1 Back to Bundles: Deflating Property Rights, Again

8/12/2019 1 Back to Bundles: Deflating Property Rights, Again

1/25

Legal Theoryhttp://journals.cambridge.org/LEG

Additional services for Legal Theory:

Email alerts: Click hereSubscriptions: Click hereCommercial reprints: Click hereTerms of use : Click here

BACK TO BUNDLES: DEFLATING PROPERTYRIGHTS, AGAIN

Shane Nicholas Glackin

Legal Theory / Volume 20 / Issue 01 / March 2014, pp 1 - 24DOI: 10.1017/S1352325213000153, Published online: 04 February 2014

Link to this article: http://journals.cambridge.org/abstract_S1352325213000153

How to cite this article:Shane Nicholas Glackin (2014). BACK TO BUNDLES: DEFLATING PROPERTYRIGHTS, AGAIN . Legal Theory, 20, pp 1-24 doi:10.1017/S1352325213000153

Request Permissions : Click here

Downloaded from http://journals.cambridge.org/LEG, IP address: 130.217.227.3 on 10 Jul 2014


2/25http://journals.cambridge.org Downloaded: 10 Jul 2014 IP address: 130.217.227.3

Legal Theory , 20 (2014), 124.c Cambridge University Press 2014 0361-6843/14 $15 .00 + 00

doi:10.1017/S1352325213000153

BACK TO BUNDLES: DEFLATINGPROPERTY RIGHTS, AGAIN

Shane Nicholas GlackinUniversity of Exeter

Following Wesley Hohfelds pioneering analyses, which demonstrated that the folkconcept of ownership conated a variety of distinct legal relations, a deationary

bundle theory regarding those relations as essentially unconnected held sway formuch of the subsequent century. In recent decades, this theory has been thoughttoo diffuse; it seems counterintuitive to insist, for instance, that rights of possessionand alienation over a property are associated only contingently. Accordingly, scholarssuch as James Penner and James Harris have advanced theories that revive the conceptof ownership, identifying some instances of property as paradigmatic,and regardingothers as conceptually subsidiary. I propose a new interpretation of the bundle theory,based on David Lewiss idea of Humean supervenience among physical particles. Icritically examine the major antibundle positions, arguing that their criticisms resultfrom confusion about the claims of the bundle theory, which remains the best account

of property rights available.

World is crazier and more of it than we think,Incorrigibly plural . . . .The drunkenness of things being various.

Louis MacNeice, Snow

I. INTRODUCTION

My aim in this paper may, at rst glance, strike the reader as somewhat odd.It is a defense of a theory of property rights that after all has been prevalent among legal theorists for most of the last century and that is taught asa matter of routine in most undergraduate property-law courses in orderto disabuse entering law students of their primitive lay notions regardingownership. 1

Early versions of this paper were presented to MLS and LLM students at the NationalUniversity of Ireland, Maynooth, and to the Irish Jurisprudence Society. I am grateful tothose audiences, as well as to Sibo Banda, Garrett Barden, Brendan Curran, Brian Flanagan, John Glackin, Gerald Lang, Garret Ledwith, Neil Maddox, Tanya Sheridan, Rachael Walsh,and Robbie Williams for their helpful discussion of this material. I am also indebted to theanonymous reviewers for L EGAL THEORY , whose suggestions greatly improved the nishedarticle.

1. BRUCE A CKERMAN, PRIVATE PROPERTY AND THE CONSTITUTION (1977), at 26.

1
http://journals.cambridge.org/http://journals.cambridge.org/



2 SHANE N ICHOLAS G LACKIN

Yet those primitive lay notions have been reinforced in recent years by several decidedly sophisticated legal theories. Among contemporary theo-rists, the previous orthodoxy is under siege, with a substantial faction now holding that the so-called bundle theory of property rights is no longertenable and that a robust relation of ownership of things can and should bereestablished. 2 The bundle theory may serve the purposes of the dimmest law student, 3 according to this new consensus, but it will not serve an ade-quate account of judicial and lay reasoning about property rights. Accord-ingly, the burden of proof rests once again upon those of us who advance adeationary analysis of such rights.

What I intend is, indeed, to deate the somewhat mystical force that isseemingly taken by layman and neoproperty theorist alike both to unify the various legal relations that we recognize as falling under the rubricsof property or ownership , and to imbue them with a distinctive moral status.The apparent unity of those concepts, 4 I argue here, is illusory and largely the result of historical accident. Whatever normative force we associate withthe concepts, moreover, is properly attributed to their component parts; nomoral or legal conclusion can or should be drawn from the declaration that some thing is property or that some person is its owner.

I begin the paper by sketching briey the outlines of theor at any rate,abundle theory of property rights. I do not go into any great detail inexpounding a particular version of it; there are, of course, as many different bundle theories as there are theorists, so what I aim to present here is a way to understand the common conceptual core of such theories. In subsequent sections, I consider and reject two of the most inuential objections andalternatives to the bundle theory, advanced by James Penner and J.W. Harris.Mygrounds for doing so give, I hope, somefurther indication of what I thinkthe contours of a defensible bundle theory to be.

I do not discuss the wider arguments made by these writers about eco-nomic and distributive justice. Though sympathetic to many of their con-cerns, I doubt that successful redistributive arguments are dependent onthe particular content of any distinctive and robust account of property orownership. Indeed, I suspect that stipulating such accounts of property andownership may amount in some cases to trying to acquire redistributive

2. Hugh Breakey, Two Concepts of Property: Ownership of Things and Property in Activities , 42PHIL. F. 239266 (2011), at 240242. The new consensus is also evident in a recent symposiumon the subject; see Daniel Klein & John Robinson, Property: A Bundle of Rights? Prologue to the Property Symposium , 8 ECON J. W ATCH 193204 (2011).

3. A CKERMAN, supra note 1, at 26.4. Some of the philosophical literature on property is concerned with the nature of con-

cepts and with the distinction, if any, between these and ideas; for an excellent discussionand overview of these issues, see Stephen R. Munzer, Property and Disagreement , in PHILOSOPHICALFOUNDATIONS OF PROPERTY L AW (J. Penner & H.E. Smith eds., 2013). While I have a view onthis question, I do not believe that anything in the present argument hangs on it; and it seemsmethodologically sound to me, as far as is possible, to try to keep these two rather dissimilarsets of philosophical issues separate from each other. I am grateful to an anonymous refereefor bringing this apparent lacuna to my attention.



Back to Bundles 3

justice on the conceptual cheap, avoiding the hard work of detailed moraland political argument. But those concerns lie well beyond the scope of this paper. My own belief is that demystifying the notion of property andownership would allow more rational public discourse about resource allo-cation, environmental protection, social justice, and so on; but while that is a motivation for the theory I develop here, it cannot be an argument infavor of its truth. 5

II. HOHFELD AND THE BUNDLE THEORY

The bundle theory, in its modern form, is primarily the result of work by two gures: Wesley Hohfeld 6 and A.M. (Tony) Honor e.7 Hohfeld makes

pioneering and perhaps unprecedented efforts to impose a measure of logical rigor on legal concepts in general, and the inuence of these effortshas proved especially lasting in rights theory; Honor es writings on theconcept of property are among his most signicant contributions to anextensive range of topics in legal philosophy and general jurisprudence. Idiscuss the relevance of both to my argument after briey outlining what Itake to be the bundle theorys key features.

Although now inextricably linked with Hohfeld and Honor e, the notionof property rights as comprising a bundle of sticks predates their work. 8

The bundle theory holds, at its most basic, that the right of a property owner is separable into a series of component right-parts; my ownershipof some chattel may comprise, inter alia, a right of exclusion, a right of use, a right of possession, and a right of alienation, none of which is con-ceptually dependent on any of the others. 9 The doctrine of estates in realproperty introduces further, more exotic interests; I may acquire the right

5. In the words of a fellow enthusiast; the bundle theory is an analytical schemeapplicableto many legal systems and the property arrangements within them. A virtue of such a schemeis that it makes few if any moral or political commitments. The analysis of property law is onething and proposals for its reform are quite another. Stephen R. Munzer, A Bundle Theorist Holds onto His Collection of Sticks 8 ECON J. W ATCH 265273 (2011), at 269.

