Claus-Wilhelm Canaris* & Hans Christoph Grigoleit**

The issue of interpretation arises when there is an ambiguity in the content of a contract. This is an omnipresent phenomenon of contract law: Even a thoroughly drafted and apparently precise contract may, under close scrutiny, prove to be in need of interpretation. Virtually every understanding of a contract includes - albeit often unsaid - an interpretation. In fact, the mere statement that a contract does not require or allow interpretation constitutes in itself an act of interpretation and the result of such an interpretation. From a philosophical point of view, therefore, the topic of interpretation of contracts forms a part of hermeneutics, i.e. the general theory of understanding human expressions. Indeed, there are some hermeneutic insights which also guide the interpretation of contracts. The founding principle of hermeneutics, for instance, that the interpretation of a word may be influenced by its context, applies to a contract just the same as to a piece of literature. The idea that a text can be "more intelligent than its author" and, accordingly, often has a meaning which the author was totally unaware of sometimes proves as useful for contracts as for other sorts of texts. More important, however, than such connections with the general theory of hermeneutics are the peculiarities of the interpretation of contracts. For the recipient of a love letter, for example, it only matters whether its author really meant his or her words, even if this understanding has found no outward expression but has remained purely inward. As opposed to this, a contract could obviously not fulfil its legal and economic function if the purely internal understanding of the author were decisive in every case. A similar statement can be made about the ambiguity of words. For example, a poet may use ambiguities on purpose as a stylistic device so that the interpreter has to highlight them, whereas the parties of a contract, and therefore also its interpreter, should avoid ambiguities as far as possible. Finally, the autonomous significance of interpreting contracts as compared to other fields of hermeneutics becomes obvious if one looks at the phenomenon of "gaps". While a gap in a contract may potentially be filled by "constructive" interpretation (cf. 3.3.2), a scholar of literature would thoroughly misunderstand his job if he supplemented a drama by a scene which the author himself failed to write or whose text was lost. The interpretation of contracts is therefore primarily determined by legal requirements and value judgements. Consequently, the legal principles on which contractual interpretation is based have to be considered first.

The interpretation of contracts must comply with their function which is primarily to enable the parties to govern their legal relations by themselves. Nowadays, the principle of freedom of contract applies as a general rule, in all European legal systems.1This is aptly expressed in Art. 1:102 (1) PECL: "Parties are free to enter into a contract and to determine its contents, subject to the requirements of good faith and fair dealing, and the mandatory rules...". Similarly, Art. 2.1. CEC states: "Les parties peuvent librement déterminer le contenu du contrat, dans les limites imposées par les règles imperatives, les bonnes mœurs et l'ordre public...". While the Treaty of the European Communities does not explicitly guarantee the principle of freedom of contract, it is generally accepted as a fundamental underlying idea2because the exercise of the expressly stated Community freedoms is only conceivable on the basis of freely negotiated contracts.

The principle of freedom of contract has several roots. From an ethical point of view, it is based upon the idea that the state must respect the autonomy of the individual. This assumption can be derived from the dignity of man and his right of "free development of his personality", of "pursuit of happiness" or similar concepts. These rights are violated if the state does not generally leave it to the individuals to regulate their relations themselves, but rather prescribes every detail; because by doing so the state turns into a guardian of the individuals, and this is irreconcilable with their dignity - a notion that has won nearly complete recognition in Europe since the age of enlightenment. Thus, freedom of contract is an expression of the legal self-determination of the human being and a sub-category of the autonomy of the individual.

In addition, the principle of freedom of contract has a foundation in political theory. It corresponds both to the ideas of democracy and of separation of powers. Democracy and freedom of contract are based upon the same basic values: on legal liberty and on the equality of the citizens. Hans Kelsen justly called the contract a "markedly democratic method of creating rights and duties" because the "subjects that are to be bound participate in the creation of the binding rule".3The link to the idea of democracy also becomes quite obvious if one considers the classic wording of Art. 1134 CC: "Les conventions légalement formées tiennent lieu de loi (!) à ceux qui les ont faites". The principle of separation of powers is strengthened considerably by a free contractual exchange4- in particular if the principle of competition is obeyed as well. This is because freedom of contract counterbalances the concentration of power in the government by shifting, within its area of operation, the competence to take legal decisions from the government to the citizen.

Finally, modern welfare economics accentuate and describe more precisely the social function of free contracting. According to the Pareto-criterion, the voluntary exchange by contract is the paradigm of economic efficiency. The individual exchange and its efficiency correspond on the collective level with the institution of the market: The mechanism of the market brings the independent transactions to perfection and, when operating ideally, creates optimal Pareto-efficiency. Thus, contract and market guarantee an efficient distribution of resources and a maximisation of social wealth, while, of course, not exhaustively solving the problem of just distribution.5Accordingly, the primary postulate of the Coase theorem to the legal system is that it should allow the voluntary exchange of goods and make it as easy as possible by lowering the cost of transactions.6

If the right of self-determination is taken seriously it necessarily includes the freedom to pursue and to agree to something unreasonable - just as the vote in a democratic decision is not subject to any control of reasonableness. Not reason but the intentions of the parties, therefore, form the basis of the contract according to the maxim: "Stat pro ratione voluntas".7This implies that the legal system only has to make sure that the decision of the parties is as free as possible not only in its legal but also in practical circumstances.

This assumption is in accordance with the fact that the economic goal of efficiency is to achieve at the greatest possible compliance with the individual preferences. Thus, the assessment of personal utility is left to the sole discretion of the individual. It is no contradiction that this utility is also measured by its monetary value, because money, in economic theory, merely has an instrumental character in relation to the individual evaluation of the utility. The individual determines the monetary value of his utility in the context of the transaction. However, this again does not preclude discussion on the issue of just distribution.

Accordingly, it is not legitimate to impose "reasonable" solutions upon the parties by means of interpretation. Rather, the goal of interpretation is to determine the intention of the parties - primarily the actual, alternatively the hypothetical intention. Only if this fails, may one resort to the "reasonable" intention; because it has then to be assumed that the parties are rational actors and thus intended something reasonable in case of doubt. The prevalence of the actual over the reasonable intention is generally acknowledged in all European legal systems8and also forms the basis of the rule of Art. 5:101 (1) - (3) PECL and of the similarly phrased rule of Art. 8 CISG (for more detail see 2.1. and 2.2.). Under Italian law, to give another example, it is widely held that the rule of Art 1366 CC "il contratto deve essere interpretato secondo buona fede" is subordinate to Art 1362 CC "Nell'interpretare il contratto si deve indagare

quale sia stata la commune intenzione delle parti...".9By contrast, it is questionable that Art. 39.4. CEC rules: "En tout état de cause, l'interprétation du contrat ne doit aboutir à un résultat qui soit contraire à la bonne foi ou au bon sens". If the parties intended such a result in an agreement it is not the task of interpretation to correct it; rather this may happen, if at all, according to the rules of law concerning the nullity of contracts.

There are (at least) two parties involved in a contract. Therefore the intention of only one party cannot form by itself the authoritative criterion for the interpretation of the contract because it constitutes a purely psychological internal state and, as such, is generally not discernible by the other party. Thus, one cannot simply resort to the in-ternal intention of the declaring party if the addressee understood the declaration differently. Conversely, the understanding of the addressee cannot be authoritative if it differs from that of the declaring party. Both assumptions directly result from the principle of freedom of contract and from the idea of self-determination. If contract terms which one of the parties neither intended nor could discern were regarded as authoritative, then his freedom of contract would be disregarded and his self-determination would virtually be turned into heteronomy.

On the other hand it would go too far to consider a contract invalid whenever the parties understood its terms differently. This would largely deprive the contract of its suitability as an instrument for effectively regulating the relations between the par-ties. It is therefore a compelling imperative of legal certainty to hold a party to a con- tract under certain circumstances, even if he was mistaken about its terms. Closely related is the notion that a party has to be protected in his reliance on the effectiveness and the terms of the contract if he understood it "correctly". It is generally fair and reasonable to hold the erring party to the "correct" terms because of, and insofar as he is responsible for, his "incorrect" understanding. Like every freedom, the freedom of contract and, accordingly, contractual self-determination correspond with responsibility, which restricts self-determination as far as it is necessary to protect individual reliance and the functioning of the markets.

