Accession, Property Acquisition, and Libertarianism

1. Introduction

If we take Robert Nozick as our starting point, then the following methods of appropriation “exhaustively cover the subject of justice in holdings”¹ as far as libertarianism² is concerned. First of all, one can appropriate unowned things “in accordance with the principle of justice in acquisition”³ or better said, original acquisition. Here the dominant view is that appropriation is effectuated by the Lockean method of mixing one’s labor with unowned resources, that is, homesteading.⁴ However, as pointed out by many libertarian authors,⁵ it is clear that if one mixes one’s labor with a thing one already owns, or joins several such things together or subtracts something from them etc., one also becomes the owner of the resulting product. Thus, we can say that the principle of justice in acquisition covers two distinct principles or methods, namely, homesteading and production. One can also appropriate things already owned by others “in accordance with the principle of justice in transfer”⁶ that specifies conditions under which property titles are voluntarily transferred by their actual holders to the new owners. Here the standard view is that unless the transfer is effectuated by coercion (as in the case of extortion) or ignorance (as in the case of fraud), it is voluntary. Finally, one can acquire title to holdings as a matter of “rectification of injustice,”⁷ for example, when the defendant is compelled to pay money damages to the plaintiff for an assault he committed. Now, even though each of these four methods of appropriation has its own complications and blurred lines, it is usually claimed that in one form or the other they exhaust the possibilities of property acquisition under libertarianism.⁸

Yet some libertarians are aware of cases which do not fall easily under any of these methods. For example, Murray Rothbard⁹ considers scenarios in which one person improves another’s car or erects buildings on another’s land without the latter’s consent and he contends that the distribution of property titles to the resulting products should not be governed by any of the above methods but rather be determined by the separability or inseparability of the improvements. Similarly, Walter Block¹⁰ discusses cases in which one party nonconsensually paints another’s house or carves a statue out of another’s granite and he proposes to solve these conundrums without deriving them from the aforementioned principles, intimating instead that the material should be more important than the labor. All of this suggests that, contrary to what libertarians explicitly claim, libertarianism does recognize one more principle of property acquisition, namely some version of the principle of accession.¹¹ Indeed, the main thesis of the present paper is that there is the fifth libertarian method of property acquisition, that it should be called by its name (accession) and that the most reasonable reconstruction thereof does not conform to Rothbard’s premonitions about separability of improvements or Block’s suggestions about the importance of the material but rather follows a different logic, often in tension with some other deeply entrenched libertarian beliefs.

Thus, generally speaking, the principle of accession has it that ownership of a thing produced from inputs supplied in good faith by two or more parties not bound by the contractual relationship should be assigned to the party whose input is most prominently connected with the thing while the supplier of the less prominent input should be compensated for his contribution by the owner of the resulting product.¹² For example, if A, reasonably supposing that he is dyeing his own wool, mistakenly mixes his dye with B’s wool, then under the principle of accession, assuming that B’s wool is more prominently connected with the dyed wool than A’s dye and labor (due to, say, its higher value or due to the fact that it persists through changes of color as the same substance),¹³ ownership of the dyed wool will be assigned to B, although he will have to compensate A for his dye and labor.¹⁴ Now what is important about this solution is that B acquires property rights to the resulting product not on the basis of homesteading (the wool was owned by B at the beginning of the entire process and it persisted as the same substance through subsequent changes of color), production (B’s wool was mixed with another’s labor and dye, not with B’s own), voluntary transfer (A mistakenly mixed his labor and dye with B’s wool and thus did not transfer his title to B) or rectification of injustice (A acted in good faith and improved B’s property beyond anything he could possibly owe B for his harmless – if it happened at all – trespass), but on the basis of the fact that A’s less prominent input accedes to B’s more prominent resource.¹⁵

In contradistinction, the received libertarian view has it that since A’s dye and labor are inseparable from B’s wool, A should lose his rights to the inputs without compensation. However, the received libertarian view is unable to explain why exactly A should lose his rights to the inputs without compensation. Or, in other words, it is unable to explain how the four libertarian principles of justice mentioned above actually justify such a verdict. For example, the received view could try to invoke the principle of the rectification of injustice and argue that even though non-culpably, A nonetheless unjustly used and modified B’s wool and so should forfeit his rights to the inputs (although if the dyed wool is more valuable than the undyed one, no additional rectification for B might be needed). Yet, even if A indeed committed a trespass by unjustly using and modifying B’s property (an allegation that is actually not made by libertarians writing about such cases), A’s infringement should be dealt with in a proportional manner¹⁶ and there is no guarantee whatsoever that forfeiture of valuable accessories would be a proportional reaction to a harmless and non-culpable trespass.¹⁷ Accordingly, the alleged fact that A committed a trespass does not entail that he should lose his rights to the inputs. To the contrary, it is perfectly possible for A to retain his rights to the inputs despite committing a trespass and only pay negligible money damages for his harmless and excusable infringement. Therefore, some independent argument would have to be presented in order to establish the contention that A loses his rights to the inputs without compensation and that B acquires them in accordance with the traditional libertarian methods of property acquisition. However, as we argue below, none of the libertarian principles of justice is able to provide such an argument and thus it is plausible to conclude that libertarianism recognizes the fifth method of property acquisition, namely, the method of accession, which should only be properly reconstructed along libertarian lines.

