I. FUNDAMENTAL DISTINCTIONS

Property and obligations—actions and rights in rem and in personam. A man’s assets are either property or obligations. The difference between the two is the difference between owning and being owed something.

This difference between owning and being owed is expressed

¹              For Gaius, and therefore for Justinian, the sequence of thought which leads to the treatment of succession immediately after property is this: in both alike we are concerned with the methods of acquisition of things, the subject-matter of the law of property being the acquisition of single (i.e.

by the Roman lawyer in the distinction between actions in rem and actions in personam. Any claim is either in rem or in personam, and there is an unbridgeable division between them. An action in rem asserts a relationship between a person and a thing, an action in personam a relationship between persons. Thus the typical action in rem (rei vindicatio} asserts that a physical thing belongs to the plaintiff, and the simplest action in personam (condictio} asserts that the defendant owes a sum of money or a physical thing to the plaintiff. The Romans think in terms of actions not of rights, but in substance one action asserts a right over a thing, the other a right against a person, and hence comes the modern dichotomy between rights in rem and rights in personam. Obviously there cannot be a dispute between a person and a thing, and therefore even in an action in rem there must be a defendant, but he is there not because he is alleged to be under any duty to the plaintiff but because by some act he is denying the alleged right of the plaintiff.

What is the practical importance of the distinction between actions in rem and 'actions in personam? One must begin by emphasizing that an action in rem is not, in form, one which compels the defendant to return the res. (This is the sense in which the Common law uses the term ‘real action’.) Since the Roman process was essentially a voluntary recourse to arbitra­tion, and since there was no machinery for the state enforcement of judgments, the result of a successful action, whether in rem or in personam, was simply an order to the defendant to pay to the plaintiff a sum of money, and it was for the plaintiff to enforce this order, if need be, by levying execution on the defendant’s person or property. In such a system there is no place for decrees of specific restitution or specific performance. And yet, having said so much, one must then add that in most actions in rem the defendant would in fact be driven to restore the object claimed. This contrast between the formal and the substantial result of an action was achieved by a simple device.

¹ The complete formula of the vindicatio ran as follows: ‘If it appears that the thing in question belongs to the plaintiff at civil law (ex iure Quiritium), then, unless at the direction of the judge the defendant restores the thing, let the judge condemn the defendant to pay the value of it to the plaintiff. If it does not so appear, let the judge absolve the defendant.’

value of the thing.

The answer to our question¹ can perhaps best be found by contrasting two constructions which can be put upon a sale. If A agrees to buy B’s book for a certain price, Roman law gives A an immediate right in personam to the delivery of the book (and B a concurrent right to the payment of the price), but gives A no right in rem until the book is actually delivered. Some other systems, however, give A both the right in personam and the right in rem as soon as the agreement is made.² The difference between these two constructions lies in their effect on third parties. There are two typical situations.

(i)   B still has the book and merely refuses to deliver it. Here, leaving aside the question considered on the previous page, it will usually make no difference whether A proceeds in rem or in personam. In either case he will obtain the money value of the thing. But B may be insolvent, and then the distinction will be crucial. For if A’s claim is only in personam it will take its place with the claims of all the other creditors against B’s inadequate assets, whereas if it is in rem A will be able, as it were, to take the book out of B’s bankruptcy and thus to satisfy his claim in full. For he is then not a creditor but an owner.

(ii)  The book has passed out of the hands of B into those of C—B has, for example, persuaded A to pay the price in advance and has then sold and delivered the book to G (this, or some variant of it, is one of the commonest frauds), or C has stolen it or acquired it from a thief.

¹              The Roman lawyer’s own first answer might have been that the practical importance lay in the ‘mesne process’ (cf. Maitland, The Forms of Action at Common Law, pp. 76 f.), that is to say in the steps which the plaintiff took if the defendant failed to defend the action. In an action in rem he was authorized simply to take the thing, since it was against this that his claim lay, whereas in an action in per­sonam he proceeded against the person or property of the defendant.

²              This is in principle the position of English law, but it is so qualified by special rules that the practical result is nearly always the same as in Roman law (see Buckland and McNair, Roman Law and Common Law, 2nd edn., pp. 291 if.).