What is the measure to determine the "correct" understanding of the contract if both parties have a different perception of it? One could take the perspective of a non-involved third party, but this would be in conflict with the fact that the declarations are generally - i.e. apart from special situations such as the declarations in commercial papers or in a corporate contract (see below 2.4.1) - not made to be received and un-

derstood by an "outsider" but by the addressee. Therefore, it is consequent to take the adressee's perspective as the relevant point of view, because the contract is an act of communication with this party only and it concerns only his or her interests. To define this perspective more precisely, one has to assume that the addressee seeks to understand the meaning of the declaration in a reasonable way and in good faith. On the hand, it is not expecting too much of the declaring party under the postulate of individual responsibility to be bound by a reasonable and fair understanding and on other hand, only reliance in such an understanding deserves protection by the law.

Therefore, this view has two sides. The declaring party as well as the addressee is held to an understanding which the latter party was able to have and ought to have had when a reasonable standard of good faith is applied. The test works equally for and against both parties: As little as the declaring party can allege his or her understanding when it is not reasonable or against good faith, he is not to be held to an unreasonable or unfaithful understanding the addressee might have had and vice versa. In this manner, the conflict described under 1.3.1 is resolved in a fair balance. Thereby, it is useful, but not necessary, to emphasise the notion of good faith besides the criterion of reasonability. The latter concerns more the aspect of rationality whereas the former has a connection to the principle of fairness. This is to say that a self-centred understanding of the contract may still seem reasonable from the per-spective of one of the parties, but it is not an understanding in good faith and therefore the legal order cannot take it as a standard for the interpretation of a contract.

Art 5:101 (3) PECL, which is modelled after Art. 8 (2) CISG, similarly refers to the perspectives of the parties by stating that "the contract is to be interpreted according to the meaning that reasonable persons of the same kind as the parties would give to it in the same circumstances". This view is generally shared in all European legal systems.10 Under Portuguese law, for instance, the rule of Art. 236 CC almost literally corresponds to Art. 5:101 (3) PECL. Obviously, the main problem for interpretation is to find out what a reasonable and fair understanding of a contract is. To give an answer to this question all circumstances of the individual case may be relevant (see in more detail below sub 2.1).

A different matter is whether a party may avoid the contract by reasons of mistake if he or she has misunderstood the content of the contract. Some legal systems, such as the German, the Swiss and de facto the Italian as well, have special rules for such a mistake ("Inhaltsirrtum"), whereas most legal systems and Arts. 4:103 PECL, 151 CEC treat such a mistake like other mistakes.11This issue belongs to the doctrine of mistake which is the subject of another chapter of this book.12It should be emphasised in this context, however, that the rules of interpretation are to be applied with priority, i.e. a remedy on the grounds of mistake may only be taken into consideration if the contract cannot be interpreted in accordance with the assumptions of the mistaken party.13

A theory of interpretation which is based on the actual intention of the parties is often called "subjective", whereas a theory which emphasises the external signs of the communicative act, such as the literal meaning of declaration in particular, is characterised as "objective". This antagonism pervades the entire history of interpre-tation, while the prevalence of one position over the other sometimes changed.14It is characteristic that Roman Law knew two different maxims of interpretation as the proposition of Papinian "In conventionibus contrahentium voluntatem potius quam verba spectari placuit"15and the word of Paulus "Cum in verbis nulla ambiguitas est non debet admitti voluntatis quaestio".16

Historically, the objective approach with its focus on the literal meaning of the words has been the starting point. This is related to the fact that in legal systems whose development has not yet reached an advanced level there is obviously a strong leaning to formalism and therefore an overemphasis of the role of the literal meaning of contract terms. In the course of legal and judicial development, however the modes of interpretation have become more refined and more flexible. Correspondingly, the idea of freedom of contract - which is the underlying principle of "subjective" interpretation - had to gain acceptance bit by bit against the original notion that only certain types of contracts are admissible.

Systematically, there are different issues underlying the conflict between "subjective" and "objective" interpretation. The first problem is the formalism caused by overemphasising the literal meaning of the contract. The result of such formalism can be that the parties are bound to the terms of the contract even if they have mistakenly phrased them or made them cover a situation which neither of the parties could have reasonably meant. Modern laws in principle reject such formalism. The French Code Civil, for instance, rules in Art. 1156: "On doit dans les conventions rechercher qu'elle a été 1a commune intention des parties contractantes, plutôt que de s'arrêter au sens littéral des termes". This formula is matched by the Italian rule of Art. 1362 CC almost literally and, likewise, the German BGB demands in § 133 BGB, "nicht an dem buchstäblichen Sinne des Ausdrucks zu haften".17Nevertheless the historical approach of formalism still has a certain influence today, albeit in modified forms (see 2.4.1 below).

The term "objective" interpretation fulfils a different function when it is used to describe the perspective of a reasonable party as the basis of inteipretation. This perception is not about overemphasising the literal meaning of a declaration, but about its divergence from the intention of the declaring party. In this respect the "objective" approach takes into account that the declaring party has not appropriately expressed his intention and therefore the addressee cannot reasonably consider the actual intention as relevant for the interpretation of the contract. In modern contract law, this as-

pect of "objective" interpretation has prevailed for good reason, as we just have described above under 1.3.2. Yet, this "objective" approach does not proceed "formalistically"in the sense of being strictly bound to the literal meaning, but it allows an unlimited variety of other criteria to be considered in addition and besides the meaning of the words see in more detail under 2.1).

With respect to this function of "objective" interpretation, one can draw a certain parallel to the dispute between the will theory and the theory of declaration, which played a major role in the 19th century legal discourse especially in Germany. This dispute has, however, no practical consequences for the theory of interpretation. This is to say that even the advocates of a strict will theory cannot argue that the intention of the declaring party determines the terms of the contract even if it could not have been recognised by the addressee. Doing so would mean to disregard the intention of the latter and would therefore be contradictory to its own premise. Even on the basis of a strict will theory, therefore, differences between the intention of a party and the "objective" meaning of his declaration can only have a practical impact on the question whether the contract is invalid or subject to rescission, which is answered to the affirmative under German law by § 119 Abs. 1 BGB ("Inhaltsirrtum"). The dispute between the will theory and the theory of declaration therefore becomes only relevant for the doctrine of unilateral mistake and does not prejudice the issue of interpretation.18

Interpretation is not only about the content of an agreement, but also about the logically prior question whether or not a contract has been formed at all. In some cases it may be doubtful whether a declaration is to be understood as an offer or just as an invitation to the other party to make an offer of his own (invitatio ad offerendum), in other cases it may be questionable whether a declaration is an offer to enter into a contract or just a declaratory message to confirm that an alleged contract has been formed before etc. To answer such questions, the same rules and criteria apply as to questions about the content of a contract.

The same applies if the parties did not express themselves in words, but merely through a certain conduct. It is common currency in contract law that a conclusive form of conduct - for example a certain motion of the head or raising a hand in a public auction - may be sufficient to imply a binding contract. This is, for example, expressly emphasised in Art. 2:102 PECL and in § 863 (1) of the Austrian ABGB. There may often be doubts about whether the conduct shown is in fact conclusive. This, too, is a problem of interpretation and therefore one has to ask how the other party reasonably had to understand the conduct in question.

The rules of interpretation in the different European legal systems largely overlap, as all national approaches are based on the described principles and their implications.

Corresponding to the principles discussed above (see 1.3.), most European legal systems and the PECL attempt to strike a balance between subjective and objective considerations.19The fundamental point for determining the content of a contract is the actual intention of parties (Arts. 2:101 (1) (a), 5:101 (1), (2) PECL). The goal of interpretation, therefore, is to establish what the intention of each party was at the time of contracting. As far as their intentions correspond, they form the content of the contract. The objective aspect of interpretation concerns the perspective that is adopted in order to determine the intention of each party. In this context, the content of each declaration has to be determined separately and the perspective of the addressee has to be adopted (cf. above 1.3.2). It is crucial how the addressee could reasonably understand the declaration in view of the individual circumstances of the case (Arts. 2:102,5:101 (3), 6:101 (1) PECL).

In ascertaining the perspective of a reasonable addressee, it is generally irrelevant whether the declarant knew or could have known that perspective. The declarant is protected sufficiently by the rule that his reasonable perspective is relevant for the interpretation of the corresponding declaration of the other party. If the two declarations do not correspond, the parties do not consent and no valid contract is formed.20

When establishing the meaning of a contract, the judge must consider all circumstances which allow one to draw conclusions with regard to the intentions of the parties at the time of contracting. Special consideration has to be given to the circumstances of the negotiations. In certain cases, circumstances which occurred before the state of negotiations - e.g. particular practices which the parties have established between themselves - or after conclusion of the contract - e.g. the specific handling of a certain aspect of performance mutually agreed upon - may also be taken into account.21It should, however, be emphasised that, from a theoretical perspective, those circumstances, regardless of when they occurred, are only circumstantial evidence as to the parties' intentions at the time of contracting.