2. Libertarianism and the Accession Principle

As we pointed out above, some libertarians are aware of cases which do not fall easily under any of the traditional libertarian methods of property acquisition. For example, developing his famous left-libertarian argument for an equal share in natural resources, Hillel Steiner goes through various scenarios that are supposed to buttress his final point that self-ownership is “a sufficient basis for creating unencumbered titles both to things produced solely from self-owned things and to things produced from this equal portion of unowned things”¹⁸ but not to “products whose factors include unowned things.”¹⁹ Now the most interesting scenario discussed by Steiner – at least for our purposes – concerns restoring an antiquarian copy of John Locke’s Two Treatises. As pointed out by Steiner, “if I contractually mix labour with your antiquarian copy…, I don’t thereby become its owner. It’s not the fruit of my labour, because the title to that labour has been contractually transferred to you.”²⁰ But then he asks the question of what if I were to “mix my (uncontracted) labour with… [a land that] belongs to neither you nor me nor anyone else?”²¹ Responding to this query, Steiner admits that “[t]he answer, as we know only too well, is uncertain. For any claim, to the effect that its being infused with my labour makes this land mine, can be met with the counter-claim that, in so infusing the land, I was relinquishing my title to that labour,”²² as in the book-repair case. And to the possible counter-argument that whereas in the book-repair case I was contracted for relinquishing my labor, in the land case I am not, the response is “that even had the book-repair not been contracted, my nonetheless repairing it would still have constituted a relinquishment of my labour since the book wasn’t mine.”²³

However, there seems to be an important difference between the case in which, being aware that the book belongs to you, I nonetheless decide to repair it without your consent and the case in which, being ignorant about the fact that the book is someone’s property or mistakenly thinking that it belongs to me, I proceed with the restoration.²⁴ For in the former case I willfully expend my labor on the property of another, relinquishing thereby not only the labor itself but also facing no obstacle whatsoever – be it coercion, ignorance or mistake – for my ownership title to this labor to travel to you as well.²⁵ In the latter case, on the other hand, even though I infuse your property with my labor, thereby relinquishing it from my possession, there is no passage of title because “[f]or that waiver-generated transfer to be normatively valid – for that waiver to effect the transfer of the right in question – it is necessary that it be done voluntarily.”²⁶ Since I act in ignorance or mistake about the legal status of your property and since for a “waiver to be normally valid, [it] must be performed non-ignorantly,”²⁷ mixing my labor with your property cannot result in the title transfer. Thus, although it might indeed be uncertain who should own the book if I repaired it in good faith (more specifically, acting in non-culpable ignorance about its legal status), one thing is certain: whoever should own it, his title will be encumbered by another’s claim. For if you should own it, then your title will clearly be encumbered by my claim to the labor embedded in the book. If, on the other hand, I should own it, then my title will even more clearly be encumbered by your claim to the material of the restored book.

Now some libertarians might want to oppose this analysis by demurring to the idea that I still own my labor after it has been applied to, mixed with or embedded in another’s property and thus that I own the improved property or have a claim to it. They might say something along these lines: Well, we agree that you own your labor energy because you own yourself but not after you applied this labor to the things we own. Accordingly, you do not own or have a claim to the improved things either. By saying this, they might either mean that I cannot own my labor power once it is separated from my body and applied to external resources, regardless of the legal status of these resources,²⁸ or that I lose ownership of my labor exactly because it is applied to another’s property rather than to my own or to unowned things.²⁹ If the first option is the case, then they should readily see that their argument does not apply to cases in which it is not (only) my labor but (also) my tangible property that is inextricably mixed with another’s resources and thus that it cannot be generalized to provide a successful refutation of the accession principle as such. Besides, their criticism of the idea that one cannot own one’s labor once it is separated from one’s body and applied to external resources is, as we argue in our other paper,³⁰ unsuccessful, mainly due to perfect translatability of matter and energy.³¹

If, on the other hand, the second option is the case, then they face exactly the problems discussed in the present paper. Most of all, they have to grapple with the question of why mixing one’s labor with an unowned thing gives one a claim to the thing whereas fails to do so when the thing belongs to another. For acquiring a claim to a thing via labor-mixing has much more to do with a way in which labor-mixing is accomplished than with the legal status of the thing. For example, if one mixes one’s labor with an unowned thing but has the intention to abandon one’s labor rather than acquire ownership of the thing, one will not become its owner. If, however, one is coerced to mix one’s labor with an unowned thing, one’s labor will physically attach to the thing but one’s title to the labor will not be waived and thus the thing will contain something that belongs to the laborer. It is exactly for this reason that, for example, Rothbard believed that “the feudal or monopolist landlords have no legitimate claim to” their land and that it should be the coerced laborers, that is, “[t]he current ‘tenants,’ or peasants [who] should be the absolute owners of their property.”³² However, the same logic also applies to coerced or defrauded laborers who mix their labor with another’s property. Their labor physically attaches to another’s property but their title to this labor does not pass to the owner and so they have a claim to his property, a claim that is usually transformed into a compensatory interest.³³ Thus, it is not clear why it should be the case that the legal status of the thing should unilaterally determine what happens with the title to the worker’s labor. Rather, the opposite is the case – it is the way in which the worker mixes his labor with external resources that is crucial.³⁴

Thus, returning to Steiner’s book repair case, the fact that your title to the book will be encumbered by my claim to the labor embedded in your book explains why the case at hand does not fall easily under any of the traditional libertarian principles of justice or why the answer to the question of who should own the repaired book, “as we know only too well, is uncertain” for libertarians. For suppose that libertarians would like to assign ownership of the restored book to you. On what basis would you thereby appropriate my labor embedded in the book? Since I did not relinquish my title to this labor, you could not possibly acquire it via the method of original appropriation. Your production of the repaired book is trivially out of the question as it was not you who improved the book and the labor mixed with it was not yours either. By the same token, there was no contract between us. Rectification of injustice is also blocked by the fact, elaborated further below, that I did not damage your book – I actually improved it – and that I acted in good faith. If, on the other hand, libertarians would like to assign ownership of the restored book to me, these problems would only be exacerbated. Thus, in whichever way they would like to distribute exclusive ownership in this and similar cases, the solution would have to be based on a different principle than traditional libertarian methods of homesteading, production, voluntary transfer and rectification.

