Original modes of acquiring ownership

This category is commonly divided into original ius civile and original tus gentium modes. This is yet another indication of the way in which the modern classification (original/ derivative) bisects the Roman categories.

7,2,1 Usucapio

(Inst.Gai.2.40.-61., Inst.2.6., D.41.3., 4.-10., C.7.22.-31.)

Modestinus, Encyclopaedia, book 5: Usucapion is the acquisition of ownership by continued possession for the period prescribed by law. (D.41.3.3.)

Usucapio was prescriptive acquisition—a person became owner of property by hold­ing it for the prescribed period of time (See Hausmaninger, Gamauf, and Sheets, Casebook, 124-49 as well as R (Lewis) v Redcar and Cleveland Borough Council & Anor [2010] UKSC 11). It was of great importance in practice since it allowed persons who possessed property to convert their interest into full dominium. The consequence was to reduce the problems caused by uncertainty over ownership:

Gaius, Provincial Edict, book 21: Usucapion was introduced for the public weal, to wit, that the ownership of certain things should not be for a long period, possibly permanently, uncertain, granted that the period of time prescribed should suffice for owners to inquire after their property. (D.41.3.1.)

The persons who benefited from usucapio were primarily those who had acquired property innocently, but who lacked domlmum over it through some flaw in their ownership. This situation primarily arose, for example, through an unintentional and undetected breach of the requirements of the mancipatio act (cf. Inst.Gai.2.41., :'43,). Usucapio gave the holder ownership, thus helping to reduce potential litiga­tion. The origins of usucapio can be traced to at least the Twelve Tables, which laid down that land could be acquired by two years' use, one year sufficing for /either things.

7.2.1.1 Possession

The holder had to have legal possession, i.e. the possession that was protected by interdict (see generally 6.5). Possession was the fundamental requirement of usucapio, as the very name signifies—'the taking by use'. It was insufficient for the holder to have mere custody (detentio) of the thing: he had to have possession.

7.2.1.2 Lapse of time

The requirement was for there to be two years' continuous possession of land, one year for other things. The periods seem to have been rather short, but are under­standable when it is remembered that Rome was relatively small in size and popula­tion at the time of the Twelve Tables (cf. Inst.Gai.2.42.; 44.). In such a community, where property was closely guarded, and where the phenomenon of the absentee landlord was probably an infrequent occurrence, the prescribed periods appear to have provided ample time for the true owner to interrupt the possession of the holder (see Prichard, A. M., 'Early Usucapio' (1974) 90 LQR, 234-45). Later, when Rome expanded, there was certainly an arguable case for lengthening the required periods. But they were not altered until Justinian’s reign, some 1,000 years after their original enactment. Justinian extended the periods to three years for movables and ten years for land (twenty, if the parties did not reside in the same province).

7.2.1.3 Continuity

There had to be continuous, uninterrupted possession for the required period. If the holder's possession was interrupted, he had to start again—he could not add the earlier period to the later one (see Gordon, W. M., ‘Interruption of Usucapio' (1962) 9 RIDA 3, 325-33, for a discussion of the case where a possessor alio nomine decides to possess suo nomine).

Gaius, Provincial Edict, book 21: Possession is broken in fact when someone is forcibly evicted from possession or the thing is seized from him. And, in such a case, the possession is broken not only against the person who seizes the thing but against everyone. (D.41.3.5.)

Civil interruption occurred when a legal claim was made to the property. In early law it seems that the very claim itself amounted to an interruption, but in classical law it was held that there was no interruption until judgment was given against the possessor. However, if the required period of possession elapsed between joinder of issue and judgment, the plaintiff was entitled to recover the property (if he proved his case) from the possessor since the judge had to make his decision based on the facts as they were at joinder. Although the classical rule concerning civil interrup­tions was clearly an improvement on that in early law, it was open to potential abuse: a possessor could try to delay proceedings against him so that time would. run in his favour. It is possible that Justinian restored the original rule that interrup- : tion occurred when proceedings were commenced.

In late classical law an exception was allowed to the rule that an interruption, necessitated the start of a fresh period of possession. If the possessor sold property, the buyer could add the seller’s period of possession to his own, providing that both had been in good faith at the commencement of their respective periods of possession:

Paul, Piautius, book 13: The period that the vendor possessed before the sale runs for the purchaser. (D.41.3.14pr.)

What if the possessor died before the lapse of the required period—could his heir add his predecessor's period of possession to his own?

Javolenus, Letters, book 4: The testator's possession runs for his heir, provided that there is no intervening possession by another.

The heir stepped into the shoes of the deceased: the period of possession of the deceased and his heir was regarded as one continuous period—death did not consti­tute an interruption. It was thus irrelevant if the heir was not in good faith, provid­ing that the deceased had begun his possession in good faith.

7.2.1.4 Good faith (bona f/des)

It seems that the requirement that the possessor must be in good faith was not recognized as an essential of usucapio until the later Republic. What was meant by good faith in this context? The basic test was whether the possessor believed that his possession was held honestly (cf. Inst.Gai.2.43.). For example, he would be in good faith, despite knowing that he was not the owner, if he held possession as the result of a praetorian interdict in his favour, or because he was a bonitary owner. Moreover, the possessor was in good faith if he mistakenly believed (as the bona fide possessor did) that he was the owner, providing that the mistake was one of fact, and reasonable in the circumstances:

Neratius, Parchments, book 5: But a man may usucapt a thing which he believes to be his, although his belief is unfounded. This, however, is to be understood in the sense that a rea­sonable and plausible error will not prevent the usucapion of a possessor... (D.41.10.5.1.)