The content of a declaration is to be determined objectively only to the extent that the addressee can reasonably rely on the "normal" use of language, "the regular" meaning of a certain conduct etc. It is therefore not justifiable to use an "absolutely objective" standard, i.e. a standard which is entirely independent from the individual situation and purely based on the declaration itself or its wording. Moreover, it is almost impossible to obtain such an "absolutely objective" perspective. In evaluating an act of communication, it is generally unavoidable to respond to the individual circumstances of the case, e.g. the branch of trade involved, the objectives of the parties etc. Otherwise, there is little chance to identify the meaning of words and of other means of expression used. This is in keeping with the general rules of hermeneutics

under which an act of communication is to be understood only with reference to its context. These aspects are neglected or at least unduly simplified when the subjective approach to interpretation is, as it is commonly done, contrasted with the objective approach. At a closer look, it is the subjective perspective of the addressee that matters and this perspective is objectified only to a certain extent, i.e. by the standard of reasonable understanding.

The wording, however, is necessarily the starting point of interpretation. It is of particular importance because it is the manifestation of the party's intention. In balancing against other criteria, the language used is normally to be given a rather significant weight. The wording may not impetuously be disregarded by arguing the reasonableness of a solution which is far from or even not at all in accordance with the text. This point may be demonstrated by using an example drawn from the Comments to Art. 5:101 (3) PECL, given there as an illustration of the objective method.22

A clause in an insurance contract provides that the policy covers the theft of jewellery only if there has been "clandestine entry" into the place where the jewellery was. An individual, A, pretends to be a telephone repairman and presents himself at Madame B's home to repair her telephone. A distracts B with some pretext and takes the opportunity to steal her jewels. The insurance company refuses to pay up, on the basis that there has been no "clandestine entry". On a reasonable interpretation entry gained by fraud is a form of "clandestine entry".

This example shows the risk of not taking the wording of the contract seriously, while the interpreter realises his own evaluation of a reasonable term. By its literal meaning, the term "clandestine entry" hardly encompasses "entry gained by fraud". One would use the term "clandestine" if the victim did not notice the entry of the thief, but not if she voluntarily admitted him into her home not realising his identity and intentions. "Entry gained by fraud" may surely not be considered to be within the core meaning of the expression "clandestine entry", but at its periphery at most. Therefore, the parties' interests and the purpose of the contract clause must be ana- lyzed thoroughly. In this respect, significant differences between the two situations become evident. If B did not suspect that an unknown person had entered her home, she had no reason to take special care to protect her jewels from theft. On the other hand, if she deliberately admits a telephone repairman into her home, the need to lock the jewels away or keep an eye on them is apparent. The crucial factor in interpreting the ratio of the expression is the aspect of control: It is immaterial whether the thief really is a telephone repairman or just pretends to be, whereas it is decisive that B, despite being aware of an unknown person's presence, left the jewels unguarded and let herself be distracted by some pretext. This could have happened with a real telephone repairman as well. With respect to the purpose of the insurance, B's need and worthiness of protection against theft by a pretender is almost as low as against theft by a real telephone repairman and not nearly as high as against theft by unknown intruders. Thus, the case is much closer to a situation which is definitely not covered by the insurance than to one which falls into the core meaning of the clause. That is why with respect to the purpose of the insurance contract and the clause in question, its

wording is to be construed narrowly. Theft after "entry gained by fraud" is therefore not covered by the insurance. This example shows how the criteria of reasonableness may be applied in a rational and methodical manner: The usual meaning of the words and the purpose of the clause to be interpreted must be combined with the double comparison of the situation in question with those definitely within and those definitely without the scope of the clause. Moreover, in the example, B cannot successfully invoke the contra proferentem rule (cf. below 2.4.2). Although the contract was phrased by the insurer, there are no remaining doubts which justify the application of the contra proferentem rule after careful analysis of its wording and purpose and taking into account the parties' interests.

If the parties' intentions correspond, but deviate from the regular understanding of their declarations, then neither party reasonably relies on the objective meaning of the contract. In such a situation, therefore, the corresponding intentions of the parties have priority over the "regular" or "correct" meaning of the declaration (falsa demonstratio non nocet). This priority is based on the above assumption that there is no "absolutely objective" perspective underlying the interpretation. What matters is the individual perspective of the addressee. This perspective is not purely shaped by the "bare" declarations, but also by all the circumstances which allow conclusions concerning the actual intention of the other party.

The priority of the corresponding intentions is largely agreed upon in all European legal systems.23 One formula, which is often used (e.g. in Art. 5:101 (1) PECL and in Art. 39 II CEC), states that the corresponding intentions of the parties have priority over the literal meaning of the words.24 If, for instance, the object of a sale is denominated as "Haakjöringsköd" and both parties take it for granted that this refers to whalemeat, even though "Haakjöringsköd" - according to the general use of language - means sharksmeat, then the contractual agreement is about whalemeat.25 This also applies if the parties knowingly use the term sharksmeat in the wrong sense in the written contract (e.g. if they fear that trading openly with whalemeat might damage their reputation). In such situations, the priority of the corresponding intentions is in line with the generally recognised rule of simulation (Art. 6:103 PECL):26 The true intentions of the parties prevail even if they have stipulated otherwise in an apparent contract.

The same applies if one of the parties realises or if it is obvious to him that there are certain intentions underlying the other party's declaration that contradict the normal understanding (Art. 5: 101 (2) PECL).27 Here, the interpretation deviating from

the normal understanding does not correspond to the common intention. Yet, the adressee's reliance on the regular use of language is not worthy of protection. Rather, the addressee can reasonably be expected to reveal the discrepancy if he prefers not to be bound to the intention of the other party. In a modified version of the example given above, this means that the seller owes whalemeat if the buyer knows that the seller mistakenly uses the term "Haakjöringsköd" for whalemeat, notwithstanding that the buyer intends to enter into a contract about sharksmeat (which might be the case because sharksmeat is more valuable).

Before we turn to the discussion of further rules of interpretation, some general characteristics of the law of interpretation have to be considered.

Although the principles discussed under 2.1 and 2.2 appear to be widely accepted and quite well founded, it is difficult to phrase them in more concrete terms and to complement them by further rules. The reason for this difficulty lies in the nature of communication and its fundamental dependence upon the circumstances of the individual case.28 It is possible to establish general rules, for instance by establishing pragmatic guidelines based on the experience of normal communication. However, the number of potential rules is practically unlimited. Moreover, any rule is prone to mistakes due to peculiarities of the case and must allow for a wide range of exceptions.

It is therefore not surprising that, even in continental systems, interpretation is largely governed by general principles and judge-made law.29 The most recent codifi-ation, the Dutch Burgerlijk Wetboek, deals very briefly with the topic of interpretation.30 The draftsmen of the German Civil Code (BGB) deliberately abstained from stating detailed rules of interpretation since they wanted to avoid instructing the judi-ciary in "practical logic".31 Nevertheless, detailed rules can be found in Arts. 1156 et seq. of the French CC, in Arts. 1362 et seq. of the Italian CC and in Arts. 1281 et seq. of the Spanish CC. The PECL also provide rules related to interpretation in Arts. 2:102, 5:101-107 and so does the CEC in Arts. 39-41. It is noteworthy that only limed binding force is attributed to the French rules of interpretation.32 The Cour de Cassation holds that a judgment does not have to be reversed purely on the ground that rules of interpretation have not been observed.33 The rules in the PECL are - in parts - consciously drafted in a way that courts may abstain from applying them in exceptional cases.34

A clear line has to be drawn between "real" rules and the mere enumeration of aspects that ought to be taken into consideration in the process of interpretation. In Art. 5:102 PECL, some factors are named that must be taken into account when interpreting a contract.35 This kind of list is supposed to outline circumstances that are of particular importance in determining the perspective of the reasonable addressee. Yet such a list will always be exemplary and very general, as demonstrated by Art. 5:102 PECL. The perspective of the addressee is characterised by an indefinite number of individual factors. Moreover, the choice of possible factors will always be the self-explanatory expression of common sense. Thus, there is little use in enumerating the factors relevant to interpretation: where such an enumeration is specific, it will always be incomplete; where it is general, it will only state the obvious.