6. W ESLEY N. HOHFELD , FUNDAMENTAL LEGAL CONCEPTIONS AS A PPLIED IN JUDICIAL R EASON-ING: A ND OTHER ESSAYS (1923). Since the only commercially available editions of Hohfelds work at this time are facsimile reprints of at best intermittent legibility, the reader is invited tocontest the veracity of all passages quoted from Hohfelds work in this paper.

7. A.M. HONOR E, Ownership , in M AKING L AWS BIND 161192 (1987). Opinions differ as to whether Honor e, who sought to outline the liberal concept of full individual ownership( id.), can properly be grouped with the bundle theorists; at any rate, the bundle theorists whofollowed made extensive use of his schema.

8. The rst known use of the term in this context is in J OHN LEWIS, A TREATISE ON THEL AW OF EMINENT DOMAIN IN THE UNITED STATES (1888), at 43; The dullest individual amongthe people knows and understands that his property in anything is a bundle of rights. The rst metaphorical reference to a bundle of sticks , however, is BENJAMIN N. C ARDOZO, THE P ARADOXESOF LEGAL SCIENCE (1928), at 129: The bundle of power and privileges to which we give thename of ownership is not constant through the ages. The faggots must be put together andrebound from time to time.

9. See, e.g., Justice Mosks dissent in Moore v. Regents of the Univ. of Cal., 793 P.2d Cal.479, 509510 (1990).




of occupation in an apartment from you, perhaps supplemented by vari-ous easements and rights of access, though my right to alienate it may berestricted by the terms of the license. You may in turn have only a life in-terest in that apartment, with a reversion vesting in a relative upon yourdeath. These arrangements are subject to revision over time in responseto economic and social circumstances and can thus vary substantially from jurisdiction to jurisdiction, with many local peculiarities developing. Thecommon law envisages, for instance, such collective provisions as those of the Israeli kibbutznik and the American Hutterite; 10 if the property is Irishand the owner indebted, the High Court may encumber it with a judge-ment mortgage 11

The bundle theory regards these individual and separable rights, orsticks, as having no substantive, essential connection to each other. Tothe extent that any two or more of them tend to accompany each other,their conjunction is contingent rather than intrinsic; from a logical point of view, we might as easily have tied any other combination of the possibleinterests in the chattel or property together in the bundle. In our ordinary commercial transactions, we can and do take these bundles apart, redis-tributing the sticks among others bundles, and replacing them with new ones. 12

Who, then, is the owner of some particular piece of property? Thequestion, for a bundle theorist, involves something like what Gilbert Ryle would have termed a category error 13 Since the number and kind of suchinterests in the property, all vested in different individuals, may be almost unlimited depending on the legal jurisdiction, it makes little intuitive senseto pick out any one interest bearer as the owner. No particular interest orright, and moreover no particular combination of those interests and rights,is either necessary or sufcient to establish ownership. 14 The lawyer, as Jeremy Waldron notes, will not be interested in nding out which of [the

10. See, e.g., Robert C. Ellickson, Property in Land , 102 Y ALE L.J. 1315 (1993), at 13461348.11. See J.C.W. W YLIE, IRISH L AND L AW (4th ed. 2010), at 867884.12. In practice, thenumberand form of such fragmentations and redistributionspermitted

is limited by theso-called numerus clausus principle, principally as a result of the in rem nature of property rights that will play a large part in this discussion; see T.W. Merrill& H.E. Smith, Optimal Standardization in the Law of Property: The Numerus Clausus Principle , 110 Y ALE L.J. 170 (2000);and Joseph Singer, Democratic Estates: Property Law in a Free and Democratic Society , 94 CORNELL L.R EV . 10091062 (2009), at 10211029. For a beguiling account of the complexities of property-right arrangements, however (and of the degree to which the lay publics understanding of their subtleties often outstrips that of legal specialists), cf. Robert C. Elickson, Of Coase and Cattle: Dispute Resolution among Neighbors in Shasta County , 38 STAN. L. R EV . 623687 (1986). Asdetailed below, there are also relations of logical entailment between, e.g., duties and theircorrelative rights; we may regard these, however, as instances of the same stick viewed fromdifferent perspectives.

13. See GILBERT R YLE, THE CONCEPT OF MIND (1949), at 16ff.14. A fee simple interest with possession in land is widely considered to be the closest

analogue to absolute ownership; nevertheless, even such an interest is subject to potentialstatutory restrictions on bequest or to compulsory purchase by the state. See, e.g., both themajority opinion by Justice Brennan and the dissent by Justice Rehnquist in Penn Cent.Transp. Co. v. New York, 438 U.S. 104 (1978).



Back to Bundles 5

various parties] really counts as an owner. His only concern is with thedetailed contents of the various different bundles of legal relations. 15

Any unitary notion of ownership, then, denoting a single canonicalrelation between person and property, seems to drop out of the legal pic-ture altogether, surviving as a mere folk-legal concept in the discourse of laymen. Among the cognoscenti, even Ackermans dimmest law studentrecognizes that the term is at best an imprecise placeholder for what really matterswhich individual bears the particular stick or sticks relevant to thelegal dispute in question.

That, at any rate, is the essence of the bundle theory of property rights. Property and ownership are therefore amorphous or shapelessconcepts, 16 failing to pick out consistently any determinate legal relationor set of legal relations. Rather, their extension changes from one occasionto the next; while everyday talk ascribes an essence or core content tothe concepts of property and ownership, no such entity in fact exists. Forthe bundle theorist, then, the terms represent a sort of primitive hangoverfrom outdated theories that legal science has now dispelled, in much the way that character traits like courage and romance are still widely creditedto the heart.

The theory has a distinguished philosophical lineage. It might not havebeen approved by Hume himself but is certainly in the spirit of his injunc-tion to commit to the ames as sophistry and illusion anything that isneither observable (like the sticks) nor knowable a priori. 17 Indeed, my claim is that it represents a special case of what David Lewis terms Humeansupervenience, or the doctrine that all there is to the world is a vast mosaicof local matters of particular fact, just one little thing and then another. 18That is to say, anything true we can say about the world ultimately reduces tostatements about the distribution of perfectly natural properties and rela-tions; there are no necessary connections between the individual perfectly natural properties and relations; and those perfectly natural properties andrelations are the intrinsic properties of subatomic particles and the spa-tiotemporal relations between them.

Supervenience more generally is a philosophers term of art for a rela-tion between sets of properties; A -properties supervene on B -properties if and only if all differences in A -properties must be accompanied by differ-ences in B -properties. Thus, for instance, the brittleness of objects is saidto supervene on their physical microstructure; if one windowpane is moreshatter-resistant than another, then the atoms composing the two must be

15. JEREMY W ALDRON, THE R IGHT TO PRIVATE PROPERTY (1988), at 29. Cf. the discussion of the train-sparks in R ICHARD A. POSNER , ECONOMIC A NALYSIS OF L AW (8th ed. 2011), at 63.

16. Cf ., e.g., Simon Kirchin, The Shapelessness Hypothesis , 10 PHILOSOPHERS IMPRINT 1-28(2010).

17. D AVID HUME, A N ENQUIRY CONCERNING HUMAN UNDERSTANDING pt. 3, sec. 12 (L.A. Selby-Bigge & P.H. Nidditch eds., Oxford University Press, 1978) (1748).

