There are different ways or organising a law of contract. That is as much as to say that there are different ways of responding to the central tasks which contract has to perform.

So the right starting point is not the mode of organisation adopted by one system but those central tasks which have to be addressed by all. What are they?

To start with, leave the law out of it.

That is a simplified story about a game played according to real rules. The word �promise’ is not essential, but the idea of promising is. �Will you meet me at the theatre at 7.00?’ �Yes, I won’t be late.’ No promise as such, but plenty of moaning if lateness happens.

An invitation to rely can be expressed thus: �I want you to regard me as bound.’ That is what it comes down to. �Regard me as under an obligation.’ In whatever language it is actually done, that is the idea of a promise. The promissor puts himself under an obligation. He wants to incur the obligation, and he uses language (together with conduct and circumstances) to impose the obligation on himself. Notice, �I promise (or “vow”) to smoke no more cigarettes.’ Or, on a desert island, �I promise (or “swear” or simply “I must”) to remain cheerful.’ These are specially difficult because, though I try to impose the obligation on myself just as in the case in which I make the promise to you, there is nobody but myself (in the metaphor of binding) who can pull the rope. A vow to God helps with this problem.

The last paragraph shows that invitations to rely and the means of putting oneself under an obligation are not co-terminous. I put myself under an obligation to get you to rely on me, but I sometimes put myself under obligations for other purposes. There is room for a prolonged discussion. Which comes first? But no room here. It is safe to say that in human society the reason why people put themselves under obligations is, overwhelmingly, to engender reliance.

Where does the law come in? We have been talking about an extra­legal phenomenon. The law’s first task in the area is to say what has to be done if one wants to incur a legal obligation.

There is a second major task. And it is not in complete harmony with the first. A given promise, using that word in the widest sense, will induce a given reliance. But the reliance will not be explicable solely in terms of any words used. �Will you look after my suitcase till tomorrow morning?' �Yes, it will cost you 50 pence.' This is a common enough transaction. He has undertaken (promised) to look after my case and I am relying on that promise. But what exactly am I expecting of him? Must he stay up all night with a gun across his knees? What happens if the station burns down and my case with it? What if he gives it to another customer by mistake? The point is, the exchange of words is only a brief shorthand. It does not cover every eventuality. Even if there is a sheet of �small print' there will always be questions unanswered by the words used or their logical implications.

This second task derives from the laconic nature of undertakings. It is intending to build up packages of obligations inherent in transactions and, in particular, in ones which commonly recur. Reference is often made to �implied terms'. I much prefer natura contractus, the nature of the contract. The business is, to settle a version of what reasonable men would say if they set out to specify all the rules of, say, sale or hire or pledge and so on. That package states �the nature of contract'; and then, subject to more or less restriction according to the needs of the age, people can exclude or extend it as suits their circumstances and the nature of their bargain. �Implied terms' would be innocuous were it not that it leaves no separate language for terms genuinely intended but not expressed. The natura contractus is the law's version of the deal. It embraces matters which the parties themselves will never have thought of and whose inclusion in the contract cannot realistically be referred to the actual but tacit intent of the parties.

The reason why this second task is not completely in harmony with the first is touched on at the end of the previous paragraph. In fleshing out the skeletons of commonly recurring deals, the law imposes obli­gations which are reasonably incidental to the parties' intents but which cannot be referred directly to their agreement. If I sell a car I can say afterwards that I never agreed to assume any obligation in respect of quality; and when I have to pay damages in respect of a latent defect I may maintain that the obligation is as much �imposed by law' as the obligation to compensate for negligent injury inflicted in the course of driving.

We have identified two major tasks which a law of contract has to perform. It has to answer these questions: (i) What promises are binding at law? (ii) What package of unmentioned rules does a given promise carry with it? These are short versions of the questions, not intended to vary the sense of what has been said already.

The second of these tasks exercises a very definite influence on the organisation of contract. It favours the identification of specific con­tracts, the formation of a list of commonly encountered transactions in which people make and receive promises. Because a list is ideal for working out standard packages of law suitable for different contexts. On the other hand, the first (Which promises are binding in law?) pulls rather in a different direction. It invites a general answer, free of specific content, an answer in terms of mode as opposed to content. As, for example, when they are seriously intended; or, when they are in writing; or, when they are followed by a hand-shake. The reason why this question favours this kind of answer is that it takes little imagination to perceive the potentially infinite diversity of content. Hence, an answer in terms of content instantly meets the objection that the list would have to be too long. It would be unmanageable. Or, if it were short, it would simply fail to meet people's reasonable needs.

The last paragraph supposes two possible principles of organisation. One based on a general test: all promises, whatever their content, are legally binding if... The other, specific rather than general, based on content: promises are binding if they are promises to sell, buy, lend, borrow, pledge deposit and so on. On the other hand, if it is right to say that the law of contract has to perform both the tasks which have been identified, and if both pull in different directions, it will be difficult for a system to settle exclusively on one or the other pattern. Some sort of compromise will be necessary. It is an important differ­ence between Roman law and the common law that, though both compromises, Roman law puts the emphasis on specific contracts and the common law on general modes of contracting.

The Roman law of contract was dominated by eight specific con­tracts: sale, hire, partnership, commission (mandatum), loan for con­sumption (mutuum), loan for use (commodatum), deposit and pledge. This list covers the whole range of commercial and social life. But it is in the nature of any such list to be less than completely comprehensive. The contractual figures in the list rise like islands in the sea, an archipelago not a single continent.

The need for a general mode of contracting, free from specific content, was met in two ways. First, from earliest times the contract of stipulatio could be adapted to any content. To refer to it as �the contract of stipulatio' is to align it with sale, hire and so on. That is safe enough, once the difference has been observed. Stipulatio depended on promising in a particular manner, by the exchange of question and answer and probably using particular words. The spirit of the thing is captured if we imagine one person saying �Do you solemnly swear to...?' and receiving the answer �I solemnly swear it.' It is, quite literally, a form of words. So, strictly, it is a mode of contracting, not a type of contract. Secondly, there developed, by way of long-stop, a doctrine of contracts without a name, innominate contracts. According to that doctrine any agreement involving reciprocal performances became actionable by a party who had done his part. The roots of the doctrine are classical. As definitions were tidied up, something had to be done for those whom only the need for tidiness squeezed out. Barter, a horse for a kingdom, was taken out of sale by the requirement of a money price. And you will not find it in the list. We may be sure though that there never was a time when barter was remediless. Ad hoc, the gap was filled. The question is whether the ad hoc solution of this kind of problem by actions in factum or actions praescriptis verbis gave way during the classical period to a general doctrine. Perhaps only much later.

There should be questions in your minds as to why, if stipulatio was old and flexible, there ever was a need for actions on particular transactions and then, later, a need for another general doctrine. I shall deal with that in the context of stipulatio itself.