B, while the other construction allows him to proceed against either B or C. Once again, of course, this difference will be of practical importance only if B is insolvent or has disappeared, but one eventuality or the other is not at all improbable. The man who sells to C in breach of his contract with A is not likely to await the bringing of an action, and even if he can be caught he is not likely to be still solvent. The problem which the law has to solve is the legal version of the eternal triangle. Which of two innocent parties is to suffer for the act of a dishonest third? There can be no solution which is both capable of practical application and ideally just. The law must choose the one which is most in the interest of the community at large.

Contract and conveyance. There is a further gulf set in Roman law between rights in rem and rights in personam. The ways in which they come into existence are kept distinct: the act which creates a right in personam does not create a right in rem, and vice versa. The most important application of this principle is in the dis­tinction between contract and conveyance. A contract creates rights in personam but cannot create or transfer rights in rem. For this a conveyance is necessary, i.e. an act recognized by law as appropriate for the purpose. We have just seen that in a sale of a book ownership does not pass to the buyer by the making of the contract, but only by the actual delivery of the book. The delivery is the conveyance. If the object sold were not a book but a res mancipi such as land or a slave, a formal conveyance by mancipatio or in iure cessio would be necessary.¹ And a convey­ance was likewise necessary for the creation or transfer of rights in rem other than ownership.²

This separation of contract and conveyance is rooted in the sound though inarticulate principle that rights in rem, since they potentially affect everyone, should not be secretly created or transferred, whereas there is no such objection to the secret creation of rights in personam. Hence it is that the Roman con­veyance is essentially a visible and public act. In iure cessio takes

* See below, pp. 105 f.

²The principle was not, however, maintained inviolate even in the classical law (e.g. the creation of hypotheca by a simple agreement—below, p. 152) and in the law of Justinian, when the formal conveyances have disappeared, it has been largely forgotten (cf. below, pp. 119 f.).

place in open court.¹ Mancipatio,² though in form a private act, requires the presence of six Roman citizens in addition to the parties, and the reason for so large a number must have been the desire for publicity rather than the need for subsequent proof, since, by contrast, the formal contract of stipulatio, which created only rights in personam, could be validly made without any witnesses at all. Similarly, the informal conveyance by delivery (traditw} involved in principle the visible transfer of the object itself.

There is however a conflict, which any system of law has to face, between the interest of the community at large in the publicity of conveyances, and the desire of individuals to con­duct their affairs in private. In Roman law the latter for the most part eventually prevailed. Both in iure cessio and mancipatio must in the classical law have been quite ineffective to ensure publicity. For in iure cessio was by then a very empty form, and the six participants in the mancipatio, though adequate enough for a small community, could constitute no hindrance to secrecy in so vast a society as imperial Rome. Even traditio, which increasingly replaced the formal conveyances, could eventually be performed without any visible transfer of the object, so that in the late Roman law, as in developed English law, a con­veyance was often, like a contract, nothing more than a docu­ment drawn up between the parties.³

This blurring of the line between contract and conveyance is perhaps inevitable in a sophisticated society, and modern systems usually adopt other methods to ensure publicity for the creation and transfer of rights in rem. The only wholly effective method is registration—to require that all such creations and transfers shall be entered on a public register. This method is increasingly used in the modern world, but can obviously in practice only be applied to land (or to some exceptionally identifiable movables such as ships or motor-cars), and even so it cannot be initiated without an accurate survey and a body of skilled administrators. It was quite unknown to Roman law.⁴ In regard to movables

¹ See above, p. 63.                                                            ² See above, ibid.

³              See below, pp. 119 f. There was a similar struggle to achieve secret conveyanc­ing in English law, culminating early in the seventeenth century in the recognition of the device of a bargain and sale for a term followed by a release.

⁴              There was something of the kind for land in Roman Egypt, but precisely what the effect of registration was is not clear.

there can be no simple solution. The Roman law, as we have seen, in principle required an actual delivery before ownership could pass. So long as this rule could be strictly maintained it was possible to say that no ownership could pass without the passing also of possession, but it could not also be said that no possession could pass without the passing also of ownership. This would have meant simply that the distinction between ownership and possession was obliterated. The thief would have owned what he had stolen. A less extreme rule might have excluded the thief by equating only the bona fide possessor (e.g. the man who innocently buys from the thief) with the owner, but this would have made a great inroad on the inviolability of ownership. And both the distinction between ownership and possession and the inviolable character of ownership were to the Roman lawyer fundamental.¹

Classifications of res. The Roman law classifies ‘things’ in many different ways for different purposes, but only three of these classifications need be considered here.