18. D AVID LEWIS, 2 PHILOSOPHICAL P APERS (1987), at ix; cf. Helen Beebee, The Non-Governing Conception of Laws of Nature , 61 PHIL . & PHENOMENOLOGICAL R ES. 571594 (2000).




differently arranged. Conversely, a different arrangement of atoms does not entail a different level of brittleness, and my coffee mug may be just as proneto shattering as a windshield while clearly being differently constituted. 19

For the subatomic particles discussed by Lewis, we may substitute individual sticks in the bundle of property rights. The bundle theory holds,then, that anything we wish to say about what are commonly termed prop-erty and ownership can be said by reference exclusively to the propertiesof the sticks in question, that none of those sticks or their properties arenecessarily connected with or related to any of the others, 20 and that noth-ing above and beyond these sticks and their properties and relations is of relevance to truethat is, legally validjudicial decision-making. State-ments about property and ownership in every case are therefore eitherredundantbeing reducible in full to their constituent claim-rightsorsimply false, invoking mysterious entities of which we have neither evidencenor a coherent understanding. 21

In metaphysical terms, the bundle theory is therefore, like any Humeantheory, a naturalist and an antiessentialist, position. It is naturalist insofar asit denies the existence of any further relation of property or ownershipover and above the specic claim-rights; once these have been inventoried,there is nothing more the law needs to know. It is likewise antiessentialist,because it denies that there is any particular core claim-right or set of suchrights that determines how the concepts of property and ownership areto be applied. Depending on the variant, it may deny outright that any legalrelation or group of such entities exists that corresponds to those concepts(an antirealist bundle theory), or it may hold that numerous such relationsor collections of them correspond to the concepts, which may be dened

19. See, e.g., J AEGWON K IM, Concepts of Supervenience , in SUPERVENIENCE AND MIND: SELECTEDPHILOSOPHICAL ESSAYS (1993). An accessible introduction to the topic is Brian McLaughlin &Karen Bennett, Supervenience , THE STANFORD ENCYCLOPEDIA OF PHILOSOPHY (Edward N. Zaltaed., 2011), available at http://plato.stanford.edu/archives/win2011/entries/supervenience/.

20. This is not to say, pace Henry Smith, that thebundle theory must attribute the fact (t)hat sticks come in standardized clumpsfee simple, defeasible fee, life estate, future interests,easements, and so on to mere happenstance; H.E. Smith, Property Is Not Just a Bundle of Rights 8 ECON J. W ATCH 279291 (2011), at 284. That the relations are not necessary or intrinsicdoes not mean they must be merely arbitrary. Thus it is entirely inaccurate to charge bundletheorists with analyzing a hypothetical diamond by counting atoms and ignoring important causal features of its physical structure ( id. at 279); what is denied is not that those causalfeatures exist but that they exist necessarily . Cf. Henry E. Smith, Property as the Law of Things , 125H ARV . L. R EV . 16911726 (2012), at 1709ff.

21. Perhaps the best-known expression of this view is the criticism of Local 1330, UnitedSteel Workers v. U.S. Steel Corp., 631 F.2d 1264 (1980), in Joseph Singer, The Reliance Interest in Property , 40 STAN. L. R EV . 611751 (1988). The justices, Singer argues, wrongly dened theissue as a search for the owner of the property. They then assumed that, in the absence of

specic doctrinal exceptions to the contrary, owners are allowed to do whatever they want withtheir property ( id. at 621). To search for the owner when many parties have compellinginterests, he argues, is fundamentally wrong. It is simply not the right question. To assume that we can know who property owners are, and to assume that once we have identied them theirrights follow as a matter of course, is to assume what needs to be decided ( id. at 637638).Instead, the courts ought to decide who wins the dispute on grounds of policy and morality,and then . . . call that person the owner ( id. at 638).



Back to Bundles 7

in any number of ways according to judicial convenience and none of which possesses any special theoretical signicance (a promiscuous realistbundle theory). 22 But for practical purposes, the difference between these variants is merely notational.

The greatest impetus to the bundle theorys popularity came from the work of Wesley Hohfeld, who rst demonstrated that the apparent unity of the property and ownership relations in fact concealed a myriad of distinct legal relations, in particular claim-rights, privileges, powers, andimmunities. 23 Understood thus, moreover, it becomes clear that the prop-erty and ownership relations could be understood not as holding between aperson and a thing but ratherand less mysteriouslyas holding betweena great number of persons regarding a thing. 24 Accordingly, Hohfeld pro-poses to replace the traditional distinction between rights in rem and inpersonam with a new terminology:

A paucital right, or claim (right in personam), is either a unique right residingin a person (or group of persons) and availing against a single person (orsingle group of persons); or else it is one of a few fundamentally similar, yet separate, rights availing respectively against a few denite persons. A multitalright, or claim (right in rem), is always one of a large class of fundamentally similar yet separate rights, actual and potential, residing in a single person (orsingle group of persons) but availing respectively against persons constituting

a very large and indenite class of people. 25

Thus, following and quoting with approval the position of Chief JusticeHolmes, 26 Hohfeld declares that the apparent distinction in kind could be

22. See, e.g., JOHN DUPR E, THE DISORDER OF THINGS : METAPHYSICAL FOUNDATIONS OF THEDISUNITY OF SCIENCE (1993).

23. HOHFELD , supra note 6, at 9697. The most important precursors to Hohfelds theory are H ENRY TERRY , SOME LEADING PRINCIPLES OF A NGLO-A MERICAN L AW (1884); JOHN S ALMOND, JURISPRUDENCE , OR , THE THEORY OF THE L AW (1902); and J OHN CHIPMAN GRAY , THE N ATURE AND SOURCES OF THE L AW (1909). Joseph Singer, The Legal Rights Debate in Analytical Jurispru- dence 6 W IS. L. R EV . 9751069 (1982), provides a useful summary of its historical context anddevelopment.

24. For a clear recent expression of the contrary view, holding that Hohfeld and his suc-cessors have got things exactly backward in this regard, see Henry E. Smith, Property as the Law of Things , 125 H ARV . L. R EV . 16911726 (2012), at 1692. Smiths view is that the informationcosts of keeping track of such a myriad of interpersonal relationships renders the Hohfeldian view radically impractical (see infra note 50) and that property therefore provides a platformfor the rest of private law ( id . at 1691) by simplifying these relationships as a much smaller andstandardized number between persons and things. But the bundle theory, as Smith acknowl-edges ( id . at 1695ff.), is an analytical device that aims to reveal the fundamental nature of property relations and quotidian property talk; that it is usually far more efcient to abbreviatethe multitude of fundamental relations into the everyday vocabulary is not something that bundle theorists commonly deny, any more than physicists will typically eschew talk of theordinary physical objects and properties that they nevertheless hold to be analyzable without remainder into arrangements of subatomic particles and their properties.

25. HOHFELD , supra note 6, at 72.26. All proceedings, like all rights, are really against persons. Whether they are proceed-

ings or rights in rem depends on the number of persons affected. Tyler v. Ct. of Registration,




reduced to a mere difference in quantity; a supposed right in rem is simply a great many rights in personam bundled together:

Suppose that A is the owner of Blackacre and X is the owner of Whiteacre. Let it be assumed, further, that, in consideration of $100 actually paid by A to B,the latter agrees with A never to enter on Xs land, Whiteacre. It is clear that As right against B concerning Whiteacre is a right in personam, or paucitalright; for A has no similar and separate rights concerning Whiteacre availingrespectively against other persons in general. On the other hand, As right against B concerning Blackacre is obviously a right in rem, or multital right;for it is but one of a very large number of fundamentally similar (thoughseparate) rights which A has respectively against B, C, D, E, F, and a great many other persons. It must now be evident, also, that As Blackacre right

against B is, intrinsically considered, of the same general character as As Whiteacre right against B. The Blackacre right differs, so to say, only extrinsi-cally, that is, in having many fundamentally similar, though distinct, rights asits companions. So, in general, we might say that a right in personam is onehaving few, if any, companions; whereas a right in rem always has many suchcompanions. 27

Note that the apparent distinction between contractual and property rightshas also vanished; Hohfeld nds such merely formal, extrinsic distinctions 28

spurious where the underlying substance of the right inquestion is invariant. A nal signicant feature worth noting in Hohfelds analysis is the so-called correlativity axiom, 29 which joins the concepts of claim-rights, priv-ileges, powers, and immunities with their respective correlates of duties,no-rights, liabilities, and disabilities. The axiom states that there is a sym-metrical relation of strict implication between each pair; to take the most relevant and well-known example of rights and duties, my right of exclusivepossession of an apartment strictly implies everybody elses corresponding

175 Mass. 71 (1900), at 76. Cf . O.W. Holmes, Privilege, Malice, and Intent , 8 H ARV . L. R EV . 114(1894).

27. HOHFELD , supra note 6, at 7677.28. Cf. Mossoffs reading of State v. Shack, 277 A.2d N.J. 369 (1971); Shack is a prime

example of how property rights disintegrate under the bundle conception of property . . . .Under such an approach, there is nothing really left to property that distinguishes it fromany other in personam legal entitlements that the government distributes and regulates. AdamMossoff, The False Promise of the Right to Exclude , 8 ECON J. W ATCH 255264 (2011). For anotherrelevant example, consider the lease/license dichotomy; see, e.g ., Wylie,supra note 11, at 1095ff.