Although the possessor had to have good faith, he did not have to prove it: good faith was presumed until the contrary was proved. As a general rule, good faith had to be present when the possession began, but did not have to continue throughout the required period: supervening bad faith was not fatal to usucapio. In the case of sale, the purchaser must have been in good faith both at the time of agreement and of delivery:

Ðàí³, Edict, book 54:... in the case of sale, we took also to the time of contracting so that a person must both have purchased in good faith and taken delivery in good faith. (D.41.4.2pr.)

\ Exceptionally, in the case of a gift, the donee must have had good faith throughout the required period of possession, but Justinian eventually brought gifts within the general rule.

Paul, Edict, book 54: Again, Pomponius says that in cases where slaves possess something in their master's name, it is the mind of the master rather than that of the slave to which we must look... (D.41.4.2.12.)

(There were some exceptional situations where good faith was not required by the (possessor. The most important case was that of the vacant inheritance (usucapio :p>ro herede). In some circumstances, heirs were not obliged to accept an inheritance (see 8.5.1.3). If an heir did not accept, or delayed in doing so, the inheritance was vacant and thus a res nullius open to the first taker—he would become its owner (despite the lack of good faith) after holding it for a year. This rule had some intelli­gible purpose: it was aimed at persuading heirs to decide speedily whether to accept inheritances, so that family rites could be performed, debts paid, and so that prop­erty should not lie vacant without an effective owner. Usucapia of vacant inherit­ances faded in importance following a reform by Hadrian that enabled the heir to recover the property even after the lapse of one year. And a special procedure was introduced for cases where a vacant inheritance was pillaged. This proved neces­sary because in theory theft could not be committed in such circumstances—the property was ownerless.

7.2.1.5 lusta causa

The taking of possession had to be the result of some transaction or cause which was normally a basis for lawful acquisition, e.g. sale, gift, inheritance, and exchange. It seems that this requirement was not recognized as essential until the later Republic. In usucapia (unlike truditio), the most commonly accepted rule was that mistaken iustci causa was insufficient.

Ulpian, Sabiims, book 31:... those people are mistaken who hold that if a man takes posses­sion of a thing in good faith, he can usucapt it as his own, and it is irrelevant whether he did or did not buy it, whether or not it was given to him, provided that he thinks he bought it or received it as a gift, because there is no effective usucapion unless there be, in truth, a legacy, a gift, or a dowry, although the recipient believes so. (11.41.3.27.)

As the passage makes clear, there was juristic disagreement on this issue. Some jurists took the view that mistaken lusta causa was sufficient. For example, Pomponios states:

Pomponius, Submits, book 22: You delivered to me a slave whom, erroneously, you thought that you owed me on a stipulation; if I know that nothing is due to me, I will not usucapt him; but if I do not know, the more correct view is that I do usucapt him, because the very delivery, on a ground which I believe to be true, suffices to bring about a result that I possess for myself what is delivered to me. This was the recorded view of Neratius and I think it to be correct. (1).41.10.3.)

T his view has the attraction of being consistent with the rule on putative iustu causa in truditio (see 7.1.3). indeed, any distinction between usucapia and truditio in this respect is difficult to justify. However, the opposite view prevailed in the late clas­sical period, and was accepted by Justinian—mistaken lusta causa was held insuf­ficient in usucapion, as a general rule.

7.2.1.6 Property capable of usucapion

The basic rule was that usucapia applied only to property capable of being owned pri­vately. Moreover, various things were excluded from usucapion on policy grounds, e.g. dotal immovables, and land belonging to a ward in guardianship. But, the most important exclusion in practice was stolen property. Under the Twelve Tables, and a lex Atiuia of the second century BC, stolen property could not be acquired through usucapia (see Inst.Gai.2.45.) (see, on this point, Pugsley, I)., ‘ The Misinterpretation of the Lex Atinia' (1970) 17 RIDA 3, 259-72. See also, recently, Belovsky, l’„ ‘Usucapio of Stolen Things and Slave Children' (2002) 49 RIDA 3, 57-99.) A similar rule was enacted in the late Republic in the case of property seized by force. The exclusion of stolen property had serious consequences for innocent acquirers of such property, e.g. bona fide purchasers for value. They could not become owners, however long their possession. The operation oi this rule can be seen as an over-protection of the position of the domimis, and as contributing to insecurity of title for inno­cent acquirers. There seems an imbalance here between protecting the position of the rightful owner and that of the innocent acquirer. See Pugsley, Property mid Obligations, 48 ff. However, the apparent imbalance was, according to some schol­ars, partially remedied by the qualification (probably introduced by the lex Atinia) that stolen property could be the subject of usucapion (according to some) if the owner failed to claim it, having discovered its whereabouts.

There are some difficult questions concerning the usucapion of stolen property. 1'he precise relationship of the lex Atinia to the Twelve Tables is problematic. It is unlikely that the lex Atinia was simply a repetition of the Twelve Tables rule that prohibited usucapion of stolen property. Most probably the original rule was seen as too protective of the position of the domimis, and so the lex Atinia allowed some exceptions. However, it is possible that the Twelve "Tables rule applied originally to usucapion by thieves, and that it was the lex Atinia which prevented the usu­capion of stolen property by anyone. But that view is inconsistent with a text by Gaius (Inst.Gai.2.49.) that suggests that the original rule in the Twelve Tables was aimed at preventing usucapio by a third party rather than by the thief, as the lat­ter could not usucapt in any case because of his lack of good faith. However, since the requirement of good faith is thought to have been a development of the later Republic, Gaius' clear implication that it was required at the time of the Twelve Tables is puzzling.