Similarly, Rothbard considers the case of a man who unknowingly purchases a stolen good and improves it. Who should be the owner of the improved good and should the improver be compensated for his bona fides improvements? For Rothbard, “the answer depends on the moveability or separability of these improvements.”³⁵ If the improvements are not moveable or separable, then the owner of the stolen good should become the owner of the improvements while owing no compensation to the improver. Thus, in the scenario in which Brown steals a car from Black and sells it to Robinson who in turn adds a radio to the vehicle, Rothbard believes that “since the radio is separable from the car, he should be able to extract the radio as legitimately his own before returning the car to Black,”³⁶ suggesting thereby that Robinson is not allowed to do so after the car is returned to Black. By the same token, “if the addition is not separable, but an integral part of the property (e.g., a repaired engine), then Robinson should not be able to demand any payment or property from Black.”³⁷ And the same goes for the scenario in which Brown “steals” a land and sells it to Robinson. Here too, “the criterion should again be the separability of any additions Robinson had made to the property. If, for example, Robinson had built some buildings on the property, then he should be able to move the buildings or demolish them before turning the land over to the original landowner.”³⁸

Setting aside for the time being the problematic nature of the separability criterion, the question arises: On what basis does Black acquire the title to Robinson’s inseparable improvements? Certainly, since Robinson did not waive his title to the improvements due to his ignorance, Black could not homestead them. Nor did he produce them himself from self-owned things. Similarly, Black did not acquire the title to them by way of voluntary transfer from Robinson (or more specifically, taking into consideration the fact that Black does not owe Robinson any payment in exchange, by way of gift). Finally, Rothbard does not even suggest that Robinson should lose his improvements as a matter of rectification of injustice. For one thing, if bona fides improvements of Black’s property constituted an injustice, then their separability or inseparability would not matter one iota. And yet, according to Rothbard’s criterion, if they were separable, Robinson should recover them without any compensation to Black. Hence, Black must have acquired the title to the improvements by some other method than original appropriation, production, voluntary transfer or redress. In a sense, Black appropriated these improvements only because he owned the car, that is, only because his input was more prominently connected with the final product than Robinson’s repair job.

Now if we imagine the reverse scenario to the one considered by Rothbard, we can see this even more clearly. Thus, suppose that this time Brown stole an energy storing power supply device from Black and sold it to Robinson who in turn charged his own electric car using the energy stored in the device so that it became inseparably mixed with the car. Would now Rothbard also be willing to say that “[i]n our view, then, the [energy] must be returned immediately to the true owner, Black,”³⁹ even though doing so would mean, per inseparability, giving Robinson’s car (or even only the car battery) to Black? Cleary not, for saying so would be so intuitively unjust that no further argument would be even needed to reject the theory yielding such a prediction. Instead, it is much better to conclude that it is Robinson who should become the owner of Black’s stolen energy while compensating Black for his loss. But if so, then again, Robinson’s acquisition of Black’s stolen energy would not be effectuated by original appropriation, production, voluntary transfer or redress. Rather, Robinson would appropriate Black’s energy only due to Robinson’s input, that is, Robinson’s uncharged electric car being more prominently connected with the charged car than Black’s energy.

Also, Block⁴⁰ deals with cases in which one party acquires ownership of another party’s labor and materials by virtue of the latter party mixing these resources with the former party’s property, thereby improving it and increasing its value. Although Block considers these cases under the headings of unjust enrichment and proceeds of crime,⁴¹ the basic point is essentially the same as in Rothbard, that is, the owner of a more prominent property becomes the owner of its improved version (and so also of the improvements themselves), even though he did not acquire it via any of the traditional libertarian methods of property acquisition. In a sense, the owner of a more prominent property continues to own it in the improved state (and thus he thereby appropriates the improvements) just because his unimproved property was a more prominent input to the improved property than the improvements themselves. For example, considering a case in which a house owner, A, hires a painter, B, to paint his house, but B makes “the error in good faith”⁴² and “accidentally paints C’s house,”⁴³ Block argues that “[t]he fact that B’s paint job in some objective manner increased the value of his home is entirely beside the point. To force him to pay B would involve positive obligations, anathema to the freedom philosophy.”⁴⁴