29. The term is coined in Mathew H. Kramer, Rights without Trimmings , in A DEBATE OVER R IGHTS : PHILOSOPHICAL ENQUIRIES 7112 (M.H. Kramer, N.E. Simmonds & Steiner Hillel eds.,2000), at 24:

For Hohfeld, rights and duties . . . were always correlative by denition . . . . He posited thecorrelativity of rights and duties as a denitional fundament of his theory, by explicatingthe concepts of right and duty in such a way that each entails the other; each is theother from a different perspective.


10/25



10 S HANE N ICHOLAS G LACKIN

held against my neighbor are after all separable) but to each potentially detachable part of the things over which the rights are held (for you may partition Blackacre and sell me a right of possession in the eastern but not the western half). 34 Since the sticks are so innitesimal, it is little wonderthat theorists have written of the disintegration of property. 35

III. PENNERS UNIFIED CONCEPT

So much for the bundle theory as it has descended to us through Hohfeldand Honor e and as I propose to defend it against the recent attacks. Why has its popularity declined so sharply in recent decades, to the point that recent theorists of property, in the course of outlining their own views, canregard it as abundantly refuted and in no need of further discussion? 36Even during its ascendancy, the contrary idea of an integrated notion of property existing prior to and informing the law 37 remained attractive fora variety of reasons, including its congruence with laypersons intuitionsand the purported ability to explain the clear meaning and use of the termin theory and practice 38 as well as to ground wider arguments about the justice of particular distributions of property. In particular, such a notionis thought to provide a bulwark against statist conscatory practices. 39 Asstated above, I do not propose to discuss the issues of distributive justicein the course of this paper. Rather, I focus on the ways that the conceptsof ownership and property are meant and used, according to some of themore inuential critiques of the bundle theory, with the aim of exposinga number of misunderstandings or misrepresentations of the theory. What these alternative theories share, which the bundle theory denies, is the focuson some supposed standard relationship that an owner might have to a

34. Many cities, as I am grateful to an anonymous commentator for pointing out, haveregulations that prohibit splitting city lots in two in this manner. There are, of course, excellent pragmatic reasons for regulations of this sort, which minimize the information costs imposedby excessively baroque, exotic, or ne-grained distributions of property rights; some theorists(see infra note 49) make these costs the basis of their opposition to the bundle theory. But according to the bundle theory, these are contingent rather than essential connections betweensticks, and there is nothing in the nature of property that determines that regulations must exist to link them in this way.

35. Thomas C. Grey, The Disintegration of Property , in NOMOS XXII: PROPERTY (J. RolandPennock & John W. Chapman eds., 1980), at 69. For a very clear exposition and rebuttal of Greys somewhat extreme view by a fellow bundle theorist, see STEPHEN R. MUNZER , A THEORY OF PROPERTY (1990), at 3136.

36. See, e.g., Breakey, supra note 2.37. Id. at 241.38. Id .39. Klein & Robinson, supra note 2, at 196201, provides a range of quotations to this effect;

but cf. R ICHARD A. EPSTEIN, T AKINGS: PRIVATE PROPERTY AND THE POWER OF EMINENT DOMAIN(1985), which argues that the bundle theory itselfby making each stick a compensabletakingserves this protective function. This argument is reiterated in Richard A. Epstein,Bundle-of-Rights Theory as a Bulwark against Statist Conceptions of Private Property , 8 ECON J. W ATCH223235 (2011).



Back to Bundles 11

particular physical resource that they take to be paradigmatic of property and ownership in general.

The rst such view I want to consider is that of James Penner. In a lengthy (110 pages) paper 40 and a subsequent monograph, 41 he outlines both a cri-tique of the bundle theory and an alternative, unied concept of property that rest on two main claims; the exclusion thesis, and the separability thesis. I focus on the critical rather than the positive aspect of his theory;if his criticisms of the bundle theory prove unsuccessful, then the positiveaccount loses both its motivation and many of its basic assumptions.

Penners critique begins inauspiciously. Hohfeld, he writes, was madfor symmetry between rights and duties, and he based his notion of the cor-relativity of rights to duties on it. 42 But as we see above, with what Kramerterms the correlativity axiom, Hohfeld identies a logical equivalence be-tween Xs holding a right against Y and Ys having some correlative duty toward Y. That equivalence is not merely the expression of some irrationalaesthetic whim of Hohfelds; it is, prima facie, a function of the very mean-ing of the terms. 43 So when Penner concedes that anybody believing in thecorrelativity of rights and duties is pretty much bound to describe a right in rem as a multitude of rights in personam ,44 but reassures us that (n)o one without a commitment to Hohfelds views regarding symmetry need denecorrelativity in this way,45 he owes us rather more than the expression of personal aesthetic preference for asymmetry and rejection of the in rem/inpersonam reduction as implausible that we are given. What we really needand are nowhere given is a wholly new account of the normal meaning of right and duty that establishes that they are independent, and unrelated,concepts.

Let us charitably suppose that some such denition of those terms isavailable and acceptable. Why does Penner consider the equivalence of inrem and in personam rights to be implausible? The difference betweenthe two which Hohfelds analysis omits, he argues, is that for a right inpersonamunlike a right in remit matters which particular person is theduty bearer or the right holder.

Consider the familiar Blackacre. If A owns Blackacre, writes Penner,then hemay grant any numberof rights in personam to specic or speciablepeople to make use of it, walk across it, and so on. But it matters to A whothey are, and it matters to them who A is. 46 A must know who they are in

40. James Penner, The Bundle of Rights Picture of Property ,43UCLAL.R EV . 711820 (1996).41. J AMES PENNER , THE IDEA OF PROPERTY IN L AW (1997).42. Id . at 25.43. Cf. Walter Wheeler Cook, Introduction , in W ESLEY N. HOHFELD , FUNDAMENTAL LEGAL

CONCEPTIONS AS A PPLIED IN JUDICIAL R EASONING : A ND OTHER ESSAYS (1923): Any given singlerelation necessarily involves two persons. Correlatives in Hohfelds scheme merely describe thesituation viewed rst from the point of view of one person and then from that of the other.PENNER , IDEA , supra note 41, at 2526.

44. Id .45. Id. at 26.46. Id.




order to grant them the right, and they must know who A is and the scopeof his rights in order to assess the content of their license. But in the matterof rights in rem, as when someone wishes to purchase Blackacre, no suchpersonal familiarity with A is required:

They may never see him, or hear about him, or even know if he has diedand been replaced as owner by his younger sister. Their only relationship tohim is through his property, in the sense that they can affect A only by actingon his property in some way. To them, A is only represented as his property,and whats more, he is not even represented as A. He is only represented asowner, i.e. his particular identity is completely obscure. This is no relationin personam between them and A. It is exactly the same relationship that everyone has to all the property that is not their own. 47

This is not an especially novel point. Indeed, it will be familiar to logiciansand philosophers of language as a case of the de dicto / de re distinction. 48Consider the phrase the tallest boy in the class; this picks out one deter-minate individual. But in semantic contexts such as belief and desire ascrip-tion or modal claims about possibility and necessity, termed referentially opaque,, such a phrase may be ambiguous. In the sentence Mary wants tokiss the tallest boy in the class, the phrase may refer to that specic person,

or it may function as a general description that any number of others might have fullled in other circumstances. Suppose that Mary believes Henry tobe the tallest boy; the sentence does not specify whether she wishes to kissHenry in particular or whichever boy happens to be tallest (perhaps she ismistaken, and Henry is in fact marginally shorter than Ben, who slouches).If it is Henry alone whom she desires, the sentence should be interpretedde re , or as pertaining to the thing identied by the phrase; if instead shedesires the tallest boy regardless of who he may be, the sentence should beread de dicto , or as pertaining to the words of the phrase itself.