7.2.1.7 Longi temporis praescriptio

(C.7.33., 39.)

Usucapion of provincial land was not possible since such property was not capa­ble of private ownership. This factor eventually led to the development of longi temporis praescriptio—'lengthy prescription'—which emerged in the provinces dur­ing the Empire (possibly at the end of the second century AD) as a form of prescrip­tive interest in provincial land. It initially started off merely as a defence against the owner's vindicatio, but eventually, imperial enactments—commencing with a rescript in AD 199—recognized provincial practice and gave full legal protection to long possession of provincial land (see Nbrr, D„ ' lime and the Acquisition of Ownership in the Law of the Roman Empire' (1968) 3 IJ, 352-62, for a discussion of the reasons for the introduction of longi temporis praescriptio}.

The required period of possession was ten years if the parties resided in the same province; otherwise twenty years. Most of the rules of usucapio were applied. For example, the possessor had to have good faith and there had to be iustu causa. The main difference, apart from the length of the period required, was that a possessor could add his predecessor's period of possession to his own. ‘Lengthy prescription' did not originally confer ownership on the possessor but protected him against the owner, whose right of action was barred. However, the possessor's position was con­sidered to be so strong that in the late Empire he came to be regarded as acquiring ownership, by which time the rules of usucapion and ‘lengthy prescription' were virtually indistinguishable. The two systems were blended together under Justinian with the ending of the distinction between Italic and provincial land. Usucapio was confined to movables; land, wherever situated, was made subject to the rules of longi temporis praescriptio.

In the fourth century AD, an imperial decree (possibly of Constantine) provided that possession by a non-owner for forty years extinguished the owner's right to a vindicatio in whatever circumstances the possession had originated (even if in bad faith). This law, longissimi temporis praescriptio (‘very lengthy prescription'), was later amended by a reduction of the period to thirty years under Theodosius II in the fifth century AD (see C.Th.4.14.1.). Justinian ruled that if the possession had begun in good faith, the possessor acquired full title on the lapse of the period, even if the possession was otherwise defective, e.g. for lack of iusta causa.

7.2.2 Protecting those en route to acquiring ownership through usucapio

(D.6.2.)

Two categories of possessor could acquire ownership through usucapio, namely the bonitary owner and the possessor in good faith (see Hausmaninger, Gamauf, and Sheets, Casebook, 223-31). The bonitary owner lacked dominium since the par­ties had employed (possibly inadvertently) an inappropriate mode of conveyance while attempting to transfer ownership of an object (e.g. delivery of a res mancipi), while the possessor in good faith lacked dominium because the person who trans­ferred ownership to him using an appropriate mode of conveyance did not have ownership himself. In both instances, dominium did not pass and the transferee merely acquired interdictal possession which could ripen into dominium through usucapio. The interval between conveyance and the acquisition of dominium was the period when the transferee was described either as a bonitary owner or a possessor in good faith depending on the nature of the defect (see Inst.Gai.2.40.). During that time, the position of both the bonitary owner and the possessor in good faith was rather precarious under the ius civile: they could defend their possession using interdicts, but only against immediate dispossessors and they could not prevent the rightful dominus from recovering it through vindicatio. The praetors were concerned with the potential unfairness inherent in the ius civile position. So they protected the bonitary owner and possessor in good faith by remedies that enabled them to assert their title and defeat claims. The level of protection depended on the position of the transferee. If the transferee was a bonitary owner, these remedies protected him from claims by any third party, even the domimis. If the transferee was the pos­sessor in good faith, on the other hand, these remedies protected him only from claims by any third party with a weaker right than his own, in other words, against anyone except the dominus.

The operation of the rules depended on whether the bonitary owner or possessor in good faith was a defendant or a plaintiff. If the bonitary owner was a defendant, i.e. in possession of the disputed property, his interest would be protected against the domimis (the only party who could legitimately institute the vindicatio against him) by the praetor, who would allow a defence texceptio) in the relevant formula. The exceptio would state the reason why the defendant had an interest in the prop­erty, e.g. because it had been sold to him (the usual case). If the defence was proved on the facts, the domimis was defeated—the bonitary owner retained possession and could be awarded ownership. If the plaintiff was not the domimis, but a third party, the bonitary owner could rely on possessory interdicts (see 6.5.5.1) to protect his interest. If the possessor in good faith was a defendant, i.e. in possession of the disputed property, he could protect his possession against third parties who had a weaker claim than him to the property using the possessory interdicts. As the domi­mis had a stronger right to the property than the possessor in good faith, he could successfully bring a vindicatio.

What if the bonitary owner or possessor in good faith was the plaintiff, i.e. seek­ing recovery of the property? Possessory interdicts might be available, but their applicability was somewhat limited when they were sought for the recovery of pos­session; they normally failed against the domimis. In 67 BC, the praetors introduced the actio Publiciana, which gave the bonitary owner an effective remedy as plaintiff against the domimis (and any third party) (see D.6.2. Ipr.). The bonitary owner could allege in his formula a fiction—that the appropriate period of usucapio had elapsed (see Capogrossi Colognesi, 'Ownership and Power in Roman Law', 533). Providing that he could satisfy the other conditions of usucapio (e.g. that he had acquired possession in good faith), the bonitary owner's claim to the property would suc­ceed (see Inst. Gai.4.36.). If the dominus relied on the defence that he was the right­ful owner, the bonitary owner could raise a counter-defence based on the transfer (e.g. through sale) of the property to him by the domimis. The actio Publiciana was also available to the possessor in good faith, but the level of protection was more limited. Where the possessor in good faith was the plaintiff and the dominus the defendant in possession, he could not succeed with the actio Publiciana. Where the defendant was a third party (not the dominus), both the bonitary owner and posses­sor in good faith would succeed with their claims. See Nicholas, Introduction, 125-8.

