcon¯icts in American history have involved, and been spoken about

as, con¯icts of rights That these political con¯icts have beendescribed as con¯icts between opposing rights rather than ascon¯icts between rights on one side and something else on the othersuggests that the language of rights does not attach to only one kind

of political outcome Instead, either side of an issue can and usuallywill use the language of rights in support of its position Eighteenth-century Americans who advocated independence from Britain drewheavily on the language of rights in their political writings, butAmericans who opposed independence were no more loath to userights language in their anti-independence arguments In the nine-teenth century, when the issue of slavery split the United States,abolitionists and Republicans argued their cause in the name of awhole battery of rights, ranging from the natural rights of people to

be free to the rights of free labor and free speech Southerners andslaveholders, however, also marshaled arguments from rights, in-voking the right to property and the rights of states Later con¯icts inAmerican history featured the right to strike against the right towork, the right to dispose of one's property in whatever way one

1 I explore these ideas in more detail in chapter 1, with reference to scholars such as W V O Quine and Quentin Skinner.

chose against the right to regulate private property in the publicinterest, the right to equality against the right of association, therights of the federal government against the rights of the states, andthe ``right to life'' against the ``right to choose.'' In short, thelanguage of rights has not shown itself systematically partisan to anyone political vision at the expense of all others It has been a versatiletool, suitable for many different agendas.

It could, of course, be argued that both sides in all of thosecon¯icts shared elements of the classical liberal framework and thatthe language of rights is useful to all those (but only those) who arewithin that political family Whether all of American politics can bedescribed as liberal is the subject of a venerable scholarly debatewhich cannot be satisfactorily resolved here I suspect that signi®cantcurrents in American political thought can indeed be described asilliberal and that those currents, like others, have managed to use thelanguage of rights For present purposes, however, it is enough to saythat a liberalism broad enough to encompass all major currents inAmerican politics necessarily encompasses everything, or almosteverything, that is worth studying in American political discourse.Whether or not rights language can only support liberal ends, it canand has supported virtually every signi®cant agenda that has beenpursued in American politics

If it is not true, as both the critical legal scholars and thecommunitarians allege, that rights discourse systematically privileges

a certain kind of political outcome, it still might be true that arguing

in terms of rights is not an intellectually coherent way of framingpolitical or moral debate After all, the critical scholar might pointout, the litany of con¯icts above suggests that we cannot resolveissues by reference to rights, because rights so often con¯ict with oneanother There is a sense in which that critique is well made It istrue that rights con¯ict, so it is true that asking which side in adispute is the possessor of applicable rights will not always yield aclear resolution to the dispute It does not follow, however, that thelanguage of rights is an incoherent way of talking about politicalissues, because con¯icts among rights raise problems only if onebelieves that we should be able to settle substantive questions byreference to rights alone In other words, the charge that rightslanguage is an insuf®ciently good adjudicatory framework is signi®-cant only if one assumes that the sole or highest function of rightslanguage is adjudicatory That assumption is common, ironically, to

the liberal theorists who would use rights as means of adjudicationand the communitarian and critical legal scholars who object thatrights discourse is inadequate as an adjudicatory framework.

The language of rights is, however, at least as much about rhetoricand re¯ection as it is about adjudication Rhetoric is one of theobvious ways in which language and politics come together, and,considered as a species of political rhetoric, the practice of rightsdiscourse is entirely coherent As I argue throughout this book,analyzing rights discourse as a form of political rhetoric exposespatterns that leave the meanings of various rights claims more clearthan they would otherwise be There is a gap between the commonliteral understanding of statements like ``I have a right to privacy''and a different set of meanings that the statement carries, meaningsthat say more about the political values of the speaker than about anobjective moral or legal order and which may be all the better forthe refocusing, given the real de®ciencies of rights discourse as ameans of describing such orders Not all uses of rights language aremerely rhetorical, of course: careful re¯ections of judges and scholarsconcerned with rights often differ in character from the assertions ofagitators and political skirmishers Accordingly, the rules governingthe uses and best interpretations of rights language are not rigidlyidentical for all circles of discussion But no sharp line dividesjurisprudence or political philosophy from politics itself Indeed, thebest way to understand even the self-conscious rights philosophy ofsophisticated liberal theory is not as abstract philosophy alone butalso as political discourse

Not only does liberal rights theory have a political aspect, but itsuses of rights language are politically powerful One reason why theliberal theorists who use the language of rights are likely to defeattheir critics on the ®eld that really matters ± that is, the ®eld ofpolitics ± is precisely that they use the language of rights rather thandeconstructing it In the world of American political discourse, astrong claim of rights, any rights, tends to overpower a subtleintellectual argument about the shortcomings of rights as a concept.When I say that the political ®eld is the ®eld that matters, I saynothing with which those critics should disagree Critical legalstudies has long insisted that philosophy is inescapably political, andcommunitarian thinkers generally argue, in an aspirational vein,that the philosopher should strive to bring philosophy and politicstogether It is ironic that many of those who most explicitly argue for

the nexus between academics and politics cannot or do not use thepolitical value of academic discourse as effectively as the liberalswhom they critique That pattern is especially characteristic of thecritical legal scholars Some communitarians, in contrast, haveshown themselves more able to use the language of rights when it isuseful to advance their political arguments, irrespective of the factthat communitarianism also contains a strong strain critical of rightsdiscourse.2 Their willingness to use the language of rights greatlyincreases the possibility that their political arguments will be per-suasive to American audiences, and indeed communitarian ideas arenow ascendant in American political discourse Using the language

of rights may dilute communitarianism's intellectual distinctiveness,but it also contributes to the political potency of its platform