Something of this sort, I take it, is at the root of Penners concern. If I amgranted a license in personam to walk Blackacre by its owner, I must havethe ability to identify him, and the phrase the owner of Blackacre shouldconsequently be read de re . But if I am not granted such a license, I have,like everybody else, a duty in rem toward the owner of Blackacre,, whoever that may be (i.e., read de dicto ). Penner explains:

Owing individuated, separate duties to particular property-owners would pre-sumably require knowing what owners held what property in order to under-

stand what those duties are. But we dont. And thats because our duty is not

47. Id., at 27.48. Introduced in W.V. Quine, Quantiers and Propositional Attitudes , 53 J. PHIL . 177187

(1956).



Back to Bundles 13

to trespass on the property of others. We are under one duty to the plurality of property holders however their property is distributed among themselves. 49

The question here, which Penner appears to beg, is why I must have per-sonal knowledge of somebodys identity in order to hold a right of or beara duty toward them in personam. We might suppose that I must have suchknowledge if I am claiming a special or exceptional right; if I am assert-ing, for instance, that I, unlike everybody else , have been granted a licenseto walk Blackacre. But that would be the case, if true, only because theexceptional nature of my right, as against the default rights and dutiesof everybody else, would require it to be specically granted. 50 This in no way entails that the rights and duties we assume in default of being specif-

ically exemptedsuch as the duty to avoid trespassing on the property of othersare any less in personam than the exceptional ones we are grantedspecically. When Penner presumes that we must know which owners(de re ) hold what property in order to understand our in personam duties,he is helping himself to just what is at issue: the claim that an undifferenti-atedduty in rem cannot be reduced to a multitude of individual inpersonamduties.

Suppose that Mary, some years later, is out walking in the countryside.She happens upon Blackacre and briey considers taking a shortcut across

it. I had better go the long way around, she tells herself eventually, sinceI have a duty to Blackacres owner not to trespass on the land. Supposefurther that Mary believes her old schoolgirl crush Henry to be the owner;nothing about the legal nature of the duty hangs on whether or not herbelief is correct. Whoever the owner is, Mary bears an (otherwise) identicalduty to that person not to trespass. The distinction between de re and de dicto meanings can be problematic, but only in contexts of referential opacity, when the intended reference of the sentence may fail. Since the law does

49. PENNER , IDEA , supra note 41, at 27. Thomas Merrill & Henry Smith, What Happened to Property in Law and Economics? , 111 Y ALE L.J. 357398 (2001), use this as the basis for theirinuential information-cost account of property-as-exclusion; see also Larissa Katz, The Regu- lative Function of Property Rights , 8 ECON J. W ATCH 236246 (2011), at 239; Henry Smith, On the Economy of Concepts in Property , 160 U. P A . L. R EV . 20972128 (2012); and Henry Smith, Emergent Property , in PHILOSOPHICAL FOUNDATIONS OF PROPERTY L AW (J. Penner & H.E. Smith eds., 2013),especially sec. 5. My own view, which coincides with Munzer, Property and Disagreement , supra note 4, is that context and heuristics, as in any area of human life, reduce the supposedcognitive burden of in personam property rights considerably.

50. Even then, the right might be granted to me de dicto and without personal acquaintance

as, for instance, the individual renting the Gate Lodge of Blackacre or the oldest single manin the village; de re reference and personal knowledge are not necessary for in personam rights.For this reason, we can dismiss Penners later concern that in personam rights have theirspecic right-holders and duty-owers essentially and that the title embodied in those rightscannot therefore be transferred to others; James Penner, Potentiality, Actuality, and Stick- Theory , 8 ECON J. W ATCH 274278 (2011), at 277. I thank an anonymous reviewer for bringingthis point to my attention.




not trafc in such statements, 51 the ambiguity identied by Penner cannot be legally signicant.

Against the view that property is a structural composite,, Penner goes

on to propose an alternative view of property as a single right protectinga single, identiable interest. 52 The argument for this conclusion is es-sentially a burden of proof argument, drawing on the intuition of Occamsrazor. 53 This invocation is surprising, to say the least; William of Ockhamsinjunction that entia non sunt multiplicanda praeter necessitatem 54 would serveperfectly as an epigram for the bundle theory and its insistence that thephenomenon of property can be fully explained on the basis of the individual sticks without any appeal to a wider, all-encompassing relation of property ownership.

Penner, although he elsewhere explicitly divides his alternative account into an exclusion thesis and a separability thesis, 55 states his view of property most succinctly thus:

The right to property is the right to determine the use or disposition of analienable thing in so far as that can be achieved or aided by others excludingthemselves from it, and includes the right to abandon it, to share it, to licenseit to others (either exclusively or not), and to give it to others in its entirety. 56

This certainly captures much of the intuitive force of the folk-legal concept of property However, the attempt to shoehorn various aspects of property law intoand various other aspects of law out ofthis concept of a single,coherent right 57 can only be described as tortuous. 58

51. Actually, this is not quite correct. I suspect, for instance, that a case of de re / de dicto opacity in the context of a demonstrative referenceas discussed in David Kaplan, On the Logic of Demonstratives , 8 J. PHIL. LOGIC 8198 (1978)may be the best theoretical explanation of the celebrated distinction between the common-law contract cases Cundy v. Lindsay, (1878) 3 App. Cas. 459, and Lewis v. Averay [1971] 3 All ER 907; [1972] 1 Q.B. 198. But that is another

papers work; the claim is sufciently accurate for our current purposes.52. Penner, Bundle , supra note 40, at 739.53. Id.54. Entities are not to be multiplied beyond necessity.55. PENNER , IDEA , supra note 41, chs. 45.56. Penner, Bundle , supra note 40, at 742.57. Id. at 754.58. Here is the summary of just a single instance:

My submission about the proprietary character of choses in action, then, is that to the extent that we regard choses in action, these rights in personam , as property rights, we doso because of their relative personality poverty in relation to other rights in personam . What makes these problematic property rights is the fact that while the relationship ishumming along and parties are meeting their obligation, when banks are honouringtheir depositors; balances, dividends are paid, and debt payments are made on schedule,these rights full very much the role of property that money does. When things go awry,however, when holders are apt to lose shareholder suits or actions against their debtors,the rights revert, in a sense, to their in personam origins.

PENNER , IDEA , supra note 41, at 131.



Back to Bundles 15

The problem is not that Penner is unsuccessful in the two hundred orso pages that the endeavor takes up. The problem, reminiscent of ThomasKuhns famous account of the Copernican revolution, 59 is rather that suchan extensive effort is necessary at all. As Kuhn describes, the Ptolemaic as-tronomers who predated Copernicus believed the sun and planets to orbit the Earth in cycles of perfectly circular form. As more and more accurateobservations began to be made, however, it became necessary to postulatefurther circular oscillations around the cycles themselvesepicyclestopreserve the theory. By the time of Johannes Kepler, who rst questionedthe utility of the epicycles, the ad hoc patches necessary to support thegeocentric/circular orbits theory had become absurdly complex. What dis-tinguished the Copernican theory was not that he could make better andmore accurate predictions, for he could not; rather, it was the fact that histheory did not need to be supplemented by such a massively complicatedauxiliary structure to bring it into line with observation. A sufciently de-termined theorist, both W.V. Quine and Pierre Duhem demonstrate, canreconcile any recalcitrant data whatsoever with his theory; 60 the questionfor Penner is not whether his efforts are successful but whether they are worth the trouble. 61

There remain, in any case, theoretical problems with Penners analysis.Penner draws a contrast between the bundle theory and his own on thebasis of the individuation of particular rights; where the bundle theorist must believe such rights to be discrete and determinate, like (m)embersof a club [which] naturally come in units called persons,, Penner holdsthat they are in fact like pieces of a cake which can be sliced in any way we wish; they are no more than momentary functional descriptions made with a particular legal concern in mind. 62