The difficulty in setting up a precise scheme of rules also affects the question whether interpretation is a matter of fact or a matter of law. This distinction is relevant especially with regard to the scope of review of the trial court's decision on appeal. Interpretation is always based on facts, namely, the subjective intentions of the parties and other individual circumstances of the case. The core question of interpretation, whether a binding contract has been formed on the established facts and what its content is, necessarily requires an additional legal judgment.36 This follows directly from the assumption that the perspective of the reasonable addressee determines the outcome of interpretation. Because of the numerous and particular circumstances that potentially need to be taken into consideration, it is difficult, however, to clearly distinguish fact-finding from the application of law. As a consequence, findings of trial courts concerning interpretation should be reviewed with restraint, in a similar way as with findings of fact. In order to secure the priority of the trial judge's verdict, some legal systems deal with interpretation as a matter of fact, but nevertheless allow some limited review on appeal.37 It is more accurate and therefore preferable to regard interpretation as a matter of law while limiting reversion of the trial court's findings to cases in which the result of interpretation is evidently inconsistent with a legal rule.38

Admittedly, there is no strict concept of priority between the different aspects which play a role in the process of interpretation such as the meaning of the words, the purpose, the context or the origins of a contractual clause. These aspects are more or less loosely combined with one another. At one time a certain aspect prevails and another time a different aspect. Often the aspects are weighed by their persuasiveness. So there may be cases where a "weak" argument drawn from the meaning of the words of the agreement has to step back behind a "strong" argument derived from the purpose of the clause or the contractual context, but in other cases a narrow understanding of the words may prevail, because the purpose of the clause does not speak

clearly enough for the opposite or is even vague or indistinct. Hence, interpretation is a process of balancing.

Inevitably, a balancing approach bears considerable uncertainties. But this does not mean that interpretation of contracts evades rational inspection or is even an irrational and purely decisionistic procedure. Rather, balancing works as does reasoning in general: one gathers as many arguments as are worth considering, weighs them by their persuasiveness and strikes a balance between them if they are in conflict. Yet, the uncertainties of the balancing approach explain why it is so difficult to develop abstract and clear rules for interpretation. They also explain why it is preferable to leave some discretion to the trial judge in the process of interpretation and to limit the scope of judicial review on appeal.

Even though there is no strict priority between the different aspects of interpretation, some of them carry special weight from an abstract point of view without regard to the particulars of a case. First of all, this applies to the meaning of the words of the agreement. The words used are usually the manifestation of parties' intention as well as the object of their reliance (see 2.1 above). In addition, the meaning of the wording marks the border line with "constructive" interpretation, which may only be passed under certain conditions (see below 3.3.2). Thus, to overcome the hurdle of the literal meaning, very strong arguments are needed. The situation is similar with respect to arguments which are drawn from the purpose of a contractual clause or the purpose of the entire contract, since the parties use the contract as a means to pursue their specific goals. Moreover, the focus on the purpose of legal arrangements generally prevails in modern jurisprudence. In many cases, however, the purpose remains unclear or can only be determined on the basis of other aspects such as the literal meaning of the words, the context or the origin.

Even though some of the criteria, such as the meaning of the words or the purpose, have important weight, they nevertheless remain elements within the process of balancing. Therefore, they may have to step back behind other arguments if those turn out to be stronger in the particular case. This would be different in the case of a clear rule. If, for example, a rule stated that one has to consider only the aspects expressed in a written document and that one may only refer to external aspects if they are known to everyone - as it is stated with respect to the interpretation of negotiable instruments (see below 2.4.1 ad finem) -, it would clearly follow, that external circumstances in the process of formation of the contract would not be allowed to be taken into account. In this case, interpretation is not guided by a "more or less" as in a balancing process when reasons have to be weighed against each other, but by a "yes or no".

The Roman Law principle cum in verbis nulla ambiguitas est, non debet admitti voluntatis quaestio39 is still effective especially in the French doctrine of clauses

claires et précises. As far as a contractual clause in a written contract is phrased unambiguously, the Cour de Cassation holds that there is no interpretation to be carried out and only the objective meaning is relevant. Thus, there is an irrebuttable presumption that the unambiguous clause is the correct and complete expression of the parties' intentions.40

Prima facie, the French doctrine appears to be an exception to the rule that the corresponding intentions of the parties have priority over the objective meaning of the declarations. This would be irreconcilable with the principles governing interpretation: If one party can prove that both parties had corresponding intentions which deviate from the unambiguous meaning of the written declarations, none of the parties has reasonably relied on the objective meaning. However, this objection does not prove strong if one takes a closer look at the doctrine of clauses claires et précises since it is held permissible under this rule to consider external circumstances to determine whether or not the term in question is ambiguous.41 This is unavoidable also from a practical point of view, as there is no purely objective standard which allows a determination of the plain meaning of the words used. This leads to the second objection: The doctrine of clauses claires et précises is inconsistent because it presupposes a "regular" interpretation for determining the plain meaning or, respectively, an ambiguity, while claiming that interpretation is not allowed.42 This contradiction becomes apparent in a case where the Cour de Cassation decided on a shipment of sugar damaged on transport. There was a precise term in the contract providing for an allocation of transport risks. Contrary to the wording of the contract, however, the buyer was willing to accept delivery and to pay in full despite the damage the shipment had suffered, because the price of sugar had risen and therefore acceptance of the damaged shipment was preferable to rescission of the contract. The court ruled that, on the one hand, the clause concerning transport damages was clear, but, on the other hand, that the clause, being inconsistent with the apparent intentions of the parties in the situation given, was subject to interpretation.43

The obvious question is why the doctrine of clauses claires et précises has nevertheless not yet been abandoned. An important aspect might be that interpretation under French law is a matter of fact and thus there is, as a general rule, no review on appeal.44 The conclusion that a declaration is unambiguous in a certain sense is a means to reverse the decision of the trial court on appeal. Thereby, interpretation becomes - not openly but practically - subject to a review of evident mistakes on appeal as has been argued above.

Related, but not identical to the doctrine of clauses claires et précises is the parol evidence rule which has played an important role particularly in the English legal tra-dition. Under this concept, extrinsic (parol) evidence is, as a general rule, not admis-sible to add to, vary, subtract from or contradict the terms of a written document.45

Here the objection mentioned above is equally valid, i.e. none of the parties reason-ably relied on the objective meaning in case they had corresponding intentions that were wrong", or if they agreed on further terms of contract not embodied in the written document. Again, it has to be added that an interpretation which claims to leave aside circumstances that are not embodied in the written document is practically impossible. Accordingly, the rigid ban on external circumstances also had to be loosened under English law by way of recognising many exceptions. In 1986, the Law Commission even started doubting whether the parol evidence rule has any actual binding effect at all.46 It has to be admitted, however, that usually the parties consider the written document the complete and final expression of their legal relationship. Yet this assumption can be appropriately accounted for by regarding it a rebuttable presumption:47 Whoever claims circumstances not embodied in the written document to be valid bears the burden of proof for this claim.

This legal situation generally remains unchanged if the parties agree on a written form clause or a merger clause. Such a clause only implies a higher probability that the parties intend to restrict the content of their agreement to what is stated in the written document, thereby cancelling or excluding other (deviating) agreements. Such an intention, however, is subject to modifications: There is always the possibility that the parties - contrary to the clause - intend to leave certain other agreements intact or to provide new terms on certain issues after conclusion of the contract. The written form clause and the merger clause can be justified - like any other term of contract - on the basis of mutual consent and, therefore, they can be altered by mutual consent at any time.

It follows that a written form clause and the merger clause only strengthen the presumption of the written agreement being complete. This is in line with tendencies in most European legal systems48 and with Art. 2:106 PECL concerning clauses that aim to exclude other amendments to the contract than written ones. Not fully convincing, however, is the regulation of Art. 2:105 (1) PECL which attributes unconditional validity to individually stipulated merger clauses. This rule - along with its legislative ratio - demonstrates, again, that a rigid rule intending to limit interpretation to the written document is inadequate. The first indication for this can be found in Art. 2:105 (3) PECL, stating that declarations made at an earlier stage can be used for the interpretation of the contract. Furthermore, the merger clause is set aside according to Art. 2:105 (4) PECL if otherwise reasonable reliance of one party would be infringed. Finally, there is a reference in the comments on the PECL that a merger clause is also ineffective with respect to stipulations made separately from the contract.49 Again, these restrictions qualify the merger clause as giving rise to a rebuttable presumption, but do so in an unnecessarily complicated manner.