Similarly, the recent critical race theory movement in Americanlaw schools tries to support its political agenda by adding anaf®rmative view of rights to the critical legal studies paradigm It isnot yet clear how, if at all, a legal theory can coherently merge thedeconstructive posture of critical legal studies with the celebration oftraditionally conceived notions of rights My own suspicion is that itcannot be done, at least not from a purely philosophical perspectivewhich judges theories by their intellectual consistency Critical racetheory, however, is avowedly political, measuring its success not just

by those abstract philosophical standards but also by the concreteresults it can achieve for its vision of society.3 Just as the rights-af®rming communitarians have helped make communitarianism amore signi®cant political force, the use of rights language by criticalrace theorists may make them more powerful political players thantheir critical legal studies predecessors In both cases, the uses ofrights language raise questions about how much of a scholarlydiscourse should be understood as different from political rhetoricand how much should be interpreted in the ways appropriate for theanalysis of everyday political claims

This book, then, analyzes how politicians, lawyers, and phers in the United States have actually used the language of rights.Rights discourse, I argue, should be understood as a coherent social

philoso-2 Compare, for example, the attitudes toward rights of more politically engaged ians like Amitai Etzioni with those of more philosophical communitarians like Michael Sandel.

communitar-3 On critical race theory's framework and goals, see Angela Harris, ``Foreword: The Jurisprudence of Reconstruction,'' California Law Review 82, 741 (1994).

practice, available to a wide range of political agendas and includingwithin its scope the political argument of philosophers as well aspoliticians, though more sophisticated in some contexts than inothers At the crudest level, Americans often claim the satisfaction ofany needs or interests they consider important and deserving ofspecial protection as ``rights.'' These claims are limited in only twoways First, ``rights'' is an umbrella term encompassing entitlements,liberties, powers, and immunities, concepts that will be explicated inchapter 1 Second, ``There is a right X'' means approximately ``X isimportant and should be protected.'' No other principle, formal orsubstantive, underlies all the claims of rights that Americans advance

or even accept This understanding of what it means for something

to be a right might seem rough and permissive, but it does capturethe way that the concept of rights actually operates in much ofAmerican politics

In more disciplined rights philosophy, rights are often the sions or waystations of an argument rather than raw normativedeclarations Theorists who think about rights in a justi®catory wayrecognize that they must give arguments for why certain thingsshould be rights and that those arguments should not be reducible toemotivism, and theorists who reason in what is now called ``re¯ectiveequilibrium,'' revising in turn a set of tentative principles andtentative implications until a desirable system is found, can have

conclu-``rights'' stand in on either side of the ledger Rights so conceived arerevisable working constructs rather than ontological truths.4It is notthe case, however, that these sophistications make the patterns ofacademic rights discourse wholly different from the cruder rightsdiscourse of simple advocacy More re®ned versions of the same tworules described above do in fact capture the way the conceptoperates in much of political philosophy The four categories named

in the ®rst rule bound academic as well as non-academic rightsdiscourse; one difference between the discourses is that manyacademics as well as judges are loath to use ``rights'' indiscriminatelyacross the four analytic categories, preferring to restrict rights to one

or two among the four Many also try to narrow the second rule,developing theories ostensibly aimed at giving more speci®c content

to rights than ``X is important and should be protected.'' Both ofthose projects, however, have political undercurrents, because both

4 The idea of re¯ective equilibrium is discussed further in chapter 1.

have the effect of limiting the rhetorically powerful category of rights

to a subset of the propositions that it might be used to support.Theorists with different political commitments tend to producedifferent formal conceptions of rights which, in turn, support thedifferent commitments that suffuse the formal theories Decidingwhich formal conception of rights is best can therefore be much thesame as deciding which set of substantive political commitments isbest In other words, even sophisticated arguments about rights arefrequently surrogates for arguments about the substance of politics

To try to settle political questions by reference to theories of rights isproblematic at best, because theories of rights do not supplyevidence of an order prior to and regulative of politics They arepart of what is contested in political discourse, and the student ofpolitics can learn a great deal about the commitments of politicalactors, including political philosophers, by analyzing how they usethe language of rights

Not only do formal conceptions of rights re¯ect substantivepolitical commitments, but the substantive political commitmentsthey re¯ect change in response to changing historical conditions.Characteristically, the change has been reactive, as people respond

to a new set of problems by articulating a new set of rights I say

``reactive'' rather than ``adaptive'' because the major pattern ofdevelopment in American rights discourse has been one of concretenegation: innovations in conceptions of rights have chie¯y occurred

in opposition to new adversities, as people articulate new rights thatwould, if accepted, negate the crisis at hand In the ®rst years of theAmerican republic, for example, conceptions of rights were fre-quently grounded in opposition to the British colonial administra-tion In the nineteenth century, political and judicial conceptions ofrights in the American North were often framed in opposition to theslave-labor system of the South During and after World War II,Americans participating in every kind of rights discourse from thecrudest to the most re®ned rethought the content and form of rights

in ways inspired by opposition to European totalitarianism andespecially to Nazism At the end of the twentieth century, the rightsdiscourse of American law, politics, and political philosophy bearsthe mark of all of these in¯uences Thus, when the substantive aims

of rights discourse change, new rights and sometimes new theories ofrights arise to carry the content of the new political commitments.Moreover, the relationship between the substantive commitments

and the formal, conceptual aspects of rights theory is reciprocal:after new crises shape new commitments and new commitmentsshape new approaches to rights, linguistic and conceptual elements

of the new approaches to rights can in turn help shape substantivepolitical commitments

