II SALE

The emergence of the consensual contract of sale (emptio venditio') was a critical moment in the history of Roman commerce. Previously sale must have depended on an exchange of stipulationes, in which the seller promised to deliver the object of sale, and the buyer promised to pay the price.

The contract of sale came into being only if there was agreement on the essentials: this meant on the object of sale and on the price. This does not sound very demanding. But it did have two important consequences. First, it meant that it was not possible to have a contract of sale with a price to be fixed in the future or (according to some jurists) by reference to a third party, since agreement on a price was interpreted as meaning agreement on a fixed price rather than agreement on how the price was to be arrived at.

Second, it meant that there could be no sale of generic goods, such as a litre of wine or a pound of corn. These cases were not regarded as satis­fying the requirement that the object of the contract must be identified. A sale of generic goods was therefore regarded as coming into being only when the actual goods which were to be the subject of the sale had been

identifiably separated from the rest of the goods; or where the whole stock of goods rather than its individual parts was the object of the sale (Gaius, D.

These rules about certainty of price and object seem to go back to the days in which sale was a cash-sale, a transaction in which money and goods were exchanged on the spot. In developed Roman law, however, and even in the later republic, there was no need for the buyer and the seller to perform their parts of the contract at the same time, any more than this is necessary today. Classical Roman law never broke away from its refusal to recognize sale of generic goods. This was certainly not because trade in such goods did not take place: there are plenty of exam­ples of it in the Digest as well as in wrecks at the bottom of the Mediterranean Sea (cf Petronius, Satyricon 76). The conventional view is that the law could afford the luxury of this restrictive approach precisely because the stipulatio was still available as a means of entering into a con­tract for sale of generic goods. Curiously enough, however, there is no evidence that this approach was followed. Various substitutes for generic sale could be employed, such as loans repayable in kind, promises to pay penalties for non-delivery (Paul, D. 19.1.47), and so forth; but the lacuna in the law is perhaps smaller than has generally been appreciated: large quantities of goods could readily be sold, provided they were identified at least as a mass, and this would present no problems if they were housed in a warehouse or on board a ship, while agricultural produce might often be sold before it was harvested, making its identification straightforward (Ernst 1996: 276-99, 339-43).

In an ordinary contract of sale (emptio venditio) the whole point of the contract was that without need for express stipulation certain legal effects were automatically produced.

(i)    The seller impliedly warranted his title to the goods, so that the buyer who was dispossessed by a person who turned out to be the true owner was automatically able to sue the seller for breach of contract. Initially the buyer had to demand a stipulatio for indemnity or for a penalty (conventionally double) in the event that he was evicted from possession of the goods. (A special action was however available as of right if the goods had been made over by the formal conveyance of man­cipatio.) In due course it became established that as a matter of good faith the seller was obliged to give this promise - so that he could be sued for breach of contract if he did not. The logical conclusion of this develop­ment was reached when the guarantee became implicit in the contract of sale itself. This appears to have happened by the early second century ad (Julian, D 21.2.8).

This development illustrates the significance of the fact that sale was a ‘good faith' contract. What this meant was that the parties' dealings with one another were assessed in any eventual litigation on the basis of what good faith demanded; and so, without any need for adding further express promises or undertakings, the law on sale kept pace with the customs of trade and commerce: it was open to a judge to find that the failure of a party to act in accordance with ordinary commercial stan­dards was not consonant with good faith and therefore amounted to a breach of contract. The standard of good faith therefore gave the con­tract extraordinary vitality and flexibility.

(2)   The seller impliedly warranted the quality of the goods. Initially the buyer took the risk of defects in the goods: caveat emptor. If he wanted a guarantee against particular defects, he would have to take it expressly by stipulatio.

Here we see something of the same pattern as with the warranty against eviction: to start with, the aediles set out a list of defects which the seller should expressly promise were not present; it later became pos­sible to insist that the promise be given; and finally the warranty was implied in the contract. This is essentially the same regime as now governs sales of goods made in the course of business in Britain: the Sale of Goods Act 1979 sets out implied contractual terms on the quality and the fitness for purpose of goods sold. So far as Rome is concerned, the precise chronological steps by which the aediles' liability gradually came to extend to all sales everywhere are uncertain, but the trend was firmly in the direction of a general implied warranty of quality. Where the goods were defective, the buyer was able to reject the goods and reclaim the price within six months or seek a rebate on the price to reflect the presence of the defect within twelve months.