There is an additional problem if written form is required by law with regard to the contract to be interpreted. In this case, the question is what kind of relationship exists between interpretation and the legal requirement of form. In order to answer this question, two sub-questions need to be distinguished (which is often neglected): First, the content of the contract must be determined. Second, it has to be established whether the contract is valid with that specific content. With regard to the first sub-question, the content of the contract is to be determined on the basis of the general rules on interpretation of written documents. At this level, the requirement of form does not pose any restriction on interpretation. The starting point is therefore the presumption of the written document being correct and complete. This presumption is rebutted if one of the parties proves by way of external evidence that the actual intentions of the parties deviate from, or go beyond, the objective content of the written document. At the level of interpretation, there is no need to disregard extrinsic evidence due to the form requirement,50 as form requirements are generally not aimed at binding the parties to something that contradicts their corresponding intentions. Having established the content of the contract, one may turn to the second sub-question, i.e. whether the contract, with the terms given, is valid or invalid due to a breach of the form requirement. Generally, there can be no breach of the form requirement only on the ground that circumstances not embodied in the written document are relevant in order to determine the content of the contract, for it is practically impossible to phrase a written clause that is not in any need of inteipretation on the basis of external circumstances. Thus, the requirement of form must, at least, be limited to ensuring that the actual intention is somehow indicated in the written document.51 Yet, even if there is no such indication in the document with regard to a certain point, the contract is not necessarily invalid. Rather, the validity of the contract should be determined depending upon the specific form requirement and its objective.52 This objective may frequently but not always render the contract invalid. For example, under German law the contract of sale of real estate remains valid even though the actual piece of land has been designated wrongly in the written document, if it can be established what piece of land the parties actually wanted to refer to.53 This is because the requirement of form governing real estate sales aims to alert the parties to the general risks of such contracts and makes them seek legal advice by a notary public, and this ratio is fulfilled even if the technical designation of the piece of land is wrong.

Finally, particular caution is necessary with regard to cases where third parties who did not originally participate in the conclusion of the contract rely on its content and where the transaction is specifically designated to evoke such reliance. This is especially relevant with respect to negotiable instruments and corporate contracts. With regard to negotiable instruments, for instance, the general rule is that extrinsic circumstances can only be considered in the process of interpretation if they are known to anyone.54 A similar rule applies to corporate contracts if legitimate interests

of third parties, for example future shareholders or creditors, require such a restriction.55 In these cases, the third party might be solely depending on the content of the written agreement alone as the basis for his dispositions. Generally, the third party has no access to the individual circumstances accompanying the conclusion of the contract and, therefore, his perspective differs from the point of view of the original parties to the contract. The original parties do not deserve the regular protection by law if (and because) they exchanged offer and acceptance with the initial concern that third parties might rely on their declarations. In these cases, it is justified to generally have regard only to the written content of the contract and to circumstances obvious to anybody while other external circumstances are to be disregarded.

It is widely accepted that an ambiguous clause is to be construed against the party on whose initiative it was inserted into the contract. This rule has its foundation in the Roman law principle "Cum quaeritur in stipulatione, quid acti sit, ambiguitas contra stipulatorem est".56 The contra proferentem rule is laid down in Art. 5:103 PECL and in Art. 40 (3) CEC. Moreover, this rule is an integral part of the already existing Community law in Art. 5 of the Directive 93/13 on Unfair Terms in Consumer Contracts. In the European jurisdictions, the contra proferentem rule is to be found not only in regulations that implement Directive 93/13, but also in other contexts.57

The contra proferentem rule is mainly based upon the concept of deterrence. The party who introduces a clause into the contract can and should ensure its transparency and, respectively, avoid the uncertainty associated with ambiguous terms. This uncertainty is detrimental, since the other party is not sufficiently aware of the scope of his rights and duties when concluding the contract. Also, it is more difficult for the other party to evaluate the outcome of a dispute. The contra proferentem rule, moreover, preserves the legal status quo which would exist without the clause in question. If one party introduces a clause, it usually worsens the other party's existing legal position by establishing duties or limiting rights. The contra proferentem rule aims to achieve that the other party's 'well-earned' legal position is only restricted to an extent that is made perfectly clear in the contract. This idea is sometimes also expressed by demanding that clauses which limit essential duties or rights of one party have to be interpreted narrowly.58 Accordingly, disclaimer clauses are often construed very restrictively.59 Of course, this always bears the risk that the restrictive interpretation de facto serves to implement a legal prohibition.

The contra proferentem rule only applies to terms which were not individually negotiated (Art. 5:103 PECL; Art. 39 III CEC). If a term was negotiated between the parties there is no unilateral responsibility for an unclear formulation. Individual negotiation of a term also reduces the need for protecting the affected legal position.

Some rules connect the interpretation against one of the parties with the role this party plays in a specific contract. For example, there are rules in French Law which, in case of doubt, provide for an interpretation against the creditor (Art. 1162 CC) or against the seller (Art. 1602 CC). The idea of the contra proferentem rule might provide a certain justification for these rules which originate in Roman Law. In many cases, however, the contract is negotiated in detail or the debtor or the buyer is responsible for the drafting of the contract. If this is the case there is - possibly with the exception of promises without recompense60 - no plausible reason to put the creditor or the seller at a disadvantage. In a liberal contract system, interpretation cannot depend on social aspects, namely on the relative economic strength of the parties, not to mention that creditor and seller are not necessarily economically more powerful than debtor and buyer. Consequently, the rule on uncertainty contained in Arts. 1162, 1602 (2) CC can, if at all, only be justified as an inaccurate version of the contra proferentem rule.61

Numerous other rules of interpretation are practised beyond those already discussed. For instance, the proposition that an individual agreement takes preference over terms which were not subject to individual negotiations (Art. 5:104 PECL) is important and widely recognised.62 This idea is based on the legitimate reason that the individual negotiation of the parties allows a more precise conclusion with regard to the intention of both parties than an abstract reference. According to another, equally convincing rule, an interpretation which avoids rendering the agreement void or meaningless is generally to be preferred (Art. 5:106 PECL; Art. 40 II CEC).63 This is based on the generally justified assumption that the parties want to achieve the goals of the contract by reasonable and legal means.64

Apart from this, it should be emphasised again that one cannot expect rules of interpretation to give precise guidelines in making decisions. Therefore, one should be cautious when formulating them in binding form. For example, the rule in Art. 5:107 PECL and Art. 4.7 Unidroit Principles that, as far as contracts in various languages are concerned, the original version generally is authoritative, is unconvincing. The. underlying assumption that the original version reflects the intention of the parties most clearly is speculative. In particular, this rule fails to consider that the parties

have usually drafted the contract in different languages in order to have a better understanding. As another example, the norm of Art. 5:105 PECL,65 stating that the interpretation has to show consideration for the contract as a whole, appears to be superfluous since it is not a genuine rule but only one obvious aspect which has to be taken into consideration for interpretation (see 2.3.1). Aspects like these, which are only potentially relevant to interpretation, can be compiled into catalogues such as Art. 5:102 PECL. From the point of view put forward here, however, it is preferable to abstain from this kind of regulation for reason of the self-evidence and almost unlimited number of such potentially relevant criteria.

It is not always possible to solve the problems of interpretation by reverting to the wording of the contract or to the clear and concurring intention of the parties. In practice, contracts often contain gaps.

A gap in the contract may arise in case the parties' intentions diverge and this divergence cannot be resolved by means of interpretation. In such a case, there is no agreement with respect to one element of the contract. As a general consequence, the contract is void.66

Such a dissent occurs extremely rarely. Normally, a dissent is ruled out either because the parties notice their disagreement or because the rules of interpretation demand that both declarations be understood in the same sense. In general, the corresponding interpretation of both declarations is ensured by the rule that each declaration has to be interpreted from the perspective of a reasonable recipient. In most cases, it follows that the same circumstances of the case are decisive for both parties. That means that the perspectives of the reasonable recipient and, accordingly, the meanings of the declarations are the same for both parties.