The language of rights, then, is a powerful and coherent methodnot just of re¯ecting but also of privileging substantive politicalcommitments, a method suitable for an extremely wide variety ofends and regularly employed by academics, politicians, lawyers, andvirtually everyone else who argues about political issues And yet,rights discourse has failings Notably, as the coming chapters willillustrate, it frequently permits obfuscation in political debate,because claims of rights often enable bare assertions to pass asreasoned arguments Perhaps that very feature, clearly a short-coming from a rationalist point of view, explains part of the lastingappeal that rights discourse has held in American politics Becauserights language offers advocates the possibility of presenting asser-tion as argument, people more concerned to advance a substantiveset of political ends than to illuminate a philosophical question areoften well served by harnessing the language of rights Suchadvocates sacri®ce precision of thought, and they expose themselves

to the charge that their arguments cannot, in the end, be grounded

in anything outside their own political commitments But that is nomore than to say that their politics are rooted in their politics, whichmay be the case already

There is here a temptation to say that this problem of circularityshould be avoided by not permitting rights language to obfuscatedebate and instead insisting that the true grounds underlying a claim

of rights should be considered and argued about directly If onetheory of rights relies upon a strong notion of individualism andanother on a normative conception of nature, perhaps it would bebetter to decide con¯icts between them not by arguing about rightsbut by asking about individualism and nature directly Unfortunately,

it would then be necessary to look for the grounds of individualismand naturalism, and, having grounded those concepts in yet otherconcepts whose grounds would have to be sought in turn, we mightsoon question what we really meant when we said we were lookingfor the ``true grounds'' of the normative argument in the ®rst place.Clifford Geertz told a now-famous story about being confronted by

a man who denied the reigning scienti®c understanding of planetary

orbits, insisting that the earth is in fact borne aloft on the back of anelephant The elephant, he said, was standing on a turtle Whenpressed as to what supported the turtle, he replied that it wasanother turtle; in fact, he explained, it was ``turtles all the waydown.''5Rights are not rights all the way down, but they may not bestanding a step or two from the bottom, either, perhaps becausethere is no bottom to stand upon and perhaps because it is unclearwhat a bottom would look like if we found one That rights cannot

be the ultimate justi®catory ground of a normative argument is thusnot a conclusive argument against using rights discourse in politics,because it may not be the nature of normative theory to yield thatkind of ground Knowing that rights are not themselves the grounds,however, does mean that we should be careful not to regard them assuch That caution is all the more important in a culture likeAmerica, where the language of rights carries substantial rhetoricalpower And it should be applied to the discourse of philosophers andjudges who think re¯ectively about rights as well as other kinds ofpolitical actors whose theories are thinner, because philosophers andjudges are partly political rhetoricians, perhaps inescapably so.Language and politics come together for them as well as for others,shaping their theories of rights

5 Clifford Geertz, The Interpretation of Cultures: Selected Essays (Basic Books, 1973), pp 28±29.

Rights theory and rights practice

When two people compete in a game of chess, they each try to winaccording to the same set of rules The means of achieving victoryare identical for both of them and known to both players in advance.They may ®nd in®nite ways of playing the game within the rules thatset permissible moves and victory conditions, but those rules andconditions are prior to the game Nothing that either player can dowould suddenly increase the size of the board, or permit one player

to move twice in a row, or let one player declare victory by, say,taking the other player's queen as opposed to the king The rules ofthe game are static and de®ned outside the play of the game itself;playing the game consists in adhering to those rules rather thanchallenging or trying to reshape them

Law and politics have their share of games or competitivesituations like games When a legislature or a court is going to decide

a controversial issue, advocates for rival outcomes use power,rhetoric, argument, and whatever else they can muster to try tosecure a favorable outcome They compete with one another, trying

to out-argue and outmaneuver their opponents, and the competitionamong them is a kind of game with certain patterns and restrictionsthat might be thought of as rules There might be a rule specifyingthat nothing will be a law that does not receive the support of amajority of some legislature, or that one cannot secure someone'svote by promising to pay him or her millions of dollars, or thatnobody may be convicted of treason without the testimony of twowitnesses One of the most important ways in which these games oflaw and politics differ from games like chess, however, concerns therelationship between the rules and the play of the game In politicaland legal argument, part of the contest is over how the issue indispute will be characterized and what kinds of arguments will count

as valid or superior When the same question could be presented as a

10

matter of free speech or a matter of community decency, it matterswhich presentation prevails; if something is agreed to be a matter offree speech, it matters what the decision-makers take ``free speech''

to mean The struggle to de®ne the grounds and terms of anargument is a struggle to set the rules of the game for a particularcontest As such, it is often the most important element of thecontest, because setting the rules can go a long way towarddetermining the outcome In legal and political discourse, then,shaping the rules is not something that happens before the game isplayed but is itself the subject of a contest, and attempts to shape therules are not preliminaries to the game but moves within the gameitself

In one of the most important books of recent legal theory, Law'sEmpire, Ronald Dworkin offers an account of law that recognizes thisinterconnection between conducting legal arguments and arguingabout what the rules of legal argument should be.1 Dworkin arguesthat legal interpretation, the activity required of judges, consists inmaking the law and the legal system the best that they can be, anddeciding which construction of the law makes the law the best that itcan be involves choosing some theory of how the law in generalshould be understood That choice is inescapably normative, and,once the choice is made, the normative theory chosen is supposed toset the rules for legal argument and interpretation Nevertheless, thechoice of theory should not be understood as occurring outside of orprior to legal argument itself Theorizing about law, Dworkin knows,

is part of legal argument; interpretive legal theories are not tions that stand outside the game but rather moves within the game.The descriptive and the prescriptive collapse on this model, astheories about the law are seen as attempts to construct law in oneway rather than another De®nitions of legal concepts or canons ofinterpretation therefore must be seen as part of, rather than prior to,the contest that is legal argument

descrip-It is therefore ironic that Dworkin's leading contribution to thetheory of rights is a de®nition of rights as a legal concept and thatDworkin sometimes treats that de®nition as regulating substantiveargument rather than being part of what legal and political argu-ment contests His de®nition is that a right is a metaphorical trumpcard, held by an individual, that can prevent the government or

1 Ronald Dworkin, Law's Empire (Harvard University Press, 1986).

Rights theory and rights practice 11

society at large from doing a certain thing, even if doing that thingwould be in society's general interest.2 Thus, if I have the right tofree speech, I cannot legitimately be silenced even if my keepingquiet would be better for society My right to speak trumps society'sinterest in my silence Some things simply cannot be done or denied

to individuals, Dworkin says, and we call the guarantees of thoseimperatives ``rights.''