Controversy surrounds the actio Publiciana. It does not appear to have been mentioned in juristic literature until the end of the first century AD—long after its apparent introduction. It may be that the date of the edict that introduced the action was considerably later than 67 BC—a Publicius was known to have been prae­tor in AD 93. But, the later date was surely too late for a major praetorian reform, whereas 67 BC falls very much within the period of the most rapid development of the ius honorarium.

There is uncertainty, too, about the scope of the actio Publiciana. Whether it was originally intended to apply to bonitary owners is not clear. It may be that the action was introduced primarily as a remedy for bona fide possessors (compare 6.5.5.1) but came to be applied for the benefit of bonitary owners as well. If this supposition is correct, the actio Publiciana may not have been as major a reform as has been traditionally thought. That would perhaps explain why, if it had been introduced in 67 BC, there was a lack of juristic comment on it for so long. An alternative view is that the actio Publiciana was originally introduced for bonitary owners, and later adapted for bona fide possessors. For the controversies about the actio Publiciana see, e.g., Diosdi, Ownership, 154 ff. and Pugsley, Property and Obligations, 51 ff.

Bonitary ownership eventually was regarded as being as beneficial as dominium. Of course, once the bonitary owner had held the property for the required length of time, he became the actual dominus.

Let us now consider the other modes. The following are commonly classified as original iusgentium modes.

7.2.3 Occupatio

(Inst.Gai.2.66.-9.)

Occupatio is the law of 'first-taking': the first taker of ownerless property (res nullius) becomes its owner (see Ruddy, F. S., 'Res Nullius and Occupation in Roman and International Law' (1968) 36 University of Missouri LR, 274-87). Property could be ownerless because it had never been owned before, or because the previous owner had intentionally ceased to be the owner (see Hausmaninger, Gamauf, and Sheets, Casebook, 150-62). The principle may seem, prima facie, somewhat crude and unsub- tle, but in practice it constitutes a sensible and relatively uncomplicated method of resolving disputed ownership of such property. Indeed, 'the first in time prevails' is a principle that is well-known to modern lawyers. It reflects that basic human response—T got there first'. The most important application of occupatio was in rela­tion to things that had never had owners, enemy property, and abandoned things.

7.2.3.1 Things which had never had owners

This category included a variety of objects, which had not previously been owned and were therefore classified as res nullius. In Roman law, the capture of wild animals attracted juristic attention since it was an important application of the rules of occu­patio, but this category of ownerless things also included, e.g., islands arising in the sea and gems found on the seashore. 'Wild animals' included birds, bees, and fish:

Gaius, Common Matters or Golden Things, book 2: So all animals taken on land, sea, or in the air, that is wild beasts, birds, and fish, become the property of those who take them... (D.41-1.1.1.)

Animals were classed as wild or tame by nature (ferae or mansuetae naturae), the test being whether the animal belonged to a species designated as wild or tame, not whether it was wild or tame in its habits:

Gaius, Common Matters ar Gohlen Things, book 2:... there are those who have tame deer which go into and come back from the woods but whose wild nature has never been denied. (D.41.1.5,5.)

The apparent simplicity of this classification of animals was deceptive. There was some difficulty in assigning animals to one class or another, mainly because of the existence of wild varieties of some domestic animals, or of hybrid, semi-wild species. See McLeod, G., 'Wild and Tame Animals and Birds in Roman Law', in New Perspectives, 169-76. Nevertheless, the classification was not without merit, as is evidenced by its continued survival in modern legal systems.

Did it matter whether animals were classed as wild or tame? Certainly—the rules as to ownership differed, animals tame by nature being regarded as ordinary mova­bles. But to acquire ownership of a wild animal it was necessary to take effective control of it—wounding an animal, without capturing it, was probably insufficient. It was irrelevant to the question of ownership that the capture occurred on some­one else's land, although intentional trespassing could constitute insulting behav­iour (see generally 10.4). Ownership was retained for as long as the taker remained in effective control of the animal:

Gaius, Common Matters or Golden Things, book 2: Any of these things which we take, however, are regarded as ours for so long as they are governed by our control. But when they escape from our custody and return to their natural state of freedom, they cease to be ours and are again open to the first taker, (Florentinus, Institutes, book 6] [4]: other than those tamed creatures which are in the habit of going and returning. (D.41.1.3.2., 4.)

Animals with the 'habit of returning' (animus revertendi) ceased to be owned only when they abandoned the habit, i.e. lost their instinct to return. Animals without the habit of returning ceased to be owned when they escaped from custody:

Gaius, Common Matters or Gohlen Things, book 2: An animal is deemed to regain its natural state of liberty when it escapes our sight or, though still visible, is difficult of pursuit. (D.41.1.5pr.) Since the owner of a wild animal ceased to be the owner when it escaped from his control, it followed that he was not responsible for any damage caused thereafter. This unsatisfactory position was remedied to some extent by an edict of the aediles that imposed strict liability (on those who kept wild animals near a highway) for damage caused by an escaped beast (see 10.6.2).

Bees were singled out for separate attention by the jurists, although the rules applicable were similar to those stated above. To obtain effective control of bees it was necessary to hive them:

Gaius, Common Matters or Golden Things, book 2: Bees, again, are wild by nature and so those which swarm in our tree are, until housed by us in our hives, no more regarded as ours than birds which make a nest in our tree. Hence, if another should house or hive them, he will be their owner.... [4] A swarm which flies away from our hive is deemed still to be ours so long as we have it in sight and its recovery is not difficult; otherwise it is open to the first taker. (D.41.1.5.2., 4.)