Under German law, for instance, a dissent has been found in a case where both parties wanted to conclude a contract for the sale of tartaric acid by telegram. Both parties wanted to sell but due to the shortened language they failed to notice the equal intention of the other party.67 This case rather underlines, however, how seldom such a dissent occurs: Before the telegraphic declarations were exchanged one party had sent a price list to the other. Thus, one of the parties had made it clear that he wanted to sell and not to buy. Taking this into account, it would have been more appropriate to treat both declarations as congruent with the other party as buyer on the basis of the perspective of an objective recipient.68 By comparison, the English case Raffles v. Wichelhaus seems to be a more justifiable example for a dissent. Here the parties en-

tered into a contract on the sale of a cargo of cotton 'to arrive ex Peerless from Bombay'. Unknown to both parties, there were two ships called 'Peerless' which arrived at the agreed port of Liverpool at different times. If both parties really had different ships in mind and if there were no indications which allowed an unequivocal determination of the ship the contract was void.69 Finally, in France, a dissent occurred in a case where the parties' intentions did not correspond with respect to the currency in which the price was to be paid. Due to a currency reform, the agreement could have referred either to old or to new Francs.70

The case of differing declarations has to be strictly distinguished from the situation that the parties have not made any provisions at all with respect to certain questions. This may be the case either because they did not consider the question at all or because they deliberately abstained from dealing with it. As long as this gap does not affect fundamental elements of the contract such as, in particular, the parties, the subject matter of the contract, and the price, the contract is enforceable (cf. Art. 2:103 PECL). In that case, the questions for which no provisions were made have to be solved under the rules provided by law.71

All European legal systems contain supplemental rules to complete contractual arrangements. In France, e.g., they are called règles supplétives, in England terms implied in law,72and in Germany dispositives Gesetzesrecht. The necessity of these rules becomes obvious if one takes into account that the parties can never provide for all eventualities. Thus, the existence of suppletive law prevents contractual incompleteness (or voidness respectively) and thereby reduces the cost of negotiations and drafting. Suppletive law can, however, only achieve its goals if it responds to the typical interests which can be attributed to the parties of the particular type of contract. Therefore its content has to reflect the kind of arrangement which reasonable parties would have made if they had thought of the issue in question. Moreover, suppletive law serves to guarantee the fairness of contracts. While the parties generally do not have to justify the content of their contracts in terms of substantive justice in a legal system governed by freedom of contract, the opposite obviously applies to the legislator (respectively to objective law). If one specifies the applicable form of justice, contract law is primarily to be governed by the rules of commutative justice, whereas the appeal to principles of distributive justice is restricted to rare exceptions.73

In continental Europe, suppletive law is to be found mainly in the codifications of private and commercial law. It is not possible, however, to achieve a comprehensive codification because the parties can construct the contract freely according to their own interests. Moreover, even a detailed elaboration of all known types of contracts

would go far beyond the scope of a concise codification. Therefore, the body of suppletive law in a code will necessarily be subject to further concretisation by judicial case law which will grow in importance with increasing age of the codification. English Law shows that the supplementation of contractual gaps on the basis of objective law is possible and necessary even in the case where the pertinent rules are not codified at all.74

The rules of suppletive law do not always offer legal results that meet the particular requirements of the case at hand. Therefore, courts often attempt to derive particular solutions from the actual contract itself, i.e. from the common intention of the parties, even if the parties did not express their ideas explicitly.

As stated above (cf. 1.4), the content of a contract can be derived not only from the explicit declarations but also from the conduct of the parties. This kind of implied intention does not necessarily have to concern the conclusion of the contract as a whole. Rather, it is possible to derive implied provisions from the agreement which -in addition to its express terms - have to be acknowledged as a binding element of the contract because they reflect the actual intention of the parties (cf. Art. 6:102 (a) PECL). In particular, this is the case when a certain term is a necessary precondition for the meaningful performance of the explicit agreement. This kind of connection justifies the assumption that the parties actually intended the provision in question. The more obviously the explicit agreement depends on a term not explicitly agreed upon, the more likely the conclusion is justified that, according to the actual intention of the parties, the condition is an unexpressed element of the contract as well.75 An obvious example is the case of a car rental contract, which obliges the lessor to hand over the ignition key to the lessee even in absence of an explicit term to do so. This kind of recourse to implied elements of a contract is an essential tool of contractual interpretation. If it were not possible to consider these supplements, the parties would be forced into using an excessive amount of wording which would make the drafting process unnecessarily complicated and expensive.

Another important source of determining non-explicit intentions are the usual customs of a particular field of commerce. Provided that a solution to a specific question is established by custom, it is generally justified to assume that the parties were implicitly referring to the customary solution (Art. 1:105 (2) PECL; Art. 32 (1) lit. c CEC).76

Yet, often the connection between the express terms of the contract and a certain problem is not compelling enough to justify a solution on the basis of the parties' actual intention. Some jurisdictions, however, nevertheless provide for contractual complementation by way of interpretation in such cases.77 For instance, when filling gaps by constructive interpretation, the German BGH reverts to the rule which the parties themselves would have agreed upon with regard to the contract and the max-ims of good faith and common usage.78 On this basis, the Court ruled , for example, that two doctors who had swapped their practices were barred from opening a new practice in the immediate vicinity of the old one for a period of two to three years.79The French Cour de Cassation allows for considerable freedom in the interpretation of contracts as well. For instance, in a case where a radio station had contracted with an author to compose a radio play and had accepted the play without objections and paid for it, the court used the idea of contractual interpretation to find that the radio station actually had to broadcast the play even though there was no explicit agreement on that issue.80

In such cases, the only basis of interpretation can be the parties' hypothetical intention. However, in many cases, the distinction between the (implied) actual and the hypothetical intention of the parties is merely a gradual one. Therefore, a clear line cannot always be drawn.81It is a characteristic feature of constructive interpretation that the supplementary rule cannot be deduced exclusively from the contractual provisions. Instead, it requires an additional normative judgment with respect to the content of the agreement which goes beyond the reasonable recipient's perspective. For example, the German BGH reverts to the principle of good faith and common usage in this context. Moreover, Art. 6:102 PECL points in the same direction stating that, implied terms of a contract cannot only arise from the intention of the parties but also from good faith and fair dealing.82

It is, however, not only difficult to distinguish constructive interpretation from the case of implied intent, but also from the provisions of suppletive law. The reason is the additional normative assessment which is necessary in both cases. In some cases, it can even be doubtful whether a distinction is fruitful at all or whether the method of constructive interpretation in fact uses the contract and its interpretation to covertly formulate rules of objective law. This objection is not to be taken lightly. The reference to the parties' intention conceals that constructive interpretation profoundly

interferes with the contract in two ways: on the one hand, the omission of a contractual term is, generally speaking, just as meaningful as a positive agreement. Thus, the court has to disregard this negative exclusiveness83of the agreement if it implies constructive clauses into the contract. On the other hand, even if one generally acknowledges the need for a complementing clause, there is always a certain discretion with respect to the particular legal result. Here again, constructive interpretation takes the solution of the pertinent issue away from the parties' autonomy and negotiation.84It follows that constructive interpretation is only permissible under restrictive conditions. It is, in particular, not possible to justify it by normative considerations alone. Rather, the parties' particular declarations have to indicate clearly and specifically that the constructive interpretation would have complied with the hypothetical intention of both parties at the time the contract was concluded.85

The distinction between constructive interpretation and suppletive law becomes particularly important in legal systems where mechanisms for the adjustment of contracts in the case of unforeseen circumstances exist in suppletive law, i.e. apart from interpretation. Those mechanisms of adjustment, such as the German rules of the 'foundation of the transaction' ('Wegfall der Geschäftsgrundlage'; cf. § 313 BGB), are justly governed by strict requirements, which should not be circumvented by constructive interpretation. If, however, a specific mechanism for the adjustment of contracts to exceptional circumstances is missing, as, for instance, still appears to be the case in France, constructive interpretation can function as a "safety valve" with respect to an overly strict reading of the principle of pacta sunt servanda. By contrast, the open acknowledgement of a legal rule for contractual adjustment is preferable because, in many cases, the recourse to the parties' intention is merely fictitious.

All European legal systems acknowledge that the contract is not only the source of principal, but also of collateral duties which need not be described explicitly. These duties are simply founded on the parties' intentions if, in case they are not observed, performance cannot be perfected properly. The obvious example is again the duty to hand over the keys in a car rental. These duties, however, have to be distinguished from duties which cannot be derived directly from the principal duty and which aim at a more general protection of the other party's rights and legally protected interests while performing the contract. An example would be the duty of a painter not to damage the principal's furniture by drops of paint.