Dworkin's de®nition holds a central place in contemporary rightstheory, perhaps because it so powerfully captures two prevailingintuitions about rights First, it ties rights to individualism Liberalslike Dworkin have placed protection of individuals against the will ofthe community at the center of their concerns for centuries; JohnStuart Mill's argument that there is a circle around every individualthat society may never invade and Immanuel Kant's images ofindividual dignity and the kingdom of ends are two easy examples.Accordingly, liberal theory has long associated rights with individu-alism That view of rights remains dominant today, both amongliberals like John Rawls and Joseph Raz who approve of rightsframeworks and critics of liberalism like Michael Sandel and MaryAnn Glendon who take more skeptical views.3 Dworkin's de®nition

of rights as individualist trumps admirably articulates this widelyshared idea about the nature of rights Second, the metaphor of thetrump card subtly acknowledges that rights can con¯ict with eachother A trump card, as the term is used in card games like bridge, is

a card that wins any round of play if no higher trump card is played,and so Dworkin means to say that possession of a right defeats anynon-rights-based considerations in a legal or political con¯ict Butthere is more to the metaphor Sometimes, in cards, more than onetrump card is played in a single round In those cases, the trumpwith the highest value prevails and the others, although trump cards,lose Dworkin's de®nition thus incorporates the legal realist andcritical legal studies criticism that rights can con¯ict with oneanother, but it does so while preserving the idea that rights provide acoherent framework for settling disputes

2 Ronald Dworkin, Taking Rights Seriously (Duckworth, 1977 [1991]), pp 91±93, 189±191, 269; Ronald Dworkin, A Matter of Principle (Oxford University Press, 1985), pp 2±3.

3 John Rawls, A Theory of Justice (Harvard University Press, 1971), pp 3±4; Joseph Raz, The Morality of Freedom (Oxford University Press, 1986), p 166; Michael Sandel, Democracy's Discontent: America in Search of a Public Philosophy (Harvard University Press, 1996), p 33; Mary Ann Glendon, Rights Talk: The Impoverishment of Political Discourse (Free Press, 1991), pp 47±48.

When Dworkin applies his de®nition to concrete cases, however,

he sometimes argues that the substance of rights must be or not becertain ways simply because those consequences follow from hisde®nition, as if the de®nition were evidence of the nature of rightsrather than an attempt to construct rights in one of several possibleways In so doing, he winks at Law's Empire's insight about de®nitionsand interpretations and presents his de®nition of rights as trumps asprior to the contest of rights discourse rather than a move within thediscourse itself Consider the argument that Dworkin makes aboutindividualism and a contested concept called ``the right to know.''The ``right to know,'' of course, is something that journalists claim

on behalf of society in support of their quest to discover and publishguarded information It is related to but not coextensive with theright of free press One who invokes the right to know claims thatthe public is entitled to have access to government documents orcourtroom records or whatever else the right is applied to, and theargument for the right to know is customarily advanced in terms ofempowering citizens to monitor the activities of government.Dworkin argues against the existence of such a right to know, and hisargument follows syllogistically from his de®nition Recall that onDworkin's de®nition, rights are things that only individuals canhave.4 It follows that no non-individual can use the language ofrights to protect its interests and that society itself, the very opposite

of the individual, cannot have any rights This stance is not asubstantive claim, Dworkin might say, but merely an analyticnecessity Nothing that is in society's interests can be a right, becauserights are by de®nition things that stand against the general interest

of society It is thus analytically senseless to speak of the rights ofsociety; indeed, Dworkin says that predicating rights of society is

``incoherent'' and ``bizarre.'' The right to know, however, is alleged

to be a right of society at large, and its application is alleged to begrounded in the general interest, not in protecting individualsagainst the general interest Dworkin therefore concludes that thosewho believe in the right to know are committing a category mistake

4 Dworkin does grant that ``individual'' should not be limited only to actual human individuals, saying that ``legal persons,'' such as corporations, may have rights as well (Taking Rights Seriously, p 91n.) Why he and other theorists like Joseph Raz make this concession is discussed further in chapter 3; as I argue there, it may be due to their preference, when confronted with data that their theories do not map well, to rede®ne the data rather than change their de®nitions This nuance does not, however, affect the current argument.

