I. HISTORICAL DEVELOPMENT

The idea of debt. We have provisionally defined a contract as an enforceable agreement, and the later classical lawyers would probably not have demurred to this definition.

Roman law was indeed precocious in that, as early as the Twelve Tables, it had a formal act which embodied the bare external essentials of agreement. This was the stipulatio, an exchange of a question and answer in formal words. In its earliest form the prospective creditor said ‘Do you solemnly promise [to pay me 500, or, to convey to me your horse] ?’ and the prospective debtor replied ‘I solemnly promise.’ fSpondesne...? spondeo.’’} But the validity of a stipulatio, like that of any formal act, came from its form and not from the agreement which the form no doubt embodied. For agreement was neither necessary nor sufficient.

Similarly, if we take as an example of the third type of debt a loan of money which has not been repaid, the later classical lawyers saw here an agreement. The borrower had agreed to repay the money, and it was the breach of this agreement which was the basis of the lender’s claim. But for the early law there was simply a debt: the basis of the lender’s claim was that he had paid to the borrower a sum of money which he, the lender, now ought to have. The difference between the two approaches can be seen if one takes another example. If X pays a sum of money to Y in the mistaken belief that he owes it, he may reclaim the money from Y. For the later classical lawyer, as for the modern lawyer, this is not a contract, since there has been no agreement; the obligation arises, in Justinian’s classi­fication, quasi ex contractu. But in the eye of the early law there is no significant difference between this and a loan—in each case a debt arises from the payment of a sum of money which the recipient is not entitled to retain.

Promissory and ‘real’ debt. The origins of contract lie probably in these two types of debt, the one deriving from a formal act, the other from an informal payment or transfer. But their scope was limited. The latter was clearly confined to debts of a specific thing or sum, and it is likely that the stipulatio was similarly restricted to promises to convey specific things or money.^{[68] [69]} Nevertheless there is a vital difference between them. The debt by stipulatio is a promissory debt, whereas the other is a ‘real’ debt, i.e.

For the development of commerce the concept of the pro­missory debt is essential. That this is so can be seen from a glance at the fundamental commercial contract of sale. A merely cash sale, i.e. one in which the thing sold and the price are exchanged simultaneously, and in which, therefore, there is no promissory element, raises few problems in the early law. Until the exchange is made there is nothing for the law to recognize; after the exchange is made there is little or nothing for the law to enforce.¹ But for anything more than the most rudimentary commerce a cash sale is not enough; what is needed is a pro­missory or credit sale, i.e. one in which either the payment of the price or the transfer of the thing or both are to take place at a later date. The seller or the buyer or both must be able to bind the other to the performance of a promise.

For the essential promises of buyer and seller (to convey the thing and to pay the price) the stipulatio, even with its restricted scope, was sufficient, but other commercial transactions require more than promises to convey a thing or pay a sum of money. Such are the many transactions involving the performance of services as, for example, the building of a house or the carriage of goods. For these, as for undertakings incidental to a sale (e.g. that the thing sold is free from defects), something more flexible is necessary. The first step was apparently to adapt the content of the stipulatio without formally extending its scope: the promisor promised to pay a penalty if he failed to perform the service in question. What was in substance a promise of a service was formulated as a promise of a sum of money.

The consensual contracts. Meanwhile, however, there had been a development of a different kind—the emergence of the principle that in certain typical transactions (which we call the con­sensual contracts, e.g. sale and hire) the parties could be bound by a mere formless agreement. The acceptance of this principle was one of the most important factors in the adaptation of the law to the commercial needs of a vast empire. And here, as in the earlier development of the stipulatio, Roman law was pre­cocious. English law did not finally recognize a purely executory agreement (i.e. one in which neither party has yet performed his side of the bargain) until the seventeenth century, and Greek law never did so. At how early a date the principle was accepted in Roman law it is impossible to say. The consensual contracts were certainly established in the first century b.c., but our knowledge of the course of legal history in the second century, which in this and in many other respects must have been the great formative period, is far too exiguous to permit of anything but conjecture.¹

Unilateral and bilateral contracts. There was a further difference between the stipulatio and the ‘real’ debt on the one hand and the consensual contracts on the other. To use modern terms, the former were unilateral, the latter bilateral. A unilateral contract is one which creates only rights in one party and only duties in the other; a bilateral contract is one which gives rise to reciprocal obligations, each party having both rights and duties.² In a loan of money, for example, the lender has a right to the repayment of the loan, and the borrower a duty to repay it, but the borrower has no rights and the lender no duties.