This second group of collateral duties is also acknowledged in all European legal systems. With respect to the issue of interpretation, the most important question whether it is possible to imply from the parties' intentions a general contractual duty to take reasonable care. This would correspond to the French doctrine that derives a general obligation de sécurité from the contract.86 Moreover, a famous English example for the contractual construction of such a duty is the Moorcock case. In this case, the defendant was contractually obliged to unload the plaintiff's ship at his jetty. When the tide went out the ship stranded and was damaged. The court established liability of the defendant on the ground of breach of his implied duty to take reasonable care.87

Yet there are two important arguments against a general duty of care based on the contractual agreement. First, there is generally no sufficiently clear indication to this end in the parties' declarations, in other words: establishing such a duty based on contract is fictitious. In terms of the parties' intentions, there is a clear difference between a general duty of reasonable care and other collateral duties directly aimed at achieving the contractually defined goal of performance. Accordingly, it is sufficient to have the parties bound to the contractual goal of performance in order to establish duties adhering directly to the principal duty, whereas establishing a general duty of reasonable care requires balancing the parties' conflicting interests.88 Second, it is objectionable to make the duty of care depend on whether or not a contract was concluded. For example, there would be no relevant reason to deny liability if the ship had been damaged at a time when the negotiating parties had not concluded a contract yet. Even if the contract was invalid, e.g. if the parties without noticing had actually reached no agreement on the price, the question of liability should not be dealt with differently.

The reason for the frequent assumption of a general contractual duty of reasonable care is that liability in contract can usually be established more easily than liability in tort. Liability in tort is restricted, as compared to contractual liability, in some European legal systems with respect to the principal's responsibility for his agent, the compensation of pure economic losses and/or the burden of proof.89 On the basis of the reasons given above, however, the question whether the collateral duties are contractual in their origin does not address the right issue. Rather, one should ask whether the rules of contractual liability - if they in fact provide a more "liberal" framework - apply even though the duties in question are not contractual in their origin, but imputed by law. This position has gained more and more support in recent times and was statutorily recognised under German law on the occasion of the reform of contract law in 2001 (see §§ 241 II, 311 II, III BGB). The reason for this legally imputed liability under the rules of contract law is that the parties necessarily grant each other an opportunity of interfering with the legal goods and interests of the other

that goes beyond the actual contractual performance and cannot reasonably be provided for in advance. The duty of reasonable care is correlated with this opportunity. It protects the parties' reliance on the beneficial effects of mutual cooperation and helps creating such reliance. The recognition of a duty of reasonable care and the resulting liability under the rules of contract law thus support and facilitate the conclusion of contracts. This concept allows one to establish a duty of care even in cases where no contract has been concluded yet or where the contract is invalid. It suffices that the parties - while negotiating or performing a contract - grant each other the opportunity to interfere with the legally protected interests of the other. The details of this issue, however, are beyond the scope of our topic in this article.90

BIBLIOGRAPHY: Académie des Privatistes Européens (Coordinateur: G. Gandolfi) Code Européen des Contrats, 2001 (quoted: CEC); J. Beatson Anson's Law of Contract, 28th edition, 2002; C.M. Bianca, Diritto Civile, III, II Contratto, 2nd ed. 2000; G. Cian; A. Trabucchi Commentario breve al códice civile, 5th ed. 2001; J. Ghestin Traité de droit civil, La formation du contrat. 3rd edition, 1993; J. Ghestin et al. Traité de droit civil, Les effets du contrat, 3rd edition, 2001; W. Flume Allgemeiner Teil des Bürgerlichen Rechts, II, 4th ed. 1992; H. Kötz Europäisches Vertragsrecht vol. I, 1996; O. Lando; H. Beale (eds.) Principles of European Contract Law, I and II, 2000 (quoted: PECL); K. Larenz; M. Wolf Allgemeiner Teil des Bürgerlichen Rechts, 8th edition, 1997; A. Lüderitz Auslegung von Rechtsgeschäften, 1966; H. Mazeaud et al. Leçons de droit civil, Tome II, 1st vol., Obligations. Théorie générale, 9th edition, 1998; D. Medicus Allgemeiner Teil des BGB, 8th ed., 2002; Paz-Ares Rodriguez et al. Comentario del Código Civil, 2nd ed., 1993; G.H. Treitel The Law of Contract, 10th edition, 1999; R. Zimmermann The Law of Obligations, Roman Foundations of the Civilian Tradition, 1990.