That this is so, Penner believes, can be seen by reecting on what happens when someone grants another a license to use their property. On the bundle view, A, the owner, holds in his bundle of rights the millions of rights of Bto do each and every thing with As property, and the millions of rights of C, and D, ad innitum. On the grant of a license to B, A merely extracts theparticular right from his bundle with Bs name on it and transfers it to B. If it is a non-exclusive license, then A can do the same in turn for C or D. 63The prospect, he writes, boggles the mind. 64

Yet this consequencethat the bundle theory entails the existence of a mind-boggling number of discrete rightsshould be neither surpris-ing nor troubling. In outlining the bundle theory above, I compare the

59. THOMAS S. K UHN , THE STRUCTURE OF SCIENTIFIC R EVOLUTIONS (1962), at 140, 322324.60. W.V. Quine, Two Dogmas of Empiricism , 60 PHIL. R EV . 2043 (1951); P IERRE DUHEM , THE

A IM AND STRUCTURE OF PHYSICAL THEORY (1954), at 180ff.61. See also the criticisms of Penners position in Adam Mossoff, What Is Property? Putting the

Pieces Back Together , 45 A RIZ. L. R EV . 371443 (2003), at 376378.62. Penner, Bundle , supra note 40, at 754755.63. Id. at 758.64. Id.




rights not to sticks in a bundle but to atoms of property; the Humeansupervenience thesis to which I compare the bundle theory is specically one about how our everyday talk about the world is ultimately dependent on and reducible to talk about the subatomic particles of modern physics.The number of such particles making up the world may indeed be mind-boggling; but no physicist supposes that the physical world therefore hasthe structure of cake and can be sliced in any way we wish. No matterhow many such atoms there are, they are nevertheless individually discrete;Penner would hardly suppose that the members of Barcelona Football Clubfail to be individuated simply because there are 170,000 of them.

Penner restates his argument as follows:

we can actually conceive of property in terms of a right which permits anowner to do anything or nothing with his property; the disaggregative bundleof rights thesis insists that an owner may do everything with his property. Theformer view accords with the fact that the law of property takes no interest in the particular use one makes of ones property (which is not to say that criminal law or the law of taxation does not); the latter holds that the essenceof property is an innite number of rights to use a thing, in the same way that the Hohfeldian idea of a right in rem entails having millions of rights against all other people. 65

But phrased this way, the incoherence of his position becomes clear; thereis simply no formal difference at all between anything and everything inthis context. In any imaginable calculus of rights, the two notions will beexpressed by exactly the same universal quantier: For all rights x , A may exercise x or grant it to B . Once again, Penner insists on an intuitivedistinction premised on the existence of some special property or entity that exists over and above the elements into which it can be formally analyzed withoutin the mathematical senseremainder.

The ultimate reason for clinging to this distinction Penner calls the use-

lessness thesis; the bundle theory involves at least the tacit admission, that the concept of property is vague or undenable, and so . . . degenerate oruseless.66 If the theory is true, he reasons, the deationaryconcept of prop-erty that it advances can add nothing to our understanding of property law or judicial decision-making. For Thomas Grey, 67 this means that the entireconcept, though retaining some use at the folk-legal level, must ultimately be confused. On a more moderate view, such as Barry Hoffmasters, theconcept is conclusory : A statement of ownership is a conclusion drawn fromcomparing a particular combination of the incidents of ownership, existing

together in a determinate situation, with the paradigm of ownership.68

In65. Id. at 758.66. Id. at 769.67. Grey, supra note 35.68. Barry Hoffmaster, Between the Sacred and the Profane: Bodies, Property, and Patents in the

Moore Case , 7 INTELL. PROP. J. 115148 (1992), at 129. Cf . MUNZER , THEORY , supra note 35, at 39;



Back to Bundles 17

other words, judges will make a decision in any given case on the basis of the particular rights and duties of the parties and the particular demandsof justice in the circumstances and will retrospectively declare a property right to have or not to have obtained. 69

To show the problem with the uselessness thesis,, Penner examines thefamous cases of International News Service v. Associated Press 70 and Moore v.Regents of the University of California .71 In International News , Justice Pitney for the majority determined that the plaintiffs, by republishing hotbut noncopyrightablenews gathered by the defendants, had infringed thedefendants quasi property. 72 Like the dissents by Justices Holmes andBrandeis, this seemed perfectly to illustrate the conclusory view.

Instead of talking about quasi property or exchangeable values, won-ders Penner, are we not on a better footing if we can say that what the Associated Press was claiming was the right to a market monopoly, akin tothe protection generally provided by the monopolies of copyright or patent law?73 Such a property right to a legally structured market position,, heargues, makes more sense than (j)amming rights to the news or to anidea into the mold of property rights. 74 Yet this characterization of theconclusory view is plainly inaccurate and still wedded to the idea of theproperty concept as determinate. There is no need to jam such rightsinto a mold; the property concept, according to the bundle theory, isshapeless and more akin to a blanket we may simply throw loosely over whatever considerations we wish it to cover. It is hard to argue, moreover,that Penners unied concept of property makes judicial reasoning easierto predict if the naive but accurate 75 intuition he appeals to nds all three Supreme Court opinions in a landmark case confused.

In Moore , where the plaintiff sought to have his property rights recog-nized in cells excised from his body during surgery for leukemia, which were later the subject of several patent lines, Penners account is similarly at variance with actual judicial reasoning. Justice Panelli, writing for themajority, denied the claim in conversion on the grounds that a relevant statute eliminates so many of the rights ordinarily attached to property that one cannot simply assume that what is left amounts to property orownership. 76 Yet Penner rejects both this and the other opinions by Justice

Larissa Katz, Exclusion and Exclusivity in Property Law , 58 U. TORONTO L.J. 275315 (2008), at 276.

69. Cf. W.H. Hamilton & I. Till, Property , 12 ENCYCLOPEDIA SOC. SCI. 536 (1933): It isincorrect to say that the judiciary protected property; rather they called that property to whichthey accorded protection. See, e.g., the famous Australian case of Victoria Park Racing and Recreation Grounds Co. Ltd. v. Taylor (1937) 58 CLR 479 (Austl.); as well as Sports and Gen. Press

Agency Ltd. v. Our Dogs Publg Co. Ltd., [1916] 2 K.B. 880, [1917] 2 K.B. 125.70. Intl News Serv. v. Associated Press, 248 U.S. 215 (1918).71. Moore, supra note 9, at 479.72. Intl News , supra note 70, at 235236.73. Penner, Bundle , supra note 40, at 816.74. Id.75. Id. at 817.76. Moore , supra note 71, at 492.




Moskwho explicitly invokes the bundle theory 77and Justices Broussardand Arabian for failing to consider the question of whether: (e)ven if onecan regard the control rights we have over our body parts as somewhat akin

to ownership, is there nothing more to be said before we treat somethingas intimately related to the human persona as ones body as property? 78The bundle theorists answer, and that adopted by each of the justices

in Moore , is simply no; the facts of the case and the particular bundle of sticks involved provide a complete basis on which to reach the decision,and no concept merely supervening upon them need be invoked to explainthe outcome. Penner thinks this entails (t)he idea that some magic oc-curred on [Moores spleens] removal so that the researchers could claima property right in it as if claiming something unowned the instant it left

his body.79

Once more, however, the criticism works only if one is weddedto a substantive concept of property. If a property right is itself nothingmagical but merely the label we attach ex post to any subbundle of the stickssufcient to decide the case, then there is just nothing at all mysteriousabout that right transferring when certain of its constituent sticks do.

I leave the discussion of Penner with one nal point that appears to havegone unremarked in the subsequent literature. To sustain the uniedconcept of property in the admitted absence of necessary and sufcient conditions for its application, Penner overturns the entire standard eld

of classical semantics, relying instead on an alternative view based on Jonathan Suttons unpublished bachelor of philosophy thesis. 80 I offer noopinion here on the merits of Prof. Suttons criterial semantics, save tonote that it has not yet displaced the mainstream view. To completely reviseour understanding of linguistic meaning for the sole purpose of preserving aparticular theoretical view of legal rights would surely be to break a buttery upon the proverbial wheel. That Penner does so in the name of a burdenof proof argument, drawing on the intuition of Occams razor 81 is doubly curious.