But how is the conclusion that C acquired the title to B’s labor and paint embedded in the improved house without any compensation to B and without going through any of the traditional libertarian methods of property acquisition compatible with libertarianism? For notice that C did not acquire B’s paint and labor in accordance with any of the traditional libertarian methods of appropriation. As we pointed out above, the fact that B made a non-culpable mistake excludes the possibility that he validly waived the title to his paint and labor. Thus, it is out of the question that C originally appropriated B’s inputs.⁴⁵ And even more out of the question is that C acquired B’s paint and labor by way of production. By the same token, it is also clear that C did not enter any contractual relationship with B (one reason for it being that he had a contract with A rather than with B) due to which he would be able to appropriate B’s factors of production, for again, B’s actions were involuntary by way of B’s non-culpable mistake. Finally, note that Block – as with Rothbard in the cases of improving another’s car and building a house on another’s land – does not try to argue, and rightly so, that B in any way violated C’s property rights and thus that C acquired B’s inputs in accordance with the principle of rectification of injustice. As we pointed out earlier, this route is blocked for libertarians because, first of all, it is not even clear whether – even assuming otherwise problematic theory of strict liability which would have to be embraced in order to ignore B’s non-negligent mistake⁴⁶ – B’s actions can be classified as a harmless trespass, for not only B did not lower the value of C’s house, but also “B’s paint job in some objective manner increased the value of his home.”⁴⁷ However, even if B’s actions could be so classified, there would still be no guarantee that they should result in forfeiture of B’s labor and paint. The reaction to B’s trespass should be proportional⁴⁸ and forfeiture of B’s inputs might easily exceed the limits of proportional reaction. Thus, traditional four methods of property acquisition are not enough to explain C’s appropriation of B’s inputs. In the puzzle at hand C appropriates these inputs just because he owns the most prominent property, that is, the house. Since he owns the house, he also owns the inputs which only modify what he already owns. Yet, this way of appropriating inputs does not fall into any of the traditional libertarian methods of property acquisition. And so, Block’s (and Rothbard’s, for that matter) solution to the puzzle can be compatible with libertarianism only if libertarianism recognizes the fifth method of acquiring property titles, namely the method by which ownership of a thing produced from inputs supplied by two or more parties vests in the party who supplied the most prominent input and so has the strongest claim to the final product – in a word, the method of accession.⁴⁹

Now in order to see even more clearly that libertarianism must assume some version of the accession principle to deal with cases which do not fall easily under any of the traditional libertarian principles of property acquisition, consider the following loci classici of natural and artificial accession discussed in The Digest of Justinian and ask yourself the question of how libertarianism would cope with them.

(1)

River. “[I]f the force of the river should detach part of your land and bring it down to mine… [and] if it adheres to my land, over a period of time, and trees on it thrust their roots into my land,”⁵⁰ who should be the owner?

And similarly, but instead of the river abruptly and perceptively detaching part of your land and attaching it to mine, the process is imperceptible:

(2)

Alluvion. Who should be the owner of “what the river adds to our land by alluvion… [which] is that which is gradually added so that we cannot, at any given time, discern what is added?”⁵¹

If the examples discussed above – the car and the house – created an impression that the final product should accede to the passive party whereas the active party should only be compensated for her loss of the input or, what comes to the same thing, that the passive party’s claim to the final product is always stronger, then the present scenario debunks this hypothesis entirely.⁵² For neither party is active in this case and yet libertarians still have to identify the stronger claim if they want to distribute exclusive ownership to the land. At the same time, the fact that neither party is active renders traditional libertarian methods of property acquisition essentially inapplicable to this puzzle for the simple reason that unless there is an action, no homesteading, production, title transfer or violation of property rights can take place.⁵³ Moreover, libertarians cannot circumvent this problem by suggesting that each parcel should remain under separate dominions.⁵⁴ For not only would it compromise respective owners’ rights to possess and enjoy their parcels, but also fail to solve the very puzzle at hand. After all, as we can learn from Paul, once a tree “takes root, it becomes part of the land where it does so, and should it be uprooted, it does not revert to its former owner; for it has conceivably altered through nourishment in other soil.”⁵⁵ Thus, to put the same point in a more libertarian vernacular, since the land is my property throughout, your tree taking nourishment from my land becomes infused with this property and so I acquire a claim thereto as well.⁵⁶ We want to know how to divide these claims.

Hence, in whichever way libertarians would like to distribute exclusive ownership to the resultant parcel, their solution will have to be based on some version of the accession principle according to which private property rights vest in the party whose input is more prominently connected with the final resource. By way of example, Gaius’s solution in River was that “if the force of the river should detach part of your land and bring it down to mine, it obviously remains yours. Of course, if it adheres to my land, over a period of time, and trees on it thrust their roots into my land, it is deemed from that time to have become part of my land.”⁵⁷ Similarly, in Alluvion “what the river adds to our land by alluvion becomes ours by the law of nations.”⁵⁸

(3)

Parchment. “If I wrote verse or a story or a speech on your paper or parchment” with “[l]etters, even in gold,”⁵⁹ who should own the improved parchment, assuming (somehow superfluously) that gold letters are inseparable from the final work and (this time nonredundantly) that I acted in good faith?