¹  See below, p. 165, n. 5.

² The English lawyer sometimes uses these terms in a different sense, a bilateral contract being one which is purely executory and a unilateral contract one which has been performed on one side but not on the other.

¹              Once the exceptio doli was admitted (see below, p. 164), the seller would in such It was only in a bilateral contract that the duties of the parties were fully reciprocal.

Stricti iuris and bonae fidei actions. It is not only for the prin­ciple of the purely executory contract that the consensual con­tracts are important, but also for the concept of good faith {bonafides) upon which they rest. Bona fides colours every aspect of these contracts, but it is in form and in origin a matter of the formulation of the actions by which they were enforced.

Unilateral contracts were enforced by stricti iuris actions and bilateral by bonae fidei actions.¹ The distinction is one of both pleading and substance, but it derives simply from three addi­tional words in the formula of a bonae fidei action.

We have seen that in the classical system of procedure pleading was strict:² no issue could be argued before the iudex unless it either appeared in the regular formula of the action or had been added to it by way of exceptio, replicatio, &c.

¹              Some quasi-contractual actions (e.g. on negotiorum gestio and tutela'} were also bonae fidei. The term stricti iuris is not classical.

^{2size=1 face="Times New Roman">} For what follows, see generally above, pp. 23 ff.

^J See the formula of the condictio, p. 24 above.

⁴ The formula given in the note on p. 24 above is that of a bonae fidei action (the actio venditi).

^s They originated perhaps in the second century b.c. and perhaps in the Edict of the Peregrine Praetor. In the classical law they, and therefore the contracts which they enforced, were considered to belong to civil law and not to Praetorian law, but it is probable that in origin they were Praetorian creations, the contrast being between stricti iuris actions which enforced duties deriving from the law (i.e. the traditional civil law) and bonae fidei actions which enforced duties deriving not matter of substance. For it was not until the middle of the first century b.c. that bad faith was allowed to be relevant in a stricti iuris action: it was only then that the defence of fraud or bad faith (exceptio doliy was admitted to the Edict. Until then the defendant in an action on a stipulation, for example, could not plead that his promise had been induced by the plaintiff’s fraud; he was strictly bound.

Once the exceptio doli was admitted, the difference of sub­stance became less fundamental, but it remained important. In matters of interpretation, and in other ways, the idea of good faith entered far more fully into the bonae fidei contracts than into those which were stricti iuris. For example, it was possible, as we have seen, to make a contract of sale either by a con­sensual contract (emptio venditio) or by two separate stipulations. But the incidents of the contracts would be very different. If, for example, the seller knew, and the buyer did not, that the thing was defective, or if the seller knew that the thing was significantly different from what the buyer thought it to be, this was sufficient to give the buyer in a contract of emptio venditio a cause of action, even though the seller had done nothing to induce the buyer’s mistake. But if the contract had taken the form of two stipulations, the buyer had no remedy. He had stipulated for that particular thing and the seller had delivered it; the seller had fulfilled the letter of his promise and the buyer had therefore no ground for complaint.

The idea of good faith does inevitably fit more easily with bilateral contracts than with unilateral. For in a bilateral con­tract the duties of one party are the counterpart of the duties of the other. In arriving at his decision the judge must strike a balance, and in doing so he can readily take account of matters of good faith. In a unilateral contract, on the other hand, the duty of the defendant has no counterpart in a duty of the plaintiff. The judge has no balance to strike.

The classification of Gaius and Justinian. In the Institutes of Gaius the outcome of the historical development which we have been considering is expressed in a fourfold classification of obligations from the law at all but from the moral concept of good faith. Thereafter this con­cept acquired, with use, sharper outlines and came to be accepted as a part of the ordinary law.                                ¹ See above, p. 22, n.i.