Rights theory and rights practice 13

Given that rights attach to individuals and not to society, Dworkineasily concludes that there cannot possibly be such a thing as theright to know.5

The proof is entirely formal; the conclusion that the public has noright to know is entailed within Dworkin's de®nition of rights Hisargument against a public right to know need not and does notweigh the substantive questions of public access to information.Dworkin does not, for example, make an argument about the meritsand demerits of allowing television cameras inside courtrooms But

by arguing that there is no public right to know, Dworkin promotes aparticular answer to the question of whether trials should betelevised The answer he promotes is ``No.'' Technically, it is stillpossible for Dworkin to take either side on the substantive question

He could claim that television cameras should not be permitted incourtrooms and that there is no right to know that such a ban wouldviolate, or he could claim that television cameras should bepermitted in courtrooms, though on grounds other than that of aright to know It would, however, be a mistake to give too muchweight to this last possibility In the context of American rightsdiscourse, to declare that there is no such thing as the right to know

is, at least presumptively, to take sides against having cameras incourtrooms At the very least, it is to weaken the argument fortelevising trials by denying its articulators the use of a powerfulrhetorical tool: the language of rights

It may be that Dworkin opposes the uses to which the putative

``right to know'' is put, in which case his argument against the right

to know is convenient to his purpose It may even be that some ofthose uses, such as televising trials, are pernicious and deserve ouropposition Nevertheless, Dworkin's case against the right to know isnot a good way to make the point Doing nothing more than tracingthe logical entailments of a de®nition, it neatly dismisses thepossibility of a right to know and, because we know that the right toknow means certain things about cameras and reporters, encourages

us to infer conclusions about televising trials and printing

govern-5 Dworkin, A Matter of Principle, pp 387±388 When Dworkin and others use de®nitions of rights to ``disprove'' the existence of certain rights, they mean that those rights do not or cannot exist in a moral sense, prior to the law Dworkin would not contest that a proposition codi®ed into law as a right would be a right, e.g., that if a legislature enacted a ``right to know statute,'' a right to know would then exist in that jurisdiction as provided in the law None of the theorists I discuss denies that rights talk is sometimes simple legal positivism.

ment records Furthermore, it does so without showing that suchpublicity would be against the general welfare, or unjusti®ablyharmful to particular people, or unjusti®able on any other substan-tive grounds Instead, a purely formal de®nition of rights fosters asubstantive position on issues of public access to information.

The formal de®nition, however, is not a rule that pre-exists thegame of legal argument It is a move within the game As Dworkin'sown theory of adjudication and interpretation explains, his ``de®ni-tion'' of rights is not just descriptive or constitutive of some aspect ofthe game (like ``the chessboard measures eight squares by eightsquares'') but an attempt to prescribe, as among multiple possi-bilities, how that aspect will function in the game Dworkin's politicalcommitments, including his commitment to individualism, arepresent in his de®nition of rights, as Law's Empire alerts us to expect

It would be ironic if Dworkin, when making his argument againstthe right to know, forgot that de®nitions and interpretations aremoves within the game and always carry substantive commitments,such that it is dangerous to treat de®nitions of normative concepts as

®xed truths ± in a word, as de®nitive It is more probable thatDworkin knows his de®nition to be a move within the game of legaldiscourse and that he does not bother to acknowledge it as such Hedoes not preface his argument by saying ``This de®nition of rights isitself subject to challenge, because it is only my attempt to constructthe category in the way that I, subjectively and normatively, believemakes rights the best that they can be.'' Instead, he simply offers hisde®nition and winks at his theory of interpretation He knows thatwhat he presents as simply descriptive is actually normative, but hemakes his move without calling our attention to the fact

Insofar as Dworkin is a player in the game, concerned withestablishing or refuting a speci®c right like the right to know,winking at the interpretive insights of Law's Empire and forgingahead with a de®nition of rights is an effective tactic Becauseeveryone agrees that rights have force, embedding political commit-ments within a de®nition of rights is an excellent way to tiparguments in their favor If I argue that I should be permitted to do

X, I may or may not win the argument and get to do X If I arguethat I should be permitted to do X and point out that I have a right

to do X, my chances of winning the argument and getting to do Xare greater than if I make the argument without reference to myrights The same is true if I argue that respecting some other right Y

Rights theory and rights practice 15

necessarily entails my being allowed to do X As part of what it means

to be a right, each political commitment within a de®nition of rights(e.g., the commitment that people should be allowed to do X) travelsunder a privileged banner Extra weight is given to the argumentbecause it is an argument not just about X but about rights

At the same time, however, something should give us pause aboutDworkin's using his de®nition of rights as he does when he dismissesthe right to know Given his presumed awareness that de®nitionsand interpretations of legal concepts contain normative judgments,

it does not seem entirely right for Dworkin to argue from a de®nitionthat he presents as no more than descriptive We might expect himinstead to use the case of the right to know to test his de®nition,perhaps by asking whether dismissing the right to know makes thelaw the best that it can be If it did, then both his de®nition and hisargument about the right to know would be strengthened That kind

of analysis would require Dworkin to engage substantive questionssuch as whose interests are served and harmed by placing televisioncameras in courtrooms or prohibiting photographers from snooping

on celebrities Instead, however, Dworkin bypasses all such questionsand rests his argument on the de®nition of rights alone In so doing,

he relieves himself of having to defend a set of normative ments by cloaking them in the banner of rights, a category he hasappropriated for the purpose But it does not make sense to let thebanner do the persuasive work if the commitments it contains couldnot do the same work on their own Similarly, if certain propositions

commit-do not travel under the banner of rights, and if their not travelingunder that banner is due only to the way that rights have beende®ned, then we have no reason to suppose that those propositionsare less compelling than the commitments included within the goingde®nition of rights

Appropriating rights language for a particular set of substantivepolitical commitments is a widespread feature of rights discourse,political as well as academic Consider the rival approaches to rightsfound in ``will theories'' and ``interest theories,'' each of which builds

a distinct set of normative choices into its conception of rights andthen argues for its positions based partly on the strength of de®ni-tions.6``Will theories'' of rights, which are sometimes called ``option

6 For the early history of the rivalry between these two ways of seeing rights, see Richard Tuck, Natural Rights Theories (Cambridge University Press, 1979).

theories'' or ``choice theories,'' have roots in Hobbes and areexpressed in the writings of Wesley Hohfeld and H L A Hart.7