The rules on bees, as practical as any in Roman law, have found a modern applica­tion, e.g. Kearry vPattinson [1939] 1 KB 471,

7.2.3.2 Enemy property

Celsus, Digest, book 2: And property of the enemy, which is on our territory, becomes not public property, but that of the first taker, (D.41.1.51,1.)

This rule applied to enemy property captured within Roman territory in time of war, and probably to things belonging to nationals of States with which Rome did not have a treaty of friendship. However, booty captured in military action did not fall within the rule; it was usually sold or given away by the victorious general (although in theory it belonged to the State).

7.2.3.3 Abandoned property

The first taker of abandoned property became its owner providing that he had the intention of acquiring it. What was the test of abandonment? It seems that the owner must have intended to be rid of the property, no longer caring about its desti­nation before it became res millius:

Gaius, Common Matters or Golden Things, book 2: It is another matter with those things which are jettisoned in stress of seas to lighten the vessel; they remain the property of their own­ers; for they are not cast overboard because the owner no longer wants them, but that the ship may have a better chance of riding the storm. Consequently, if anyone finds such things washed up by the waves or, for that matter, in the sea itself and appropriates them with a view to gain, he is guilty of theft. (D.41.1,9.8.)

. The Sabintans and Proculians disputed the question whether the previous owner lost ownership at the moment of abandonment, as the Sabinians argued, or when another person took possession of the abandoned property (the Proculian view). The Proculians were unhappy at the possibility that important property could be ^! ownerless for any length of time. However, the Sabinian view, more convenient in ; practice, seems to have prevailed.

Occupatio of abandoned res mancipi received surprisingly little attention from the jurists in view of the potential problems involved. For example, could the first taker of abandoned res mancipi acquire dominium over the property through occupatio? Certainly, if the property was then held for the relevant period of time required for prescription under usucapio. But whether dominium could arise before.the lapse of the relevant period is unclear. The view that dominium over res man­cipi could be acquired only through ius civile modes of acquisition (and therefore not through occupatio) is problematic since, if that were the case, abandoned res mancipi might not fall into ownership again—a possibility which was not likely to have been encouraged in early law. In any case, there were ways in which dominium over res mancipi could be acquired by ius gentium modes of acquisition, e.g. avulsio (see 7.2.4.1). The most plausible view is that occupatio of abandoned res mancipi sufficed to give dominium immediately without the need for a period of prescription.

7.2.4 Accessio

(Inst.Gai.2.73.-8.)

Accessio was the inseparable attachment of things belonging to different owners, one thing being incorporated in the other. The incorporated property was the acces­sory thing: it acceded to the principal thing. The general rule was that the owner of the principal thing became owner of the whole thing. Thus, if A built a house on B's land, the house acceded to the land since the latter was considered to be the principal thing. B owned the house (cf. D.41.1.26pr.).

Accessio has to be distinguished from commixtio and confusio (see Hausmaninger, Gamauf, and Sheets, Casebook, 194-8). A commixtio occurred when things belong­ing to different owners were mixed, but were readily separable. If this was done by agreement, the resulting mix became common property; if done without agree­ment, the things could be separated and returned to their respective owners—there was no change of ownership. In the latter case, the appropriate remedies were the actio ad exhibendum ('the action to display') for the production of the thing with a view to separation, and a vindicatio, where necessary. Confusio occurred if the things that were mixed were not readily separable and it was not possible to tell which was the principal thing, which the accessory, e.g. when A mixes his wine with B's wine. The resulting mix was owned in common in proportion to the value of the parties' respective shares, unless the parties had agreed otherwise. The rules of confusio were applied in Indian Oil Corporation Ltd v Greenstone Shipping SA (Panama) [1988] QB 345 (see 11.3.3).

As regards accessio, the owner of the principal thing became the owner of the whole thing, irrespective of the bona fides of the parties or the identity of the per­son who did the mixing or attaching. Accessio applied only to things that were not readily separable. The issue of separability was partly a question of law—buildings, for example, were considered to be legally inseparable from the land—and partly a question of fact. Things were considered not to be readily separable if undue effort, skill, or cost was required to separate them, or if the separation would damage the property. The operation of the rules of accessio is best considered by distinguishing between land and movables.

7.2.4.1 Accessio affecting land

The basic rule was that anything that was attached to land became part of it, since land was regarded as the principal thing (see Hausmaninger, Gamauf, and Sheets, Casebook, 176-85). The application of the rule occurred mainly in the following cases:

(a) Buildings They were regarded as inseparable from the land on which they stood—hence, the landowner owned anything built on his land (the principle superficies solo cedit is expressed in Inst.Gai.2.73.). However, he did not necessar­ily own the materials comprised in the building: the law recognized a distinction between ownership of a building and ownership of the materials:

Gaius, Common Matters or Golden Things, book 2: When someone builds on his own site with another's materials, he is deemed to be owner of the building because all that is built on it becomes part of the soil. However, 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 by reason of the Law of the Twelve Tables which provides that no one is required to give up materials of another built into his premises but that he must pay double their value.... Hence, 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. (D.41.1.7.10.)