(i)  Res mobiles, res immobiles. Most, if not all, systems of law find it necessary to distinguish land, and the buildings which go with it, from all other property, both because of its intrinsic importance and because of the obvious fact that it is incapable of being moved. Roman law is no exception. The Twelve Tables laid down two years as the period for the prescriptive acquisition (usucapw)^{[38] [39]} of immovables, whereas one was sufficient for mov­ables ; and the distinction was relevant for some other purposes also. But it never acquired the central importance that, in the shape of the distinction between ‘real’ and ‘personal’ property,^[40] it has in English law.

(ii)  Res mancipi, res nec mancipi. A much more important dis­tinction in the original civil law was the peculiarly Roman one between res mancipi and res nec mancipi. Res mancipi were slaves, beasts of draught and burden (oxen, horses, asses, mules), Italic land,^[41] and rustic praedial servitudes (e.g. rights of way and of water) over such land. All other things were res nec mancipi. The practical importance of the distinction was that res mancipi could only be conveyed by mancipatio (hence the name) or in iure cessio; a mere delivery was ineffective to pass ownership. This was the rule of the civil law, but in the later Republic the Praetor gave to the recipient by delivery (the ‘bonitary owner’) almost all the substance of ownership,¹ and the distinction then took its place among the technical survivals which served to complicate the classical law. It survived in name until it was abolished by Justinian.

The origin of the distinction has been much debated.² Why should certain things be set apart as requiring a formal and public conveyance? If the question is framed in this way the obvious answer is that these must have been the most important things in early Rome, but it is not always easy to see why the particular things listed above, and no others, should have had this special importance. It may be, of course, that the original list was different, but at any rate in the later Republic the categories of res mancipi had become closed and arbitrary. For Gaius remarks that elephants and camels, though they were beasts of draught or burden, were nevertheless not res mancipi, because they were unknown when the list was settled.

(iii)  Res corporales, res incorporales. As we have seen,³ there is implicit in Gaius’ threefold division of the law into persons, things, and actions, a distinction between physical things, such as a table or a house, and abstract things, such as a debt or a right of way. Gaius expresses the distinction in the terms res corporales and res incorporales. It is practically important for only one reason. Incorporeal things cannot be possessed, since posses­sion requires essentially a physical holding, and they cannot therefore be acquired or transferred by any method which involves the transfer or acquisition of possession. In short, incorporeal things can neither be acquired by usucapio nor con­veyed by traditio.

of Roman ownership until it included the whole of Italy. Thereafter, illogically, the extension ceased, except that the land of some privileged communities was, by the grant of the ius italicum, treated as if it were Italian.

¹              See below, pp. 125 ff.

²              See the classical account in Maine, Ancient Law, ch. 8 and, for a discussion of more recent views, Jolowicz and Nicholas Histoneal Introduction, 3rd edn., pp. 137 ff

³              See above, p. 98.

The distinction was taken up by Justinian and has become part of the legal language of Europe and, to some extent, of the Common law. It is a convenient distinction, and it certainly corresponds to the way we commonly speak—a man will say that he has bought one plot of land, a right of way over a second, and an option on a third—but on a strict examination it is illogical. It identifies owner­ship with the object owned. The strictly comparable statement to T have bought a right of way over a plot of land’ is not T have bought a plot of land’, but T have bought the ownership of a plot of land’. In each case I have acquired a right, the right of ownership being merely more extensive than the right of way. The Roman, however, preferred to think in terms of assets rather than of rights, and saw the land and the right of way as equally assets. He felt no need to make a clear distinction between ownership and its object, since for him only corporeal things could be owned, and res corporalis therefore changes its meaning according to its context. When the Roman says that only a res corporalis can be possessed, he is referring to the thing itself; when he speaks of the acquisition of a res corporalis he means the acquisition of ownership.

Ownership and possession. The most fundamental distinction of all in the law of property is that between ownership and possession. ‘Ownership’, says Ulpian, ‘has nothing in common with posses­sion? And yet, for the most part, possession is the foundation of ownership. Thus, corporeal res nec mancipi are conveyed by traditio, a handing over of possession, and title by usucapio like­wise depends on possession. We must now examine this apparent contradiction.

lang=EN-US style='font-size:11.0pt'>