If the answer is that I should own it, then clearly my appropriation of your parchment is not governed by the libertarian principles of homesteading, voluntary transfer, production or rectification. If, on the other hand, the answer is that you should own it, then your appropriation of my labor and gold ink is not governed by these principles either. Let us first consider the principle of rectification of injustice. If I culpably damaged your parchment, I would clearly owe you compensation. This condition, however, is not met because, by assumption, my actions were not culpable, I acted in good faith and improved your parchment by covering it with gold rather than damaged it. Yet, even without culpably damaging it I might have committed a harmless trespass by using your parchment without your consent if strict liability is assumed. Granted. But it is debatable whether I indeed committed such a trespass even under these assumptions, for not only did I not damage it, I actually improved it and multiplied its value. However, the crucial thing is that even if my actions did constitute such a trespass, we can be fairly certain that forfeiture of my valuable gold would not be the appropriate remedy for something that was only a harmless (not to say beneficial) infringement. Thus, the principle of rectification of injustice does not explain your acquisition of my gold. What about the principle of voluntary transfer? This principle is also blocked by the bona fides condition. If I write my story on your parchment mistakenly thinking that the parchment belongs to me, then even though my labor and gold ink do travel to your possession, titles thereto do not, for the voluntariness of the title transfer is compromised by my mistake. Hence, if you were to acquire the property title to the final work, it would have to be on a different basis than the principle of voluntary transfer because I did not voluntarily transfer the title to the improvements to you. And since the principles of homesteading and production are entirely beside the point in this case, we can conclude that whichever solution we prefer, the principle controlling appropriation of the final work has to be some version of the accession principle according to which the final work accedes to the owner of the most prominent factor,⁶⁰ whichever factor it is, while the other party is compensated in money for providing the accessory. Incidentally, the solution preferred by Gaius was that the gold letters should accede to the parchment: “Letters, even in gold, accede to the paper or parchment just as things built or sown become part of the land. But should you claim your book or parchment from me but be unwilling to pay my writing expenses, I can, if I acquired your materials in good faith, resist you.”⁶¹

(4)

Picture. If I painted a picture on your tablet, who should own the final work, assuming (superfluously) that the paint is inseparable from the final work and (nonredundantly) that I acted in good faith?⁶²

Again, none of the four libertarian principles of justice can deal with this case properly (by the same reasoning as above). Since neither my property claim to the paint (and labor) nor your property claim to the tablet is extinguished (no waiver, no forfeiture), there are two conflicting property claims to the final product. Thus, the dilemma we face is how to distribute exclusive ownership rights to the indivisible thing to which there are two valid property claims. We need some justice principle that will be able to do so in a libertarian way (that is, with a proper respect for the property rights), even though no traditional libertarian method of property acquisition is suited for this task. Our thesis is that it is exactly the accession principle with its insistence on the prominence of factors that supplies such a solution. For it solves the dilemma with the sole focus on the property aspect of the conflicting claims (in contradistinction to, for example, the equality aspect), that is, by examining which input is more prominent than the other (in contradistinction to, for example, examining who needs the final product more or which distribution would bring about more equality). Now what is interesting about Picture is that it provides an opportunity for considering whether the principal thing should be identified along the same lines as in Parchment and thus for appreciating the intricate nature of the accession principle. Gaius thinks that it should not: “Pictures do not accede to the tablets on which they are painted in the same way as writing to paper or parchment. On the contrary, the view established itself that the tablet accedes to the picture.”⁶³ According to Thomas Wood, “this is so order’d because of the great Esteem the Ancients had for this Art.”⁶⁴ On the other hand, before Justinian finally decided what the established view should be on this case, a dispute had existed amongst Roman jurists. As pointed out by Ernest Lorenzen, “Paul put the case of painting on the same footing as that of writing, regarding the material on which the painting was made as the principal thing.”⁶⁵ Thus, it is an open question which input should constitute the principal thing and libertarians, the same as Romans, can argue about the proper criteria – the problem that we discuss in the next section. What they cannot argue about however is that some version of the accession principle (that is, some version of the prominence of factors criterion) must be adopted in order to solve such puzzles.

(5)

Statue. “When someone makes something for himself out of another’s materials,”⁶⁶ for example, a statue out of another’s marble, who should be the owner of the statue, assuming that the same two conditions of inseparability and bona fides hold?

This is an instance of the so called specificatio⁶⁷ where a manufacturer creates a thing of a new species from another’s material rather than only improves an existing thing by attaching something to it as in the above-discussed cases of classical accession. At first glance, libertarians might have an intuition that the statue should belong to the owner of the marble, rather than to the laborer. After all, the statue is made entirely out of another’s private property. Thus, how can the owner lose his title without any act of his, be it waiver or forfeiture? By the same token, how can the laborer acquire another’s private property without his consent – it seems deeply unlibertarian. Indeed, considering a similar case in which a manufacturer creates a magnificent statue (a thing of a new species) from a hunk of granite he stole – an aspect that renders it an instance of criminal or fraudulent specification⁶⁸ – from another’s quarry, Block argues that the laborer “must be compelled to return that specific carved rock”⁶⁹ without any compensation for his labor.

However, the discussion that took place in Roman times over the doctrine of specificatio furnishes us with at least two arguments against this prima facie libertarian position. First of all, an argument can be made that the statue is a new thing that did not exist before and so that the manufacturer is not only its creator but also its original appropriator. Indeed, as pointed out by Paul von Sokolowski,⁷⁰ it was the Aristotelian philosophy of Proculians that prompted them to assign ownership of the final product to the laborer in their debate with Sabinians. Since it was the form of a thing that counted as its essence for Aristotle, then bestowing a new form on a thing amounted to creating an entirely new thing that had not existed before. In turn, this implied that the old thing, that is, the material on which the manufacturer worked, must have ceased to exist. Accordingly, after the laborer’s transformation, there was nothing left to which the owner of the material could keep the property title. As Lorenzen puts it, “the metaphysics of Aristotle, accepted by the Proculians, with its emphasis on the form in the creation of things, naturally led them to regard the old material as destroyed, with the result that the owner thereof could not bring the rei vindicatio action against the manufacturer.”⁷¹ Thus, an essentially libertarian point can prima facie be made that the laborer acquires ownership of the final product by way of original appropriation. For if the manufacturer created an entirely new thing that had not existed before, then he also was the first man who occupied it or, as libertarians would like to call it, homesteaded it.⁷² Indeed, Gaius says: “Nerva and Proculus are of opinion that the maker owns that thing because what has just been made previously belonged to no one.”⁷³ And again, as pointed out by Lorenzen, many Pandectists also supported the account that “the old thing is destroyed and that ownership in the new thing is acquired through occupation.”⁷⁴