In the situation discussed in the above text, i.e. where B builds on his land with A’s materials, the legal position depended on whether there had been theft of A's materials. In the absence of theft, it seems (the texts are unclear) that A had no rem­edy apart from claiming the materials when the building came down. This was a dormant right. If B had stolen the materials he was liable for theft under the Twelve Tables for double their value. It would appear that the reference to 'double value' in the above text assumes that the builder had stolen the materials. This action was barred, however, if A chose to sue B for theft under the actio furti (see 10.3.2.2). Whichever action A chose, he retained his right to reclaim the materials when the building came down. If the materials had been stolen, but not by B, A could sue the thief and retained the right of eventual reclaim of the materials.

What if a person built with his own materials on another's land? The position depended on whether the builder had acted in good faith or not. If he had not, i.e. he realized that the land belonged to someone else, he was deemed to have made a gift of the materials to the landowner:

Gaius, Common Matters or Golden Things, book 2:... if a person were to build with his own materials on someone else's site, he would make the building the property of the owner of the site, and if he knew that the site belonged to another, he would be treated as voluntarily parting with his materials so that even if the house should collapse, he would have no v/ndi- cotiofor them.... (D.41.1.7.12.)

If the builder had acted in good faith, the position depended on whether the land­owner had later gained possession of the building; if so, then the builder normally had no remedy, apart from eventual reclaim of the materials. If the builder was still in possession, he was entitled to recover his expenditure. However, underJustinian the ius tollendi ('the right to remove') was allowed to the builder in possession, irre­spective it seems of good or bad faith on his part. This enabled the builder, when vacating the premises, to remove as much of his materials as he could without damaging the structure of the building. Some uncertainty surrounds the ius tol­lendi. It seems that it was known in classical law, but was available at that time only to the builder who had acted in good faith. The right probably did not apply where the owner was prepared to compensate the builder for not removing materials:

Celsus, Digest, book 3: Our decision is that if the owner is prepared to pay the possessor as much as he would have if he took the materials away, he should have the power to do so. There must be no indulgence to malice. If, say, you want to scrape off plaster which you have put on walls, and deface pictures, that will serve no purpose but to annoy.... (D.6.1.38.)

(b) Planting and sowing The general rule was that if anything belonging to A was planted or sown in B's land, the latter became the owner when the thing took root: Gaius, Common Matters or Golden Things, book 2: If I plant someone else's cutting in my land, it will be mine; conversely, if I plant my own cutting in someone else's land, it will be his, provided, in each case, that it roots itself. For until it takes root, it remains the property of its former owner. (0.41.1.7.13.)

Regarding compensation, if A sowed (in his own land) seeds belonging to B, A was liable for theft if he had acted in bad faith; if in good faith, A was probably liable to compensate B for the value that the seeds or plants had when they were appro­priated. Where A sowed his own seeds in B's land, he was deemed to have made a gift of them if he had acted in bad faith. If he had acted in good faith, and was in possession of the land, A was entitled to the value of the seeds; if out of possession, it is doubtful whether he had a remedy.

(c) Avulsio This occurred where an identifiable piece of land was carried by the force of a river current and was deposited against the land of another:

Gaius, Common Matters or Gohlen Things, book 2: But 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. (D.41.1.7.2.)

(d) Alluvio This was the imperceptible accretion or deposit of soil on a person's land through the action of a river. Such deposits enlarged the land by acceding to it: Gaius, Common Matters or Gohlen Things, book 2: Furthermore, what the river adds to our land by alluvion becomes ours by the law of nations. Addition by alluvion is that which is gradually added so that we cannot, at any given time, discern what is added. (D.41.1.7.1.)

It seems that alluvio was confined in Roman law to the action of rivers. The principles of alluvio have been adopted by modern systems, sometimes in modi­fied form, e.g. in English law alluvio applies to ponds, lakes, and the action of the sea. See Jackson, P., 'Alluvio and the Common Law’ (1983) 99 LQR, 412-31. The French Civil Code, on the other hand, restricts alluvio along Roman lines (Articles 556-8). Although alluvio has traditionally been thought to be con­cerned essentially with silting, it has been argued that its scope was wider than that: see Lewis, A. D. E., 'Alluvio: The Meaning of Institutes 11.1.20’, in Studies J. A. C. Thomas, 87-95.

(e) Rivers Rivers are capricious things—they flood, they dry up, they change course. Their action can pose a variety of legal problems apart from avulsio and alluvio. For example, who owns an island arising in a river? The rule was that the island was owned by the riparian owners up to the middle line of the river unless it lay wholly to one side of a riverbank, in which case it acceded to that bank:

Gaius, Common Matters or Gohlen Things, book 2: An island arising in a river (a frequent occur­rence), if indeed it appears in the midstream of the river, is the common property of those who have holdings on either bank of the river to the extent that those holdings follow the bank; but if it lies to one side of the river rather than the other, it belongs only to those who have holdings on that bank. (D.41.1.7.3.)

If a river changed course or dried up, the old riverbed was shared by the ripar­ian owners up to the middle line of the bed. If part of a riparian owner's land became an island because the river divided itself, there was no change of own­ership. What if the river flooded land? If the flood abated, there was no change of ownership; but if the waters did not recede, it seems that the ground covered by the river was partly lost to its previous owner—in effect it became a river bed, the rule of the imaginary middle line applying to the new situation. There was no question of compensation, since such changes of ownership were the result of natural phenomena, not human acts. The same was true of avulsio and alluvio.

