term conventio is that men from different motions of the mind consent to one thing, i.e., arrive at one opinion. If we look at the proper meanings of the words, both these definitions seem to apply to all the types of mutual acts we have just enumerated. But we admit that the commonest usage is to apply the term pactum, or “agreement,” almost exclusively to acts which are obligatory on at least one side, i.e., acts by which a new personal right is given, as we explained above; and that is how our author seems to understand it.
Titius defines pactum, or “agreement,” as something done by the consent of two or more, given for the purpose of licitly creating or abolishing an obligation; and therefore includes under the term “agreement” not only acts consisting in the mutual consent by which personal rights are created, but also those by which they are abolished.6 But I do not think that we apply the word “agreements” to acts by which personal rights and corresponding obligations are abolished, more than to acts by which rights, whether real or personal, are transmitted from one person to another. It is irrelevant to our purpose as being a matter merely of arbitrary law, that the Roman jurisconsults and their interpreters used this term almost exclusively of acts which create by themselves only a natural and not also a civil obligation unless confirmed by a civil law, and on this ground distinguished them from contracts. [I.9.1.i]
Promises and agreements; two senses of agreement
We have just said that in his first paragraph7 our author understands by the term pactum, or “agreement,” an act consisting in mutual consent which is obligatory on at least one side. Now in this paragraph8 he restricts the word to one species within that genus, namely, that which is obligatory on both sides. However it should be noted, that if by the division here proposed a distinction is made in the case of an act obligatory by mutual consent, between one which creates obligation on one side only and one which creates obligation on both sides, the former is not well named a gratuitous promise. For there are acts which are obligatory on one party only, and which nevertheless, can by no means be called gratuitous promises, since they include from the other party, either some transmission of right, as in loan for consumption (mutuum) (where the term implies that the receiving party in this transaction is obligating himself to the giver, because the latter is at the same time transferring to him the ownership of the money given by mutuum), or the cession or remission of a right which was previously valid against the party creating the obligation, as is often the case in transactions.9 However if you are willing to understand by the term pactum, or “agreement,” in its narrower sense, a mutual act by which an obligation is contracted by at least one party (or, which is the same thing, by which a man transfers a personal right to be valid against himself), and in return a right of some kind is transferred by the other party, then pactum, or “agreement,” in its broader sense is rightly divided into gratuitous promise and pactum in the special sense. [I.9.5.i]
[Grotius distinguished agreements based upon mere declarations of intent and the necessity of keeping faith from perfect agreements, where there is a clear sign that a right is to be conferred (Grotius, Rights of War and Peace, II.XI, p. 281 ff). Pufendorf applied this distinction specifically to promising (Pufendorf, Of the Law of Nature and Nations, III.V.6, p. 269; Pufendorf, On the Duty of Man and Citizen, I.9.6, pp. 69–70). Carmichael comments:]
The authors apply this distinction of perfect and imperfect to promises rather than to agreements in general, perhaps because every reciprocal agreement is perfect in the state of nature. I say, in the state of nature; for in civil society, only those agreements and promises which may be enforced in the courts by an action taken under civil law are considered perfect. And just as a perfect promise confers a perfect right, I do not see why an imperfect right should not follow from an imperfect obligation. [I.9.6.i and ii]
“Error” in promises and agreements
[In Roman law promises and agreements are frustrated by various forms of error, fraud, or force which occur in the making of them. In all these categories Carmichael allows fewer circumstances to void an agreement than Pufendorf. Pufendorf’s first category is: “When in promising I have assumed something as a condition, without regard to which I would not have made the promise, there will naturally be no force in the promising.” Carmichael comments as follows:]
This is to be allowed only in the following sense: If I have either expressly declared this assumption on my part as a condition of the promise, or if I thought in good faith that it was understood by the promisee from the nature of the transaction. For my own silent thoughts which I cannot reasonably believe will be understood by the party I am addressing, do not alter the sense of what I say, nor consequently its moral effect. [I.9.12.i]
[Secondly, Pufendorf says: “If I have been impelled by error to make an agreement or a contract and I discover it when the matter is whole and nothing has yet been performed, it would be perfectly fair that the privilege of changing my mind be allowed to me.” Carmichael comments:]
Yet this privilege cannot be claimed by perfect right, unless the error concerns something which the person who was in error at least thought was assumed as a condition on both sides. And in reciprocal agreements, an event which is not explicitly put as a condition is not easily understood to be such, unless it is either affirmed in the article itself actually to exist by the other party to the transaction, or it is such that without the condition which it is agreed the promisor cannot perform, it would be manifestly impossible or absurd to fulfill the promise, or finally unless it concerns the actual object or matter which is the subject of the agreement, its valuable qualities or lack of them.
But if he who has made an error in an agreement says that he tacitly assumed that the event about which he was mistaken was understood on both sides to be a condition of the agreement, despite the fact that the other party did not and reasonably could not understand that the agreement was limited by that condition, the claim of error is still not completely excluded even in this case, provided that the claimant proves by proper evidence that he really understood the situation the way he says he did (for even in natural liberty the same judgment holds about what does not appear and what does not exist, as far as the external forum is concerned), and provided that he is prepared to reimburse any loss the other party may have incurred, in accordance with what we shall argue below at pp. 84–85. [I.9.12.ii]
[Pufendorf gives as his third form of “error”: “When a mistake has occurred concerning the actual object of the agreement, the agreement is defective, not so much because of the mistake, but because it has failed to satisfy the conditions of an agreement.” Carmichael has two notes on this:]
That is, as may be understood from what has been said before, that one of the parties to the agreement not only supposed that the object was of a different kind, or of different valuable qualities, than is in fact the case, but thought in good faith that this supposition of his and the will to make a contract on such an object was understood on both sides. [I.9.12.iii]
The celebrated Titius correctly observes here that the author was not right to make the distinction he does make, since the reason why this kind of agreement does not satisfy the terms of an agreement is that it is not being made about the proposed object as it really is.10 [I.9.12.iv]
[Carmichael concludes:]
The nature of the object of the agreement and its valuable qualities or lack of them, which may be thought to have had some weight in determining either of the contracting parties to make a contract which he would not otherwise have made, are naturally understood to belong to the essence of the actual contract. Hence an error committed in any of these matters, by natural law, vitiates the contract as long as no performance has been made by either party; if discovered after the contract has been wholly or in part fulfilled, it gives a right to the injured party to withdraw from it, and to require that any performance be restored on both