Second of all, the assignment of the statue to the manufacturer as it was promoted by Proculians is sometimes interpreted in accordance with a different, also libertarian-spirited, theory. To wit, as pointed out by Lorenzen, it is sometimes contended that “the Proculians, in regarding the manufacturer as the owner of the new product, were actuated by economic considerations out of respect for industrial labor.”⁷⁵ This idea, of course, leads us to an essentially Lockean argument that giving the title to the owner of the materials – as suggested by Block – would neglect the manufacturer’s ownership of the labor mixed with these materials. It is therefore difficult to understand why Block who explicitly embraces the Lockean idea that labor can be owned and that labor ownership is the root of all external property rights should so quickly dismiss the possibility of recognizing the manufacturer’s ownership of the statue or at least his claims to compensation for the improvements.

3. Contours of the Accession Principle under Libertarianism

The crucial point of the above discussion is that if libertarians want to transform one of the competing claims to the final product into the ownership interest, they have to employ some version of the accession principle. Otherwise, they will be unable to do so without running into conflict with their own principles of justice. In this section, on the other hand, we would like to elucidate more specific contours of the libertarian version of the accession principle. Now since Rothbard’s separability criterion mentioned above is the only attempt in the libertarian literature to offer any such contours whatsoever, it is also the most natural starting point for our discussion. Thus, the first question is whether the Rothbardian separability criterion can determine the contours of the libertarian principle of accession?

The most obvious problem with this criterion is – as it should be clear from the preceding discussion – that in the case of inseparable improvements the manufacturer “should not be able to demand any payment or property from”⁷⁶ the owner of the material. However, as we showed above, the bona fides improver – due to his ignorance or mistake – does not transfer the title to the improvements to the owner of the material and so has a valid claim to the final product. Hence, leaving things as they are, namely without “any payment or property from” the owner of the material, would result in a deadlock, that is, in the very opposite of what the institution of private property is supposed to achieve in the first place.⁷⁷ For the owner of the material would not be able to do anything with the final product without the improver’s permission or else he would infringe upon the latter’s non-extinguished property rights to the improvements (and vice versa). Thus, if libertarians want to – as they do – avoid such deadlocks and assign the exclusive ownership of the improved resource to a specific party, they have to transform one party’s property claim into a liability claim,⁷⁸ thereby giving this party a Hohfeldian power⁷⁹ to demand a compensatory payment from the other party. Now it does not take much to generalize this complication to cases of separable improvements that are not recovered before the final product is returned to its owner. For in such cases the bona fides improver does not transfer the title to the improvements either and so he keeps property rights thereto also after the final product is returned to its owner.

However, the most serious problem that besets the separability criterion consists of its inherent inability to determine which input is more prominently connected with the final product and so who should be the owner of the resulting property. Notice that both in the case in which the improvement is inseparable (as the repaired engine) and in the case in which it is putatively⁸⁰ separable (as the radio) but not recovered before the return, the separability criterion assigns ownership of the improved product in the same way, that is, to the owner of the car. Thus, it is not separability or inseparability of the improvements that determines which input is more prominently connected with the final product and who should be its owner. There must be some other criterion – unrevealed by Rothbard – operating in the background. Similarly, if we consider what we called the reverse scenarios, we come to the same conclusion about the implicit criterion doing the heavy lifting here. For example, if Robinson purchased a stolen power supply device from Brown and then charged his own electric car using the energy stored in the device, should he lose the car if giving it to the victim were the only way of recovering the stolen energy now inseparably mixed with the car? Should the victim’s seizing the car not be considered a car theft rather than a recovery of his energy? Clearly, the ownership distribution in such cases must be governed by some other criterion than separability of improvements. Now the question is what is this criterion as far as libertarianism is concerned?

Judging from the above discussion, four characteristics of such a libertarian accession criterion should be relatively easy to pinpoint. First, it should only apply to indivisible things. For if the final product can be divided between the input suppliers, then the question of its ownership does not even arise as each supplier remains the owner of his respective input and can recover it from another’s possession. Accordingly, there is no need to transform anyone’s property claims into liability claims. Rather, everyone can take what is his and go his own way. Thus, for example, if Robinson, acting in good faith, installed a radio in Black’s car, then the radio – pace Rothbard – remains Robinson’s private property even after the car is returned to Black.

Second, regardless of who becomes the owner of such an indivisible thing, the libertarian principle of accession must also compensate the other party for his input unless forfeiture of this input serendipitously matches what is due as a matter of redress for a harmless trespass (if we are still convinced that such a trespass happens despite the fact that the property was improved and the improver was not negligent), although the harmless trespass caveat does not apply at all to cases such as River and Alluvion where there is no human action whatsoever. In other words, the other party’s property interests in the input must be transformed into this party’s liability interests in compensatory payment. This condition follows from the fact that acting in good faith, the party in question does not transfer the property title to the input to the owner of the final product. On the other hand, inputs become merged into an indivisible thing and so can no longer be subjected to separate dominions of respective owners. Thus, in order to avoid deadlocks and at the same time respect earlier entitlements of input providers as far as possible, property claims of the accessory supplier must be transformed into liability claims.