Roughly, will theories de®ne having a right as having an opportunity

to make a choice Interest theories of rights, which are sometimescalled ``welfare theories'' or ``bene®t theories,'' promote a differentde®nition, according to which one has a right if a condition of one'swell-being is suf®ciently important to place someone under a duty.Joseph Raz, Joel Feinberg, and Neil MacCormick are interesttheorists.8 Will theories elevate the value of autonomy, and interesttheories elevate other aspects of well-being

Sometimes, a theorist of one of these schools will attack the other

by showing that some desirable right is not possible on its terms Forexample, MacCormick attacks the will theory by arguing thatchildren should be provided with basic care and nutrition forreasons having nothing to do with anyone's choices or autonomy,including that of the children Because the imperative to care for achild does not derive from the child's autonomous choices, a theorythat equated rights with opportunities for making autonomouschoices would not include a right of children to basic care MacCor-mick continues: ``Either we abstain from ascribing to children a right

to care and nurture, or we abandon the will theory For my part, Ihave no inhibitions about abandoning the latter.''9 Even if thisargument successfully shows that the will theory cannot account forall rights, the conclusion that the interest theory is thereby estab-lished is based on a false choice There is no reason why all rightsmust be grounded only in autonomy or only in welfare; some rightscan derive from respect for human choices, and others can derivefrom other kinds of needs To de®ne the ground of rights exclusively

in either criterion is to load later arguments in favor of a set ofsubstantive political commitments, either those prizing autonomy orthose prizing other conditions of human well-being.10

7 Wesley Hohfeld, Fundamental Legal Conceptions as Applied in Judicial Reasoning (Greenwood Press, 1978); H L A Hart, ``Are There Any Natural Rights?'' in Jeremy Waldron, Theories

of Rights (Oxford University Press, 1984), pp 77±90 See also Michael Freeden, Rights (Open University Press, 1991), pp 43±49.

8 Raz, The Morality of Freedom, p 166; Joel Feinberg, Rights, Justice, and the Bounds of Liberty (Princeton University Press, 1980), p 209; Neil MacCormick, Legal Right and Social Democracy (Clarendon Press, 1982), pp 143±160.

9 MacCormick, Legal Right, p 158.

10 For an argument that is the mirror image of MacCormick's, presenting a dichotomous choice between will theories and interest theories and preferring the former, see Hillel Steiner, An Essay on Rights (Blackwell, 1994), pp 62±73 Steiner applies his theory to the

Rights theory and rights practice 17

Just as rights theorists sometimes attack one conception of rights

in the attempt to establish another, theorists critical of rights ingeneral sometimes try to attack the entire concept of rights bycriticizing a single conception of rights and presenting the short-comings of the conception attacked as if they were problems withrights as a general category.11 The conception of rights mostcommonly targeted in this kind of attack is probably the strongindividualist-based notion of rights that Dworkin's theory exempli-

®es That conception is dominant in contemporary rights thinking,and it is often easy to pass that dominant conception off as theconcept of rights itself rather than one version only Attacking rights

by attacking that conception of rights is a staple among somecommunitarian theorists Being skeptical of individualism to beginwith, they identify rights with excessive individualism and condemnrights accordingly

Consider, for example, Michael Sandel's stance toward ian rights In Democracy's Discontent, Sandel routinely merges ``rights''with individual rights, discussing rights in American history as ifrights had always and only been imagined in Dworkin's fashion, that

Dworkin-is, as attaching only to individuals.12 As a matter of history, thatpresentation is lacking: as I discuss in the coming chapters, Americanrights discourse has often predicated rights of entities other thanindividuals Sandel's collapsing of all rights into individual rights,however, helps explain his hostility to rights as an outgrowth of hisviews on liberal individualism Indeed, Sandel's attack on rightsprominently features a self-conscious attack on the Dworkinian view.The idea that rights are trumps is the theme, Sandel charges, of one

of the most infamous court decisions in American history: Lochner v.New York, the 1905 case in which the Supreme Court struck down amaximum-hours law for bakers on the ground that it violated every

question of children's rights, reaching conclusions directly opposite MacCormick's, at

pp 245f.

11 A word is in order here on the difference, as I am using the terms, between concepts and conceptions Following Dworkin and Hart, I use ``concept'' to refer to broad categories in political and legal thought generally, such as the concept of rights, or of equality, or of democracy The meanings of those concepts, however, are contested by politicians and theorists; what rights or equality or democracy means to adherents of one political party or philosophy may differ from what it means to others The rival meanings or interpretations

of those concepts are what I refer to with the term ``conceptions.'' Particular conceptions impart more speci®c meanings to capacious concepts For example, we could say that equality is a concept of which liberals and Marxists hold different conceptions.

12 Sandel, Democracy's Discontent, e.g., p 33.

individual's right to freedom of contract.13 The decision consignedbakers and untold numbers of other laborers to work seventy oreighty hours a week in unhealthy conditions, unable to seek regula-tion and relief through the political process In the grand narrative

of American constitutional development, Lochner symbolizes the lawgone bad For more than sixty years, lawyers and judges have knownthat ``Lochnerizing'' is a conceptual sin of the ®rst order WhenSandel associates rights as trumps with Lochner ± rather than withwidely approved decisions on issues like free speech or privacy ± hetars Dworkin's theory with a very large brush And because he hasmerged rights as trumps with rights in general, his attack onDworkin-style individual rights appears as an attack on rights as awhole