*Professor of Civil, Trade, Labour Law and Law Philosophy at the University of Regensburg.
**Professor of Civil, Trade and Company Law, European Private Law, University of Regensburg.
1Cf. the overview in PECL, 2000, pp. 99 et seq.
2Cf. ECJ Judgment of 05/10/1999 (Spain/Commission), C- 240/97, ECR 1999 p. I-6571, 6634 (No. 99); Grundmann JZ, 1996, 274, 278; 591.
3Cf. Kelsen, Reine Rechtslehre, 2nd ed. 1960, p. 285.
4See Bydlinski, Privatrecht und umfassende Gewaltenteilung, 2. Festschrift für Wilburg, 1975, 53, 67.
5Concerning the relation of freedom of contract and market to just distribution see detailed Canaris, Die Bedeutung der iustitia distributiva im deutschen Vertragsrecht, 1997, pp. 63 et seq.
6Cf. In more detail e.g. Posner, Economic Analysis of Law, 5th ed., 1998, § 1.2. = pp. 12 et seq.; Eidenmüller, Effizienz als Rechtsprinzip, 2. Aufl:, 1998, pp. 41 et seq.
7Cf. Flume, §1,5.
8See the overview given in PECL, 2000, pp. 290 et seq.; Kötz, p. 166.
9Cf. Cian/Trabucchi, Art. 1366, sub II. Some authors, however, attribute priority to the interpretazione secondo buona fede; see e.g. Bianca, p. 420 and 424.
10See the overview in PECL, 2000, p. 291 and the citations in note 19.
11See Kramer, Der Irrtum beim Vertragsschluß - eine weltweit rechtsvergleichende Bestandsaufnahme, 1998, pp. 87 et seq.
12See chapter 22 by Muriel Fabre-Magnan.
13See e.g. PECL, pp. 230 et seq.
14A comprehensive historical analysis is provided by Zimmermann, pp. 621 et seq.
15Pap. D. 50, 16, 219.
16Paul. D. 32, 25, 1.
17With the same wording under Austrian law § 914 ABGB.
18To the same effect Kötz, p. 171 with note 22; Ghestin, no 386.
19See e.g. Beatson, p. 31; Treitel, pp. 8-9; Ghestin, No. 390; Ghestin et al., 2001, nos. 9 et seq., Mazeaud, H. et al. No. 123; Cian/Trabucchi, Art. 1362, sub X. Moreover Art. 4.1, 4.2 Unidroit Principles.
20Such a dissent, however, is rare, because usually the reasonableness test defines identical perspectives for either party; for examples see below sub 3.1.
21This is substantially laid down under Spanish law in Art. 1282 CC.
22Cf. PECL, 2000, p. 289.
23For an overview see PECL, 2000, p. 290. For limitations see below 2.4.1. Under English law, the principle of falsa demonstratio non nocet is implemented with respect to written contracts by the equitable relief of rectification; see Beatson, pp. 339 et seq.
24Cf. e.g. under French law Art. 1156 CC. Under Swiss law Art. 18 (1) OR.
25Cf. RG 8.6. 1920, RGZ 99, 147.
26For an overview see PECL, 2000, p. 307.
27BGH 26.10.1984 BGH NJW 1984, 721. Under English law, again, the rules on rectification concerning unilateral mistakes lead to similar results; see Beatson, pp. 340 et seq. For an overview see PECL, 2000, p. 291.
28Cf. Zimmermann, pp. 638 et seq.
29For an overview PECL, 2000, p. 290 sub 1.
30Art. 3:35 BW refers to the perspective of a reasonable addressee.
31Cf. Motive I, p. 155.
32Cf. Ghestin et al., No. 35 et seq.
33See Cass, civ., 19.12.1995, Bull, civ., I, No. 466, p. 324.
34Cf. PECL, 2000, p 294.
35Similarly with regard to the question whether or not a statement constitutes a contract see Art. 6:101 (1) PECL. See also Art. 4.3 Unidroit Principles.
36Similarly Bianca, p. 413.
37E.g. under French law; cf. Ghestin et al., 2001, nos. 14 et seq.
38Cf. Larenz/Wolf, § 28 No. 132. For an overview see PECL, 2000, p. 290 sub 1.
39Paul. Dig. 32, 25, 1.
40Cf. in detail Ghestin et al., nos. 25 and 28. With regard to a similar practice under Italian law see Bianca, p. 420; Cian/Trabucchi, Art. 1362 sub X 1.See also Art. 39 (1) CEC.
41Cf. Ghestin et al., No. 27; Mazeaud, H. et al., No. 345.
42Cf. with more details Ghestin et al., No. 25.
43Cf. Req., 15.4. 1926, S. 1926, I, pp. 151 et seq.
44Cf. Ghestin et al., No. 15.
45E.g. Jacobs v. Batavia and General Plantations Trust [1924] 1 Ch. 287, 295. Under French law a similar rule is stated in Art. 1341 CC. This provision, however, does not apply if there is an ambiguity in the contract; cf. Req., 31.3. 1886, S. 1886, I, p. 260; Ghestin et al, No. 28.
46Cf. Beatson, pp. 132-134, 160-163; Treitel, pp. 176-183.
47This is substantially realised under Spanish law in Art. 1281 CC, stating the prevalence of the "intención evidente de los contratantes" over the literal meaning; cf. Paz-Ares Rodriguez et al., Art. 1282, sub II. To the same effect Beatson, p. 132.
48Cf. PECL, 2000, pp. 153 et seq.
49Cf. PECL p. 152.
50Cf. e.g. Cian/Trabucchi, Art. 1362 CC, sub IX 5.
51See BGH 9.4.1981, BGHZ 80, 242, 245.
52Cf. in detail Lüderitz, pp. 192 et seq.; Larenz/Wolf, § 28 nos. 82 et seq. = pp. 355 et seq.
53Cf. BGH 25.3.1983 BGHZ 87, 150, 152 et seq.
54Cf. e.g. BGH 30.11.1993, BGHZ 124, 263; OGH 1.12.1977, OGH SZ 50/157 (1977); Baumbach/Hefermehl, Wechselgesetz und Scheckgesetz, 22. Aufl. 2000, Introd. No. 56; G.H.Roth, Grundriß des österreichischen, Wertpapierrechts, 2. Aufl., 1999, p. 27.
55Cf. e.g. OGH 25.11. 1997 AG 1998, 199; J. Farrar/B. Hannigan Company Law, 4th ed. 1998, p. 117; G. Resta, Gli atti costitutivi e gli statuti, in G.Alpa/G.Fonsi/G.Resta, L'interpretazione del contratto, 2nd ed., Milan, 2001, p. 450. In detail see K. Schmidt Gesellschaftsrecht, 4. Aufl., 2002, § 5 I 4 = pp. 87 et seq.
56Cels. D. 34, 5, 26. With regard hereto Zimmermann, pp. 639 et seq.
57Cf. under English law Tan Wing Chuen v. Bank of Credit and Commerce Hong Kong Ltd. [1996] 2 B.C.L.C. 69. Under French law Art. L. 133-2 C. consom.; with more details Ghestin et al., 2001, No. 40. Under Italian law see Art. 1370 CC. Under Spanish law Art. 1288 CC. Under Austrian law § 915 ABGB. Furthermore Art. 4.6 Unidroit Principles.
58Cf. Palandt-Heinrichs, BGB, 2002, § 133, No. 23.
59See Hollier v. Rambler Motors (A.M.C.) Ltd. [1972] 2 Q.B. 71; Beatson, 2002, pp. 170-174; Treitel, p. 202.
60There are provisions specifically demanding an interpretation in favour of such promisor e.g. in Art. 41 s. 1 CEC, under Italian law (Art. 1371 CC), under Portuguese law (Art. 237 CC) and under Austrian law (§915 ABGB).
61See with the same result Kötz, pp. 174 et seq.
62For an overview over the jurisdictions see PECL, 2000, p. 295.
63To the same effect under English law Lord Napier and Ettrick v. R.F. Kershaw Ltd. [1999] 1 W.L.R. 756, 763. Under French law Art. 1157 CC. Under Spanish law Art. 1284 CC.
64Cf. e.g. Ghestin et al., No. 33.
65Similarly in French law Art. 1161 CC and in Spanish law Art. 1285 CC.
66Under English Law see Cundy v. Lindsay (1878) 3 App Cas 459; Beatson, pp. 321-323. Under French law Ghestin, No. 495. Under Spanish law Art. 1284 CC.
67Cf. RG 5. 4. 1922, RGZ 104, 265.
68Cf. Medicus, Rn. 438.
69Cf. 2 H & C. 906, 159 Eng.Rep 375 (1864).
70Cf. Cass. com., 14 janvier 1969, Bull. civ. IV, No. 13, p. 13.
71In order to justify the rules of suppletive law, it is inaccurate and not necessary to refer to the actual intentions of the parties; for a more detailed discussion see Ghestin et al., No. 42.
72Cf. Treitel, pp. 188-194.
73Cf. Canaris, Die Bedeutung der iustitia distributiva im deutschen Vertragsrecht, 1997, pp. 45 et seq., 75 et seq., 125 et seq.
74In more detail on suppletive law see Kötz, pp. 176 et seq.
75Cf. Art. 32 (1) lit.d CEC. Furthermore, see the definition of the term implied in fact in Shirlaw v. Southern Foundries. (1926) Ltd. [1939] K.B. 206, 227: "Prima facie that which in any contract is left to be implied and need not be expressed is something so obvious that it goes without saying; so that, if while the parties were making their bargain, an officious bystander were to suggest some express provision for it in the agreement, they would testily suppress him with a common 'Oh, of course!'."
76Cf. under English law ("term implied by custom") Treitel, pp. 194-195. Under Spanish law see Art. 1287 CC. To the same effect under French law Art. 1160 CC; Art. 1159 CC additionally refers specifically to the custom "...dans le pays où le contrat est passé".
77See for an overview PECL, 2000, p. 305.
78Cf. BGH 18.12.1954, BGHZ 16, 71, 76.
79Cf. BGH 18.12.1954, BGHZ 16, 71, 76.
80Vgl. Cass, civ., 1re, 2 April 1974, Bull, civ., I, No. 109.
81Critically in general on the distinction, Bianca, pp. 412 et seq.
82In Art. 6:102 PECL (b) "the nature and purpose of the contract" is stated as an independent consideration. This is unnecessarily complicated. Either the implied term results self-evidently from "the nature and purpose of the contract"; then it follows from the actual intention of the parties (see (a)). If this is not the case, the implied term needs an additional normative justification, which means that it has to be justified by objective principles such as good faith and fair dealing.
83The aspect of negative exclusiveness is laid down in Art. 1283 of the Spanish CC.
84Recently, it has been suggested to solve the problem of judicial interference by establishing certain duties of the parties to adapt or amend the contract by negotiation. Cf. e.g. Art. 157 CEC; in detail Nelle, Neuverhandlungspflichten, 1994. Generally, solutions by negotiation are always desirable. However, there is no evidence that it is possible to establish legal duties which, on the one hand, are »judicially workable and, on the other hand, can efficiently facilitate the process of working towards a voluntary agreement. Moreover, one has to keep in mind, that duties to negotiate can never guarantee the successful conclusion of a voluntary compromise and, thus, will never fully replace judicial interference.
85More detailed on the necessary limitations to constructive interpretation see Palandt-Heinrichs, BGB, 61. edition, 2002, § 157 Nos. 8-10.
86Cf. Ghestin, 2001, No. 48.
87Cf. (1889) 14 PD 64.
88Cf. the criticism on the Moorcock decision by Treitel, p. 193. Under French law, there is a dispute on the contractual character of l'obligation de sécurité as well; see Ghestin, 2001, No. 48 and footnote 310.
89See for a comparative analysis v. Bar, Gemeineuropäisches Deliktsrecht, vol. I, 1996, pp. 405 et seq, 459 et seq. With regard to the aspect of interpretation Kötz, 1996, pp. 184 et seq.
90With regard to the position asserted here see Canaris, JZ 1965, pp. 475 et seq.; Canaris, 2nd Festschrift für Larenz, 1983, pp. 27, 85 et seq. For a recent and detailed analysis see e.g. Krebs, Sonderverbindung und außerdeliktische Schutzpflichten, 2000.