Third, the libertarian principle of accession should consider labor to be as important a category of input as physical property. After all, self-owned labor is the very source of all property titles for libertarians.⁸¹ As pointed out by Rothbard, “all production reduces back either to labor services or to finding new and virgin land and putting it into production by means of labor energy.”⁸² It is exactly due to the fact that “our bodies produce energy… [which then is] infused into parts of the external environment, transforming their features in various ways… [that] we claim these things for ourselves as the fruits of our labour.”⁸³ Thus, labor inputs must be taken into account to the same extent as material property in determining who should be the owner of the final product. In consequence, materialistic theories – such as Sabinians’ or Block’s – that in principle would like to assign ownership of the resulting property to the owner of the material should be rejected out of hand.

Four, the libertarian principle of accession should not pay much attention to the distinction between active and passive input suppliers. For example, it should not matter for ownership distribution that the mechanic repairing the car engine was active whereas the car owner was passive, for the sculptor of a magnificent statue was also active while the provider of the incomparably less valuable slab of granite was passive and yet it is far from clear that it is the latter who should own the statue. In other words, the active/passive distinction cannot be important for the question of which input is more prominent or who should be the owner of the final product because: (1) it has no application whatsoever to cases of natural accession such as River or Alluvion which are only passive; (2) it leads to counterintuitive results in the so called reverse scenarios such as the stolen power supply device; and (3) by focusing solely on the question of who was active and who was passive, it fails to distinguish between crucial qualities of action such as voluntariness, ignorance or duress that determine whether a waiver or forfeiture takes place (for example, a defrauded sculptor is active but he does not waive the title to his labor and so his claim to the statue cannot consistently be dismissed only because he was active whereas the granite owner was passive).

Thus, the final question is, respecting these four conditions, which specific criterion should govern property acquisition by way of accession under libertarianism? As one can judge from the above discussion, two candidates clearly come to the fore: (1) specificatio via first occupation of a new thing as in the case of a statue made from someone else’s marble or wine made from the grapes of another; and (2) disparity of value via the contribution of a more valuable input as in the case of a picture painted on someone else’s tablet or a house built of someone else’s bricks. Now, although these two criteria do overlap in many cases, it is important to differentiate between them and to learn which one is superior in solving accession puzzles along the libertarian lines. In order to do that, let us make recourse to some cases from Roman law.

(1)

River. “[I]f the force of the river should detach part of your land and bring it down to mine… [and] if it adheres to my land, over a period of time, and trees on it thrust their roots into my land,”⁸⁴ who should be the owner?

As we can readily see, the first occupation theory does not deal well with such puzzles. For one thing, since there is no change of species, the resulting land cannot be considered a new thing. Accordingly, my occupation thereof cannot classify as first occupation and so I cannot be considered its owner under the first occupation theory.⁸⁵ However, you cannot be considered its owner either. After all, even though your part of land preserved its identity, the resulting riverbank also contains my parcel which preserved its identity too. This, in turn, gives us another deadlock: each of us should remain the owner of his respective parcel while they merged into an indivisible thing. On the other hand, under the disparity of value rule the detached part of your land should accede to mine as my entire real estate is ex hypothesi more valuable than a mere part of your land.

(2)

Building. “When someone builds on his own site with another’s materials,”⁸⁶ who should own the building?

Now, although the building is indubitably a new thing that did not exist before, it is difficult to maintain that another’s materials, for example, bricks or logs, ceased to exist. Quite the contrary, it is relatively easy to identify specific bricks and logs in specific places of the structure. As a result, the first occupation theory is unable to distribute property rights to the building. For, on the one hand, it seems that the building is a new thing that can be originally occupied by its builder. However, on the other hand, the individual bricks and logs did not lose their identity and so do not seem to be eligible for first occupation. Thus, on the first occupation theory it seems impossible to say who should own the building. And even if the second argument seems stronger than the first – after all, the individual bricks and logs did not cease to exist and thus cannot be originally occupied – it is worth noticing how counterintuitive a verdict it would spawn. Indeed, no authority would like to give ownership of the building to the owner of the materials. Thus, Gaius thinks that in such a case the party who “builds on his own site with another’s materials… is deemed to be owner of the building.”⁸⁷ But the reason for such a solution has nothing to do with the builder’s alleged first occupation of the new structure. Rather, it is due to the value of his input, that is, the land on which he erects the building, for “all that is built on it becomes part of the soil… [and] no one is required to give up materials of another built into his premises but that he must pay double their value.”⁸⁸ What is even more interesting, however, is the fact that “the owner of the materials does not thereby lose his ownership of them; but he meanwhile cannot bring a vindicatio for them or an action for their production.”⁸⁹ Yet, “if the house should collapse for some reason, the owner of the materials can have a vindicatio for them and have an action for their production.”⁹⁰ Which is exactly what one should expect if the identity of the materials is preserved through the changes.

Similarly, although much more informatively, when writing an opinion for the court in a famous common law case Wetherbee v. Green dealing with the problem of accession, Justice Thomas M. Cooley pointed out that it is rather easy to track the identity of specific materials as they become incorporated into a new thing such as a house or church organ. Nevertheless, no one would be willing to assign ownership of the thing to the owner of the materials. Instead, it would be the value of land or labor that would be controlling the distribution of property interests. Thus, according to Cooley, J., even though oftentimes “no difficulty will be experienced in determining the identity of a piece of timber which has been taken and built into a house… no one disputes that the right of the original owner is gone in such a case.”⁹¹ This is so because “no one would defend a rule of law which, because the identity could be determined by the senses, would permit the owner of the wood to appropriate [the thing] a hundred or a thousand times the value of his original materials.”⁹² Instead, “[t]he owner of a beam built into the house of another loses his property in it, because the beam is insignificant in value… as compared to that to which it has become attached.”⁹³ And the church organ “belongs to the maker rather than to the man whose timber was used in making it, not because the timber cannot be identified, but because in bringing it to its present condition the value of the labor has swallowed up… the value of the original materials.”⁹⁴ Accordingly, “[t]he labor, in the case of the musical instrument, is just as much the principal thing as the house is”⁹⁵ in comparison to the beam.