It does not follow from Sandel's substantive views that he mustattack rights as he does Rather than con®ning rights to anindividualist conception of which he disapproves and then deni-grating the concept wholesale, he might have chosen to advance adifferent conception of rights, one that would incorporate normativecommitments that he preferred Sandel knows that theories of rightsalways embody some set of normative commitments, or, as he puts

it, some vision of the good.14Republicans interpret rights according

to republican principles, he correctly notes, and liberals interpretrights according to liberal principles It is not clear, therefore, whySandel seems not to think that he can interpret rights in light of hisown conception of the good Surely, it cannot be a good argumenta-tive strategy to attack a popular idea like ``rights'' when one has theoption of appropriating it instead Nevertheless, Sandel declines thatopportunity He, like many other theorists, seizes on one strain inrights theory and treats it as rights theory in general, the onlydifference being that he does so not in order to establish that theorybut to condemn rights as a whole

feinberg, raz, and the rights of human vegetablesWhat Sandel sees correctly about rights, however, is that claims ofrights are inescapably normative, because rights are always inter-preted according to some vision of the good or set of substantivepolitical commitments Some theorists of rights refuse to see this

13 Ibid., p 42 14 Ibid., p 321.

Feinberg, Raz, and human vegetables 19

aspect of rights discourse, arguing as if the existence of rights could

in some cases be a purely descriptive matter Consider Joel Feinberg,who holds an ``interest theory'' of rights Feinberg's theory reasonsfrom formal de®nitions to substantive conclusions in a way similar toDworkin's theory, but it shows a different face of that technique ``Tohave a right,'' says Feinberg, ``is to have a claim to something andagainst someone.''15 Feinberg argues that claims are bound up withinterests and further proposes that ``only beings who have interestsare conceptually suitable subjects for the attribution of rights.''16

Rocks, for example, cannot have rights, because rocks do not haveinterests Animals, Feinberg says, do have interests and do haverights The argument works like this: animals prefer to be treatedsome ways and not others, and their preferences are tantamount tointerests Given that animals have interests, they are ``conceptuallysuitable'' to be rights bearers It is true that animals cannot assertclaims in support of their interests, but that is just because theycannot speak What they lack is the ability to assert, not the capacityfor having claims in a moral sense Feinberg therefore concludes thatanimals, as suitable rights bearers who have claims, have rights.17

In contrast, Feinberg argues that human vegetables have no rights

at all If assumed to be incurable, he says, human vegetables cannot

be said to have interests Under his de®nition, that makes themconceptually unsuitable to be rights bearers, so Feinberg concludesthat they cannot have rights, but he balks at the normative impli-cations of that conclusion.18 He contends that the fact that humanvegetables cannot and do not have rights is not a license to treatthem in any malevolent or destructive way one might choose Heknows that people might interpret him as saying that humanvegetables may be legitimately killed, warehoused, or who knowswhat else, and he is morally uncomfortable with that implication Hetherefore explicitly denies that his arguments about rights-bearinghave any kind of moral impact According to Feinberg, whetherhuman vegetables ``are the kind of beings that can have rights [is] aconceptual, not a moral question, amenable only to what is called

`logical analysis,' and irrelevant to moral judgment.''19

The claim that rights analysis is ``conceptual'' and not ``moral'' isquite comprehensible within the framework of analytic inquiry

15 Feinberg, Rights, p 159 Emphasis in original 16 Ibid., p 209.

17 Ibid., pp 159±267 18 Ibid., pp 176±177 19 Ibid., pp 180, 213.

Feinberg is claiming to do no more than show what consequencesfollow from a set of de®nitions In a way, he is urging us not toreason from his formal de®nitions of rights to substantive conclu-sions about political morality, because that would confuse theconceptual with the moral When he argues that animals have rightsand human vegetables do not, Feinberg says, he is not telling us what

to do when confronted with political or moral decisions But in thatcase, it is hard to understand what the argument is about If animals'having or not having rights entails no consequences for action orappraisal of action, whether or not they have rights is of littleimportance Indeed, if the de®nitions Feinberg develops and ana-lyzes bear no connection to the world of normative decisions, it isnot clear why we should be interested in his arguments at all

These last implications do not really need to be addressed,because Feinberg's arguments are, despite his protestation to thecontrary, inescapably normative Saying that some class of beings(animals, human vegetables, fetuses, ``the public'') can or cannothave rights is a political and a moral act, not just an analytic one.The concept of rights is one of the constituent concepts of politics inFeinberg's society, and proposals to prefer one or another under-standing of such constitutive concepts are necessarily political acts.20

On Dworkin's model, such proposals are about interpreting theconstitutive concepts and therefore must be normative, because theyinvolve deciding which understanding makes the constitutiveconcept the best that it can be In a similar vein, Quentin Skinnerhas argued that a dispute over the applicability of an appraisive termcannot be only a linguistic or a semantic dispute It is a political ormoral dispute as well.21 In America, ``rights'' is an appraisive term,among the most appraisive of all To say that X has a right to Y is tomake a normative statement about the relationship between X and

Y, and only if the term could be divested of its normative meaningscould its employment be non-normative That divestment is prob-ably not possible, and if theorists really could divest the term

``rights'' of its normative meaning, it is doubtful that we woulddiscuss their use of the term at all Other than to express normative

20 On the political nature of contesting concepts, see William Connolly, The Terms of Political Discourse (Basil Blackwell, 3rd edn., 1993), pp 39±40, 180.

21 Quentin Skinner, ``Language and Political Change,'' in Terence Ball, James Farr, and Russell L Hanson, eds., Political Innovation and Conceptual Change (Cambridge University Press, 1989), pp 6±23.