Some of the surviving documents provide an interesting perspective on these warranties. For example, a papyrus of ad 151 attests the sale of a slave who is warranted to be healthy in accordance with the edict; payment of double the price is also promised for the event that the buyer is evicted from possession. Both of these warranties are given by stipula­tio: particularly interesting is the fact that a reference to the aediles' edict and its list of diseases and defects is incorporated by reference in a stip­ulatio.

(3)   Once agreement had been reached on the essentials of the con­tract, the object sold and its price, the risk of accidental loss or destruc­tion passed to the buyer. This happened only when the contract was, as the Romans put it, ‘perfected', that is, any conditions to which it was subject had been satisfied, and the goods had been identified (where appropriate, by measuring them or weighing them out: Gaius, D 18.1.35.5). It has been regarded as strange that the buyer became liable at such an early point in the transaction, since he would not become owner until the goods were actually conveyed to him, for example by delivery. The result is that, at a time at which he did not own the goods, he was at risk if they were lost or destroyed in certain circumstances. The explanation for this again seems to lie in the fact that sale was originally cash-sale, and conclusion of contract, conveyance and transfer of risk all took place at the same time. That would not account for the survival of the rule at a time when the contract was regularly concluded well before the conveyance of the goods was made. But the risk which passed to the buyer was only for events which were not preventable by the seller, and this restriction was interpreted strictly: the seller was still liable if the goods were stolen, unless the theft involved overwhelming force. So the sorts of risks which the buyer assumed on conclusion of the contract were for destruction by earthquake, flooding or fire (so long as the seller was not responsible for it). The risk rule, odd though it may seem, was therefore kept within strict limits.

The terms just discussed were (ultimately) implied in every contract of sale.

Some of these pacts were in the interest of the seller, and others in the interest of the buyer. So, for example, the parties could include in the contract a term allowing the seller to call off the sale if the price was not paid by a certain day or if he received a better offer within a certain period. There were difficulties of interpretation in both of these cases: was the sale concluded immediately but subject to cancellation in the future? Or did it not come into being at all until the price was paid or no better offer was received? The Digest discusses the legal conse­quences of this choice (Paul, D. 41.4.2.3-4; Ulpian, D. 18.2.2 and D. 18.3.1).

Clearly in the buyer's interest, on the other hand, were arrangements for sale on approval (pactum displicentiae) or for tasting of wine before pur­chase, which appears to have been absolutely standard practice. An interesting example of sale on approval is given by Ulpian; as appears from the end of the text, the buyer seems to have been an acrobat who specialized in jumping from mule to mule (desultor):

This question is raised by Mela: if I have given you mules to tryout, foryouto buy them if you approve them, and for you to pay a daily re ntal for each one if you do not approve them, and within the trial period the mules have been stolen by thieves, what has to be paid: is it the price andthe rentaloronlythe rental? Mela says it makes a difference whether the sale has already been concluded or is to be concluded in the future: if it had already been concluded, the price could be claimed; if it was a future sale the rental could be. He said nothing about the appropriate actions. But I think, if the sale was concluded, the seller would have the action on sale (actio ex venditoo), and, if not, the appropriate action ought to be given against the acrobat. (Ulpian, D. 19.5.20.1)

Here again we find the same concern whether the sale comes into being at once, but is liable to be set aside, or whether it is suspended until the buyer has given his approval. The reason it matters is that the risk of loss of the goods passes to the buyer when the contract is concluded: if the contract has been concluded, the loss falls on the buyer and the seller is entitled to payment of the price.

It seems likely that inclusion of any of these terms in a contract would have affected the price of the goods; and it is clear from the way some jurists discuss them that they were regarded in a sense as being part of the price. Certainly, a buyer who is given generous terms for rejection of goods may have to pay a relatively high price; a seller who insists on stern provisions for calling off the sale may have to make do with a low one. All this gives a clear impression of the way in which the parties could negotiate the terms for their particular contract, using but modifying the basic background of the law of sale.