IV. HARRISS MINIMAL STRUCTURE

The nal anti bundle argument I am considering is that of Jim Harris. Har-riss main work on the subject 82 is careful and detailed, but it can occasion-ally be difcult to identify a particular structure to the overall argument. Accordingly, I deal with the main points in approximately the order heraises them.

77. See supra note 9.78. Penner, Bundle , supra note 40, at 721.79. Id. at 817.80. Id . at 767ff.; J.K. Sutton, Family Resemblance: An Externalist Approach to Metaphysics

(1992) (unpublished B. Phil. thesis, Oxford University).81. Penner, Bundle , supra note 40, at 739.82. J.W. H ARRIS, PROPERTY AND JUSTICE (1996).



Back to Bundles 19

He begins, oddly enough, by effectively conceding one of the centralclaims of the bundle theory. Although all of us (philosophers, lawyers, andordinary folk) seem to share an intuitive idea of what property is, 83 there isno univocal, singular concept of ownership 84 or property to reckon with.Indeed, on a number of occasions he refers explicitly 85 or implicitly 86 to thenotion of a bundle of rights. Onwhat grounds, then, does Harris reject thebundle theory?Although no single concept can do justice to property rights,he argues, we can identify a minimal core of necessary conditions basedon the twin notions of trespassory rules and the ownership spectrum, 87 which is then extended in a variety of directions by different legal systemsin response to different circumstances and desiderata.

The rst stage of Harriss argument is to attempt to x our intuitionsabout property by the use of a series of anthropological thought experi-ments, termed Forest Land, Status Land, Red Land, Contract Land, WoodLand, and Pink Land, progressively advancing in both the type of resourcesavailable to residents and the local rules limiting their use. 88 Harriss methodis to demonstrate that one concept is logically prior to another by show-ing that an anthropologist from one imaginary society could understandanothers practices only in terms of his own groups. But this seems a clearnon sequitur; we may at best conclude from these stories that concepts likeprivate property are descriptively prior to alternatives given the limits of the hypothetical anthropologists conceptual resources .

Harriss imaginary societies might be unobjectionable if their purpose were simply to establish the outlines of his own intuitions about property; but they are useless in investigating how those intuitions line up with reality. 89 Yet he immediately begins drawing lessons about necessary and sufcient conditions for the existence of property institutions, lessons that simply re-peat the assertions made about his own intuitions in the course of describingthe imaginary societies. Before long, the discussions will have yielded theconclusion that the core idea of a property institution resides in the twinnedconceptions of trespassory rules and the ownership spectrum. 90

If the elements of the minimal structure are of dubious provenance, thesophisticated structures, 91 Harris insists must be built upon it for the pur-poses of the full range of modern property talk are patently gerrymandered.Example after example is force-tted in or out of the schema in the same

83. Id. at 7.84. Id. at 132; cf. J.W. Harris, The Elusiveness of Property , in PERSPECTIVES ON JURISPRUDENCE :

ESSAYS IN HONOUR OF JES B JARUP, 123132 (Peter Walgren ed., 2005), at 129.85. H ARRIS, PROPERTY , supra note 82, at 47, 51.86. Id. at 73.87. Id. at 5.88. Id. at 15ff.89. For a stark contrast with the potential utility of actual empirical anthropological studies,

see, e.g., E. Adamson Hoebel, Fundamental Legal Concepts as Applied in the Study of Primitive Law ,51 Y ALE L.J. 951966 (1942).

90. H ARRIS, PROPERTY , supra note 82, at 55.91. Id. at 42 & ff.




epicyclical manner we saw in Penner, in openly vague and arbitrary fashion;if one poses the question, who is the owner of this house or at?, answers will refer only to those with leases of a substantial duration, although usagepoints to no particular cut-off point. 92 The result, he freely admits, is aportmanteau category, 93 delineated only by a stipulative boundary. 94

In defense of this position, Harris argues that in the absence of such a uni-ed, albeit heterogeneous, concept, we would have to regard as baleful 95the everyday claims of entitlement to particular items of social wealth. Thismay be the aim of some more radical critics of property, but there is no obvious reason we cannot, per Thomas Grey, 96 leave the folk-concept intact foreveryday use and rest the actual legal claims of entitlement on the individualsticks themselves.

Harris defends his jury-rigged property concept against some of thoseradical critics by arguing that the property skepticism of Alf Ross and KarlOlivecrona:

is based on a theoretical assumption about the ontological status of concep-tual entities: either they directly reect brute reality, or they are metaphysi-cal chimera and their employment must be explained away in psychologicalterms. It ignores the possibility that the human mind may create abstract entities which human institutions can then usefully employ for a variety of functions, without it having to be supposed that the entities belong to some

mysterious supra-sensible realm. 97

This is perfectly true as far as it goes; articial, even socially constructedkinds (such as money or lawyers ) are no less real, objective features of the world than natural kinds (such as water or dogs ). But there is a further,perfectly obvious distinction between these articial-but-real kinds and thesort described by Nelson Goodman as grue-some, exemplied by theheterogeneous property of being grue: green if examined before time t ,and blue if examined thereafter. 98 Some articial kinds, that is, do reect the real contours of our world, but this will not license just any portmanteaucategory; kosher is a potentially useful scientic classication, whateverones faith, 99 but kosher, or yellow, or smaller than a bread-bin is not.Pointing out that concepts of the former sort are legitimate does not absolvecharges that Harriss portmanteau concept is of the latter sort. The humanmind may indeed create abstract concepts, but it should not simultaneously

92. Id. at 72.93. Id. at 86.94. Id. at 62.95. Id. at 63.96. See supra note 67.97. H ARRIS, PROPERTY , supra note 82, at 131. See e.g. K ARL OLIVECRONA , Lockes Theory of

Appropriation , 24 PHIL . Q UARTERLY 22034 (1974).98. NELSON GOODMAN , The New Riddle of Induction , in F ACT, FICTION , AND FORECAST (1954).99. Cf. K IM STERELNY & P AUL E. GRIFFITHS , SEX AND DEATH : A N INTRODUCTION TO THE PHI-

LOSOPHY OF BIOLOGY (1999), at 196.



Back to Bundles 21

eschew any true semantic or conceptual essence 100 of those concepts if it wishes them to be regarded as faithful to reality, carving nature at theproverbial joints.

Harriss rival view of property and ownership, then, is unsupported by any persuasive line of argument. However, he makes direct criticisms of twoaspects of the bundle theory that its proponents will be concerned to rebut.The rst of these again concerns the correlativity axiom, while the seconddecries the conclusory view of the role of the property concept in judicialdecision-making as tautologous.

The root of Harriss attack on the correlativity axiom is, once again, anapparent misunderstanding of the in rem/in personam reduction. Thecommon view holds, he argues, that the layman thinks in terms of rela-

tions between persons and objects, while the lawyer knows that the relevant relations are between people and concerning the objects in question. Thecontrast, he insists, is a false one. 101

Indeed it is, but it also seems a straw man. Nothing in Hohfelds orHonores arguments, as Harris himself acknowledges, 102 suggests that wemust cease to speak or think even loosely about legal relations betweenpersons and objects; their aim is to analyze such talk, not to eliminate it.The point is, in other words, that we can better understand the nature of those legal relations if we reduce them to relations between persons. The

bundle theory requires no more; it displays the underlying logical structureof such relations but does nothing thereby to undermine them.The reduction is possible, Harris agrees, if two conditions are met: rst,

that all the relevant legal provisions are known and determinateno opentexture; and, secondly, that we are seeking to convey information about the legal situation at a particular moment in time. 103 Without going intothe details of Harriss discussion of the second condition, let us agree that he identies an important qualication to the bundle theory as previously elaborated. 104 At the end of Section II, I identify the sticks composing a

typical bundle as corresponding to each of Honor es incidents, indexed toeach particular potential right holder, duty bearer, and detachable part of the object in question. Since I may convey an interest in the object to youfor a limited period of time only, we should say further that the incidentsare indexed to the individual temporal parts of the detachable portions of the object. 105

The rst condition, whose terminology is drawn from Hart, 106 requiresthat it may not be an open question whether a given instance of conduct

100. H ARRIS

, PROPERTY

, supra note 82, at 142.101. Id. at 119.102. Id. at 120.103. Id. at 122.104. The issue also seems to be alluded to in P ENNER , IDEA , supra note 41, at 25.105. For the standard denition of temporal parts, see TED SIDER , FOUR -DIMENSIONALISM

(2001), at 60.106. H.L.A. H ART, THE CONCEPT OF L AW (2d ed. 1994), at 124136.




falls under the scope of the right. 107 A right in rem which is not open-textured in this way will straightforwardly reduce to a bundle of rights inpersonam whose scope is similarly known and determinate. What is the

problem with open-textured rights in rem? According to Harris, judicialdecisions regarding such rights, lacking a determinate basis in precedent,must appeal to the values of ownership generally. But in doing so, they must either be brute and tautologous in their conclusions or they must invoke aparticular relation between persons and things.