165 ex contractu. They arise, he says, either re (by the transfer of a thing—the primitive ‘real’ obligation in a contractual frame­work), verbis (by the uttering of formal words, as in the stipu­lation, Uteris (by the use of a documentary form), or consensu (the consensual contracts of which we have already spoken). Save for an enlargement of the category of obligations arising re,¹ Justinian in his Institutes follows Gaius, and it is accordingly customary to speak of real, verbal, literal, and consensual contracts. And it is in this order that Gaius and Justinian treat of them. The order has little, however, to recommend it,² and since the verbal and the literal are formal contracts and the real and consensual are informal, it is more convenient to rearrange them accordingly. But first we must consider the principal characteristic of the Roman system—that it is a law of contracts and not a unitary law of contract.

Law of contracts, not law of contract. The definition of a contract as an enforceable agreement is misleading, not only for the period of development but also, in a different way, for the mature law. It suggests a greater unity than in fact there was. Modern systems have a single concept of contract; Roman law had a list of contracts. In modern law it can, very loosely, be said that any seriously intended agreement is a contract, no matter what its content. In Roman law an informal agreement was not a contract unless it satisfied the requirements of one or other of the listed contracts. For example, the consensual contract of sale (emptio venditio} required that the parties be agreed on a specific thing to be bought and sold and on a fixed price. An agreement which did not satisfy one or both of these require­ments (e.g. an agreement to ‘sell’ for a reasonable price) was not a contract of sale.

The practical consequences of this lack of generalization were three. In the first place, the lawyers were able to work out in detail the ‘incidents’ of each type of contract. The parties to a contract will rarely have the foresight to lay down in advance what their rights and duties shall be in all possible eventualities, and the law must supply what they have left unsaid. The Roman system, by isolating under a small number of headings

¹                      See below, pp. 167 fF.

²                       It probably derives from the not very rational order of the Edict.

the characteristic transactions of life, was able to assign to each of them those legal consequences which seemed, commercially and otherwise, most appropriate. Once it had been determined that a given transaction fell under a particular heading, the established rules of that type of contract could be applied. And equally, the parties, in entering into the agreement, could know what consequences would follow where they were silent. Even modern law, in spite of its generalization in theory, resorts in practice to the regulation of typical contracts.

The second consequence is a technical one of the law of actions, but for that reason it was one which mattered much to the classical lawyer. We have seen that the classical law was a law of actions—that the question a lawyer asked was not whether a man had a right but whether he had a remedy. And we have also seen that for each cause of action there was in principle a form of action. Each type of contract, therefore, had the appro­priate form or forms of action, and the choice of the wrong form would lose a man his case. It was not sufficient for a plaintiff to prove to the iudex that he had entered into a contract. He must prove that he had entered into the type of contract specified in the formula of his action. It would avail him nothing, for example, to show a contract of hire if his action were one of sale, even though, on the point in issue, the incidents of the two contracts might be the same.

The third consequence is the most obvious—the possibility of gaps in the law. An agreement, though seriously intended, might fall within none of the recognized headings and therefore have no legal effect. An example of this is the one already given of an agreement to sell for a reasonable price. This shortcoming was, however, mitigated by the existence among the recognized headings of the stipulatio, which embodies a different principle. Methods of contracting and types of contract. The stipulatio was, as we have seen, a formal contract. It was not this, however, that gave it its importance (the other formal contracts are of little consequence), but the fact that it was a method of contracting rather than a type of contract. Any agreement could be made legally effective by being cast in the form of a stipulation. The agreement to sell for a reasonable price, for example, would become a contract if this simple step were taken. Equally, as

167 we have seen, an agreement which did satisfy the requirements of one of the informal contracts, such as an agreement to sell for a fixed price, could be cast in the stipulatory form. The stipulation, in short, supplied the element of generality which otherwise was lacking in the Roman system of specific contracts.¹ It could thus claim, as we shall see, the central position in that system.