Feinberg, Raz, and human vegetables 21

views of the kind that ``X has the right to Y'' expresses, we have littleuse for the term ``a right.'' It is unlikely, then, that Feinberg isactually interested in divesting ``rights'' of normative meaning What

he is interested in, I suspect, is ®nding a way to mitigate theunpleasant conclusion that human vegetables do not have rights, aconclusion which seems to open the way to deliberate slaughter.Rather than defend that position, he ducks by denying the argu-ment's relevance to morality

It is here interesting to compare Feinberg's view of rights with that

of Joseph Raz Raz, like Feinberg, is an interest-based rights theorist,and their de®nitions of rights are almost the same Raz argues that

X has a right if some aspect of X's well-being, alternately formulated

as X's interest, is a suf®cient reason for holding someone else to beunder a duty, and Feinberg de®nes rights as claims to something andagainst someone Both de®nitions hold that rights are based oninterests and that interests give rise to rights when they areimportant enough to justify imposing a duty on some other party.One key difference between them, however, is that where Feinbergsays that only beings with interests can have rights, Raz says that abeing can have rights if its well-being is of ultimate value.22 Thisprovision would let Raz argue that, contrary to Feinberg's conclu-sion, human vegetables can have rights All he would have to say isthat the well-being of humans is of ultimate value, which he certainlybelieves, and that human vegetables are human It seems likely thatRaz would be more comfortable with that conclusion than Feinberg

is with his, because Raz could conclude that human vegetables dohave rights and thereby avoid the implication that it is legitimate tokill or warehouse them Raz can reach this preferred conclusionbecause he loaded his ``formal'' de®nition of rights and rights-bearing with more of his important substantive moral commitmentsthan Feinberg embedded in his

The problem that provokes Feinberg's unsuccessful attempt toescape from the normative implications of his argument is similar tothe problem with Dworkin's argument against the right to know Ineach case, a rights theorist analyzes possible rights by comparingthem to a formal de®nition In each case, the rights in question areincompatible with the de®nition and accordingly pronounced non-existent Dworkin's argument reaches a desired conclusion by

22 Raz, The Morality of Freedom, pp 166±180.

building it into his premises, and Feinberg's pushes him into aconclusion that he would not have chosen Both problems stem fromthe same source: the tendency to infer conclusions about thesubstance of rights from de®nitions of their form.

three kinds of definition

I suggest that these problems are unnecessary They occur only if webelieve that formal de®nitions of rights regulate particular rights,and that belief seems unwarranted more often than not Considerthat a political theorist who offers a de®nition of ``rights'' might bedoing any of three different things First, he might be asserting thatthere exists an ontological category of moral imperatives called

``rights'' and that the de®nition offered speci®es the properties thatall members of that category possess This approach to de®nition ischaracteristic of Platonism As a second alternative, he might begeneralizing from a set of desirable normative abstractions, trying toidentify principles that would support a worthy set of rights ifadopted as de®nitive of the category He could reason back and forthbetween the particular desirable norms and the general principlesuntil he found a set of norms and principles that ®t well with eachother This approach to de®nition resembles the notion of ``re¯ectiveequilibrium'' as pioneered by Nelson Goodman and made famous inthe work of John Rawls.23 A third possibility is that he is trying toexplain how the language and the concept of rights functions insome political discourse, that is, what it means within some set oflinguistic practices to call something a right This approach ischaracteristic of the later work of Ludwig Wittgenstein

A formal de®nition of rights that purported to regulate thepossible content of particular rights would have to be a de®nition ofthe ®rst or second kind A de®nition of the ®rst, ontological kindwould regulate rights in the simplest possible way, stating un-wavering criteria for all rights A de®nition offered in re¯ectiveequilibrium would regulate the category of rights less rigidly, because

it would leave open the possibility that the de®nition could itself berevised, but the de®nition would still purport to de®ne the category

as nearly as possible and would be more successful the more it was

23 Rawls, A Theory of Justice, pp 48±50; Nelson Goodman, Fact, Fiction, and Forecast (Harvard University Press, 1955), pp 65±68.

Three kinds of de®nition 23

able con®dently to deem propositions ``rights'' or ``not rights.'' Incontrast, a de®nition of the third kind could not exclude certainthings from the category of ``rights'' on the basis of content, becausethe third kind of de®nition is attendant on actual uses of the term

``rights.'' Whether de®nitions like Dworkin's and Feinberg's are ofthe ®rst or second kind is not always clear Feinberg sometimesgestures toward the ontological mode, as when he asserts that at thecore of human dignity lies a set of ``facts about the possession ofrights,''24and both Feinberg and Dworkin treat their de®nitions as ifthey were ontological in their arguments about human vegetablesand the right to know described above Nevertheless, re¯ectiveequilibrium is perhaps the favorite mode of thought among sophisti-cated modern theorists, and it may be reasonable to presume thatDworkin and Feinberg mean their de®nitions to be so understood If

it is not clear whether their de®nitions are ontological or re¯ective,however, it is clear that they are one or the other, because theirarguments reason from de®nitions to the conclusion that a certainkind of thing cannot be a right, even though people talk about it as if

it were a right I suggest that neither of those approaches tode®nition offers the best way to understand the nature of rights andrights claims As I will discuss below, the ®rst approach is concep-tually problematic and the second systematically misses importantaspects of rights discourse In explicating rights in the context ofAmerican politics, I make use of the third approach

Let us consider the ontological approach ®rst Moral and politicaltheorists who view rights this way try to identify the formal attributes

of all rights irrespective of the normative content of particular rights

``Rights,'' on this understanding, is the name of a pre-existingcategory of moral imperatives, and the quest to identify the proper-ties of rights is the attempt to identify the criteria for inclusion in thecategory That project presumes not only that certain moral impera-tives exist a priori but also that they necessarily exist as ``rights.'' Herethe project becomes problematic Perhaps some moral imperatives

do exist a priori, but the categories with which we organize moralimperatives tend to be linguistically constructed Indeed, it is acentral insight of pragmatist philosophers from William James to

W V O Quine and Donald Davidson as well as social scientists likeMax Weber that human construction rather than natural ordering

24 Feinberg, Rights, p 151 Emphasis in original.