7.2.4.2 Accessio affecting movables

Here we are concerned with the case where movables (principal and accessory) were inseparably attached or mixed together without the consent of the owners. Roman law acknowledged a number of examples of accessio affecting movables (see Hausmaninger, Gamauf, and Sheets, Casebook, 186-93). As a starting point, confiisio (blending of liquids such as wine, honey, or molten metals belonging to different owners) and commixtio (blending of solids such as grain or coins belonging to dif­ferent owners) were excluded (see 7.2.4), but specific cases such as textlira (insepa­rable weaving of costly thread owned by one party into the garment belonging to another), ferruminutio et pliimbatura (welding or soldering an object belonging to one party to a principal object belonging to another), scriptlira (writing on or decorating parchment belonging to another with one's own materials), and pictura (painting with one’s paint on the canvas belonging to another) were identified. See Watkin, T. G„ 'Tabula Picta: Imagesand Icons' (1984) 50SDHI, 383-99, on the later history of this rule and its application in the Eastern Empire following the advent of Christianity. The resulting property generally belonged to the owner of the prin­cipal thing. The major issues in accessio of movables were (a) the test for deciding which thing was the principal and which the accessory, and (b) whether compensa­tion was payable to the owner of the accessory thing.

(a) Principal or accessory? In cases involving land there was little problem in iden­tifying the principal thing—things acceded to the land. However, as regards the accessio of movables to movables, there could be difficulties in deciding which the principal thing was. For example, what was the position where 1 painted on my canvas with your paints, or wrote a poem on your parchment? Not surprisingly, the jurists did not achieve consistency about what the appropriate test should be in such cases. For example, the following passage suggests alternative tests, based on relative size or value:

Pomponius, Sabimis, book 30: If it be asked which cedes to the other, when elements belong­ing to each of two owners are welded together, Cassius says that an assessment is to be made of the respective portions of the final product or of the value of each element. (D.41.1.27.2.)

These tests find a place in the French Civil Code, which regards the thing ‘largest in value' as the principal thing, or the thing that is largest in 'volume' if the values are about equal (Article 569). Another test, proposed in some cases, was that of independence—that thing was the principal thing that could have an existence independently of the other thing:

Paul, Edict, book 21: Whatever is written on my paper or painted on my board at once becomes mine. Although in the case of a painting some writers have held the opposite, on account of a painting's value, yet where one thing cannot exist without the other, it necessarily accedes to that other. (D.6.1.23.3.)

But the test which was perhaps applied most widely focused on the physical identity of the constituent elements: that thing was principal which most retained its iden­tity, i.e. which gave the resulting mix or attachment its essence, name, or overall character:

Paul, Sabinas, book 14:... we have to look at the overall character, so that if something be added, it becomes part of the whole. Thus, if a foot or a hand be added to a statue, a base or handle to a goblet, a post to a couch, a plank to a vessel, stone to a building, the whole belongs to the erstwhile owner of the statue, and so forth. (D.41.1.26pr.)

Consistency in deciding which thing was principal and which accessory is not to be found in the texts. Cases tended to be decided on a casuistic basis rather than by the widespread application of any one test. Paintings were an instructive case: in classical law, if A painted a picture on B's tablet, the predominant view (despite Paul's assertion in D.6.1,23.3.) was that the tablet acceded to the painting, i.e. A, the painter, owned the painting:

Gaius, Common Matters or Golden Things, book 2: Pictures do not accede to the tablets on which they are painted.... On the contrary, the view established itself that the tablet accedes to the picture. (D.41.1.9.2.)

Such a solution could be justified by the application of the identity test—it is the picture which gives the essence or overall character to the resulting thing, not the tablet or canvas. Justinian adopted the predominant classical view, but for a differ­ent reason—that of relative value. This test works well (and fairly) if there is a clear difference in the relative value of the two things, as there would be if Rembrandt had painted a portrait on your canvas. But what if you had painted on his canvas?

As regards writing, it was held to accede to parchment even if the former was more valuable, e.g. if it was done in gold leaf:

Gaius, Common Matters or Golden Things, book 2: Letters, even in gold, accede to the paper or parchment just as things built or sown become part of the land. Thus, if I wrote verse or a story or a speech on your paper or parchment, not I but you would be held to own the fin­ished work. (D.41.1.9.1.)

The distinction made between paintings and written script is problematic. The identity test comes closest to providing a satisfactory justification for the distinc­tion but, even with that test, there are problems—would not the overall character of parchment that contained illuminated script be determined more by the script than by the parchment? On the other hand, in the case of 'ordinary' writing on parchment, the identity test makes sense. The words, unlike a painting, are unlikely to dominate in a visual sense: they are meant to be read rather than viewed.

(b) Compensation Was the owner of the accessory thing entitled to compensation once the inseparable attachment had occurred? The general position depended on whether the attachment had been effected by the owner of the principal thing or by the owner of the accessory thing. If the former had attached the things and had acted in bad faith, he was liable for theft but was nevertheless the owner. If he had acted in good faith (e.g. thinking the accessory thing belonged to him) and was in possession of the property, it seems that he was not liable to compensate the other party, until the later Empire when an action was allowed for the value of the accessory thing. If the owner of the principal thing was not in possession, he could obtain the property by vindicatio on paying for the value of the accessory thing. If he refused to pay, his vindicatio would be defeated by the other party's exceptio doli ('the defence of fraud’) on the ground that it would be 'fraudulent' for the former to obtain the property without paying compensation.

If the owner of the accessory thing was responsible for the attachment and had acted with full knowledge of the facts, he was deemed to have made a gift of the thing. Where he had acted in good faith, he had no remedy if he was out of posses­sion; if in possession, he was entitled to compensation.

These rules were of general application. It is possible that the question of owner­ship and compensation was decided differently in the case of paintings. A dubious text (in D.41.1.9.2.) suggests that the owner of the tablet was entitled to the paint­ing if he paid 'the cost of the painting'.