In short, Harriss objection to the correlativity axiom is also an objectionto the conclusory view of propertys role in judicial reasoning. In Bradford v. Pickles ,108 the plaintiffs sought to restrain the defendant from sinking a well under his land under the law of nuisance, as his sole purpose in doing so

was to prevent water from percolating to their waterworks, thereby forcingthem to purchase the land from him. The House of Lords upheld the lowercourts refusal of an injunction, explicitly invoking Mr. Pickles rights as alandowner to use his property in a self-interested manner, even if doingso was churlish, selsh, and grasping. 109 In other words, argues Harris,the court argued from Pickless general, in rem rights over the land to hisspecic, in personam entitlements against the mayor and council to act as hedid. If the in rem rights were simply equivalent to the in personam rights,as the bundle theorist claims, the judgment expresses an uninteresting

tautology; Pickles is privileged to do this and the Corporation has no right to stop him because Pickles is privileged to do this and the Corporation hasno right to stop him. 110

The obvious query to raise here is: What about this objection is specicto open-textured rights? The judgments in cases decidable on precedent, where Harris has no objection to reducing the in personam rights to in remrights, would be no less tautologous by this reasoning. This is our rst indi-cation that the tautology in question may not be especially problematic.

Let us look in more detail at the nature of a judgment of this sort.

According to Harris, the giving of a verdict changes the question fromopen- to closed-textured: The House having ruled as it did, the law asstated at any subsequent time includes a no-right/privilege relationship insuch circumstances, and that will continue to be true unless the House of Lords overrules that decision or it is abolished by statute. 111 That is to say,

107. For instance, when carbon-trading regimes are rst instituted, a new class of economicassets is created. It then seems to be an open question who owns and is entitled to trade thecredits newly associated with some piece of land. I thank Prof. Yvonne Scannell for suggestingthis example to me.

108. Mayor and Corp. of Bradford v. Pickles, [1895] A.C. 587.109. Id. at 600601.110. H ARRIS, PROPERTY , supra note 82, at 124. Cf . also Brian Leiters discussion of conceptual

rule-skepticism in Brian Leiter, American Legal Realism , in THE BLACKWELL GUIDE TO THE PHI-LOSOPHY OF L AW AND LEGAL THEORY 5066 (Martin Golding & William Edmundson eds., 2005),at 6163.

111. H ARRIS, PROPERTY , supra note 82, at 124.



Back to Bundles 23

the effect of a judgment in an open-textured case is just to recognize another stick .

Whether we consider this new stick to be created de novo by the decision(judge-made) or to have been in force since the dawn of legal time, albeit never previously elaborated (the preexistence thesis) is for current pur-poses irrelevant. 112 The important question is why we must, according toHarris, interpret the decisionas the assertion of a general principle of owner-ship that he admits to be subsequently decomposable into individual rightsrather than as the recognition of an additional right that contributes to ourafter-the-fact ownership conclusions. Why, that is, must judicial reasoningbe top-down, from ownership to sticks, rather than bottom-up? 113 Cer-tainly, given the heterogeneous andjury-rigged natureof the portmanteautop-down ownership concept that Harris outlines, he cannot commend it to us on the basis of legal certainty or theoretical elegance.

The supposed basis, then, which must apply equally in the case of closed-textured questions, is that of tautology: reasoning from a (newly aug-mented) bundle of rights to a conclusion that merely restates the existenceof that bundle would be both repetitious and, Harris claims, uninformative.Repetitious it might well be; but tautologies are uninformative only if wedo not already possess the information they express. Since the term waterrefers to H 2O, the sentence water is H 2O means H2O is H2O; but thediscovery of waters chemical makeup is no uninteresting triviality. 114 Sim-ilarly, even if we regard the decisions in Bradford v. Pickles , or International News Service , or Moore , as formally tautologous, they nevertheless expressinteresting and decidedly consequential discoveries about the extent anddistribution of the proprietary interests assertible under the common law.

V. CONCLUSION

In this article I offer a defense and, I hope, a clarication of the bundletheory. I want to summarize here briey why I think that it is the only successful and the most theoretically attractive account of property rightsavailable to us.

The major innovation of my position, the feature that is most likely tocome as a surprise to anyone already familiar with at least the textbook orlecture-hall accounts of the bundle theory, will be the number and minus-cule scope of the individual rights that compose a typical bundle. The classictheoretical justication of the bundle theory, I argue, applied in a thorough-going manner, does not merely require us to consider separately the dif-

ferent incidents of ownership identied by Honor e, separately indexed

112. Donna Lyons, Dworkin and Judicial Discretion: A Critical Analysis of the Pre-existence Thesis ,11 TRINITY C. L. R EV . 112 (2008) summarizes the main issues with exemplary clarity.

113. Cf. Munzer, Property and Disagreement , supra note 4, n.43 & ff.114. Cf. Hilary Putnam, The Meaning of Meaning , 7 MINN. STUD. PHIL. SCI. 131193 (1975).


25/25


toas Hohfeld points outeach potential right holder and duty bearer.That quantity of rights already strains the bundle of sticks metaphor tothe limits; such a number of sticks would be less a bundle than a lumberyard.

As I contend, we need to go further; each of those sticks is decompos-able not just into all the separate spatial parts in which interests could beconveyed to others but also into all the separate temporal parts indicatingthe limited durations for which an interest can be conveyed. The numbersare now well beyond even lumberyard quantities; they are in a literal senseastronomical. It is just that multitudinous character that offends Penner, yet I believe it actually assists us in understanding the bundle theory by providing the new metaphor that several theorists recently call for; 115 theindividual rights may be best thought of not as sticks but as atoms To that end I draw on an inuential theory from the metaphysics of scienceDavidLewiss Humean supervenienceto illustrate the relation the individualrights have both to each other and to the judicial decisions and everyday property talk for which they provide the whole and unique basis.

When we get this metaphor clear, just like the older bundle metaphor, we see two reasons to prefer the deationary theory it expresses to theaccounts of property that rely on the existence of some entity over and abovethe individual sticks or atoms. The rst is that since the individual rightscan give a complete account of property talk, invoking anything further issuperuous and inelegant; entia non sunt multiplicanda praeter necessitatem .The second is that in fact no such further entity has been observed in the wild, nor convincingly postulated; and it is far from clear even what sort of entity it would have to be. The inationary fetish that ascribes to propertyand ownership an intrinsic, unexamined, and thinglike status 116 thereforeadds nothing to our abstract understanding of property rights, nor does it shed any light on the actual processes of judicial reasoning. Commit it thento the ames: for it can contain nothing but sophistry and illusion.

115. See, e.g., Thomas Merrill, The Property Prism, 8 ECON J. W ATCH 247254 (2011), at 252(The bundle of rights is a primitive metaphor, conjuring up a medieval peasant carryinga faggot of wood. It is time for a better metaphor.); Robert C. Ellickson, Two Cheers for the Bundle-of-Sticks Metaphor, Three Cheers for Merrill and Smith , 8 ECON J. W ATCH 215222 (2011),at 216 (I therefore urge legal commentators not to abandon the bundle-of-sticks metaphor,but rather to be aware of its limitations and to invent complementary metaphors that might counter the bundles shortcomings.).

116. Cf . H ARRIS, PROPERTY , supra note 82, at 256.

Documents

1 Back to Bundles: Deflating Property Rights, Again