The following case places a considerable strain on the first occupation theory as to its ability to both yield equitable results and to determine whether there is a change of species:

(3)

Butcher. If a butcher kills a pig of another and cuts it into joints, who should own the joints?⁹⁶

At first glance, it might seem that the pig has been destroyed – from the point of view of Aristotelian hylomorphism, the pig has lost its form of a living animal, or so it seems – and that the butcher has appropriated the joints via first occupation. However, as pointed out by Lorenzen, a majority of Roman jurists were of the opinion that “the killing of a pig and cutting it into joints did not constitute specification; whereas the making of sausage therefrom was regarded as a new species.”⁹⁷ But why would joints not be considered a new species whereas sausages would, if not for the fact that killing and cutting rather than sausage making is markedly less valuable than the pig itself and that subsuming the former under the doctrine of specificatio would yield inequitable results? Thus, as it seems, it is the disparity of value criterion, posing as failed specificatio, that controls the distribution of property rights in this case.

(4)

Parchment and Picture. “If I wrote verse or a story or a speech on your paper or parchment,”⁹⁸ who should be its owner? And if I painted a picture on your tablet, who should own the final work?

Is the original parchment destroyed in the process of writing on it? It does not seem so, and yet the writer indubitably created a new thing that did not exist before, that is, a story or a speech. So, under the first occupation theory, who should own the resulting product? Turning now to the tablet. If the parchment is not destroyed, then neither is the tablet – at least this much is true. But equally surely, the painting is a new thing too. So, does the painter appropriate the painting by way of first occupation or not? And should it not be the case that if the painter acquired property rights to the painted tablet, so did the writer to the written parchment (and vice versa, if the painter did not, neither did the writer)? Now a curious thing about these two cases is that Roman jurists solved them differently. Thus, we can learn from Gaius that “[p]ictures do not accede to the tablets on which they are painted in the same way as writing to paper or parchment. On the contrary, the view established itself that the tablet accedes to the picture.”⁹⁹ The reason for this apparent inconsistency seems to be twofold. First of all, instead of being subsumed under the doctrine of specificatio, these cases were regarded as instances of accessio, for, apparently, it was either difficult to tell for sure whether they were cases of creating a new species or treating them as such would yield inequitable results (or both). Second, looked at as accessio cases, they were distinguished from each other by differential value that parchments and tablets presented to the Romans. As pointed out by Zhang, some rudiments of the disparity of value rule already existed in Roman times and so one way of solving “the discrepancy between the written parchment and painted tablet” consisted in pinpointing that “whatever was written on the parchment was not considered to be substantially more valuable than the parchment, whereas the value of the painting was considered to dwarf the value of the ‘worthless’ tablet.”¹⁰⁰

We could keep multiplying the cases in which the first occupation theory is either indecisive or yields unfair results whereas the disparity of value criterion easily leads to equitable verdicts. However, at this stage it should be clear that from the two contenders, the disparity of value fares much better as the libertarian accession principle. If we return to the cases considered by Rothbard (improving the car of another and building a house on another’s land) and Block (painting the house of another and carving a statue out of another’s granite) and adjust for some infelicitous points and differences between bona fides and fraudulent accessions, we will see that it is the disparity of value that is doing the heavy lifting in these cases. The engine repair accedes to the car and not the other way around because the car is much more valuable than the repair. The house accedes to the land because the land surpasses (if it does) the house in value. The paint accedes to the house due to the house being exceedingly more valuable than the paint. And the statue accedes to the sculptor because it is, as Block puts it, “magnificent” and so much more valuable than a mere “hunk of granite.”¹⁰¹

4. Conclusions

In the present paper we argued that besides four traditional methods of property acquisition – that is, homesteading, production, voluntary transfer, and the rectification of injustice – libertarianism also recognizes a fifth method, namely the accession principle. More precisely, we submitted that if libertarianism wants to distribute exclusive property rights – as it does – to indivisible things produced from inputs supplied outside a contractual relationship by two or more parties, it must embrace some version of the accession principle. Additionally, we pinpointed specific passages in which libertarian scholars implicitly recognize the possibility of appropriation via accession, although their suggestions as to the exact contours of the accession principle prove to be problematic. Accordingly, we tried to determine these contours in a more systematic and straightforward way and argued that from the two best contenders – that is, specificatio by way of first appropriation and disparity of value – libertarians should opt for the disparity of value criterion as more versatile, definitive, and conducive to equitable results.

Acknowledgments

I would like to express my gratitude to Doctor Olga Peniaz from the Catholic University of the West in Angers, France for organizing a seminar devoted to the present paper during which I had an opportunity to receive many incisive comments from great scholars, including Jörg Guido Hülsmann from University of Angers, France and Dr. Peniaz herself.

I would also like to thank anonymous referees of this journal for their thorough examination of my paper and for pushing me on many points so that the present version of my work is much more developed and thought-out than the version I began with.