7.2.5 Specificatio

(Inst.Gai.2.79.)

Specificatio was a method of acquiring ownership by the creation of a new thing out of someone else's materials (see Hausmaninger, Gamauf, and Sheets, Casebook, 201-4). For example, if A made wine out of B's grapes, A owned the wine. It appears that the term specificatio was post-classical although the rules were formulated by the classical period. Specificatio had an obvious affinity with accessio, and it was not always certain which was the appropriate mode of acquisition in a given situation. (On specificatio, see Cohen, B., 'Specificatio in Jewish and Roman Law' (1958) 5 RIDA 3, 225-90 for a comparative perspective and Plescia, J., 'The Case of Specification in Roman Law' (1973) 24 IVRA, 214-21, for a thorough overview of the basic legal rules relating to specificatio. Stoop, B. C., 'Non Solet Locatio Dominium Mutare: Some Remarks on Specificatio in Classical Roman Law' (1998) 66 TR, 3-24 attempts to establish at which point in classical Roman law artisans, craftsmen, and other workers were excluded from the scope of specificatio). See, more recently, Plisecka, A., ‘Accessio and Specificatio Reconsidered' (2006) 74 TR, 45-60, as well as the obser­vations by Watkin, T. G., 'Concrete Cases which could Occur: The Problem of Specificatio in Roman law', in Melanges Gandolfi, 35-366.

7.2.5.1 The Proculian-Sabinian dispute

There was a celebrated dispute between the schools about specificatio:

Gaius, Common Matters or Golden Things, book 2: When someone makes something for himself out of another's materials, Nerva and Proculus are of opinion that the maker owns that thing because what has just been made previously belonged to no one. Sabinus and Cassius, on the other hand, take the view that natural reason requires that the owner of the materials should be owner of what is made from them, since a thing cannot exist without that of which it is made.... (D.41.1.7.7.)

The Proculian view in effect treated specificatio as a special form of occupatio: at the moment of creation the new thing was regarded as a res nullius, i.e. owned by no one, and thus open to the first 'occupier'—the creator. The analogy with occupatio was rather strained, however: the created thing can hardly be said to have been a res nullius if it was owned as soon as it was created. There were other views on the question of ownership of newly created things:

Gaius, Common Matters or Golden Things, book 2: There is, however, the intermediate view of those who correctly hold that if the thing can be returned to its original components, the bet­ter view is that propounded by Sabinus and Cassius but that if it cannot be so reconstituted, Nerva and Proculus are sounder. Thus, a finished vase can be again reduced to a simple mass of gold, silver or copper; but wine, oil or flour cannot again become qrapes, olives or ears of corn... (D.41.1.7.7.)

The 'intermediate' view, that the new thing belonged to the creator if it was not reducible to its original form, was the solution adopted by Justinian. Where the thing was reducible, the owner of the original materials owned the new thing; but it was not necessary that it should actually be reduced—what gave him ownership was that the thing was potentially reducible. What was the test of reducibility? The outcome depended on how feasible it was for the thing to be reduced: it would not be feasibly reducible if undue effort or skill were required, or if the substance of the thing was substantially impaired, e.g. if the process of reduction left little or nothing of the original material. If the material was owned partly by the creator, there was broad agreement that the creator owned the new thing—reducibility was irrelevant. If, however, the materials were owned by two or more persons, but not the creator, and the new thing was reducible, the owners of the materials owned the new thing, on the view adopted by Justinian, the principles of confusio and com­mixtio applying to determine their respective rights (see 7.2.4).

7.2.5.2 Nova species

The application of specificatio depended on whether a new thing (nova species) had been created. What was the test of nova species? The texts are inconclusive, but it seems that the new thing had to have an identity, a name of its own: it could not be simply an altered or improved thing.

Gaius states:

Gaius, Common Matters or Golden Things, book 2: In my view, however, there are those who rightly say that corn threshed from someone's ears of corn remains the property of the owner of the ears; for since the corn already has its perfect form while in the ears, the thresher does not make something new, but merely uncovers what already exists. (D.41.1.7.7.)

Thus, one who crushed grapes did not create a new thing: he simply brought about crushed grapes. But making wine from grapes, or a ship from planks of timber would be regarded as creating a new thing. It seems that some element of skill and effort had to be involved in the creation of a nova species. (See Van der Merwe, C. G., ‘Nova Species' (2004) 6 Roman Legal Tradition, 96-115, on the history and contemporary application of this rule in Scots law.)

7.2.5.3 Compensation

Ownership could be acquired by the creator even if he had not acted in good faith. The existence of bad faith was relevant only to the question of compensation, the rules being similar to those in accessio. If the creator acquired ownership, having acted in good faith, and was in possession of the new thing (the likely scenario), it seems that the owner of the materials had no remedy. If the creator was not in possession, he was entitled to the property, but only if he paid the other party for the value of the materials. If the creator had acted in bad faith, he could be sued for theft but was entitled to the new thing.

Where the creator did not become the owner through specificatio, but had acted in good faith and was in possession, he was probably entitled to compensation for his work when the owner brought a vindicatio. If he was out of possession, it is unlikely that he had any remedy. If the creator had acted in bad faith, he was deemed to have made a gift of his work.

The French Civil Code (Articles 570-1) has adopted the Roman rules on specifi­catio, but with some important differences. The owner of the materials can claim the new thing if he reimburses the price of labour, unless the value of the labour far surpasses the value of the materials used. In the latter case, the creator retains the thing but must reimburse the other party for the cost of the materials. The labour and the materials are valued as at the date of reimbursement.

22.6