Originally published in Studies in International Law: Festskrift til Lars Hjerner (Stockholm 1990) 219-233. Reproduced with permission.
GAP-FILLING BY ANALOGY
Art. 7 of the U.N. Sales Convention in
Art. 7 of the United Nations Convention on Contracts for the International Sale of Goods (hereinafter cited as "the U.N. Convention") reads as follows:
"(2) Questions concerning matters governed by this Convention which are not expressly settled in it are to be settled in conformity with the general principles on which it is based or, in the absence of such principles, in conformity with the law applicable by virtue of private international law."
This Article can be said to combine three different rules:
(2) A rule regarding gap-filling, which is to be applied when the U.N. Convention does not contain any express rule on the matter involved.
(3) A rule regarding the relationship between the U.N. Convention and national law.
Clearly, the three rules are closely interrelated, and they convey a common message. Interpretation of the Convention should be autonomous, in the sense that it should not depend on principles and concepts derived from any national legal systems. In a similar spirit, the Article provides for gap-filling through analogy, which shall be given priority over the application of national rules. At the same time, through reference to rules of private international law which point to national legal systems, it is admitted that all questions cannot be settled by the method of analogy.
The immediate predecessor of Art. 7 of the U.N. Convention is found in Art. 17 of the 1964 Uniform Law on the International Sale of Goods (ULIS). This Article prescribes that:
It is obvious that the contents of this provision have been incorporated in Art. 7 of the U.N. Convention. However, there is a difference with regard to the position taken vis-a-vis private international law and consequently towards national law in general. Art. 2 of ULIS prescribes that the rules of private international law shall be excluded for the purposes of the application of this Law, subject to any provision to the contrary in the Law. ULIS thus aspires to wholly govern all questions relating to the contracts to which it is applicable, unless there is an explicit reference to principles of private international law. In this way, ULIS is even more autonomous than the U.N. Convention. Gap-filling is to wholly be undertaken by the use of analogy. I shall return to this matter later on.
All drafts of ULIS, from the 1935 draft onwards, have contained provisions similar to Art. 17. They can be explained by a general aim to counteract the "homeward trend" that must be expected in the application of a convention that deals with matters on which there are abundant rules in national legal systems. At the same time, the whole of Art. 7 of the U.N. Convention can be compared with some provisions that occur in national legal systems and cannot have this particular function.
Provisions of the type found in Art. 17 of ULIS and Art. 7 of the U.N. Convention belong to a special type. They do not regulate any special practical issues, nor do they establish general principles such as the one prescribing that unjustified enrichment should be returned, but they present methods for the application of a statute, code or convention. The use of provisions on method represents an old tradition, although there are also all-embracing codes that omit them. It is therefore appropriate to go back and examine the predecessors. Special attention will be devoted to the subject of gap-filling.
Allgemeines Landrecht für die preussischen Staaten (ALR) of 1794, known to be one of the most detailed codes ever, contains a number of provisions regarding the interpretation of statutes (§§ 45-58). One of them, § 49, requires the judge facing a dispute to which no special rule is applicable to decide it on the basis of "den in dem Gesetzbuche angenommenen allgemeinen Grundsätzen, und nach dem wegen ähnlicher Fälle vorhandenen Verordnungen, seiner besten Einsicht gemäss".
[Decide on the basis of "the general principles adopted in the code, and according to his best judgment, having regard to existing provisions applicable to analogous cases." English translation by Professor Willem Vis]
There is thus a clear reference to analogy. However, the judge is also to bring the matter to the attention of the legislature (§ 50). This provision clearly expresses great faith in the rational strength of the Code, together with the idea that it is the business of the legislature to make all important decisions. Judgments that have no immediate support in the text of the statute are to be considered as ad hoc solutions of the problems.
The French Code civil of 1804, although also based on natural law conceptions, does not contain any similar provision, only the famous Art. 4, which states that a judge who refuses to give a decision because the law is not clear is guilty of a refusal to provide justice, "déni de justice".
The Austrian Allgemeines Bürgerliches Gesetzbuch (ABGB) of 1811 is explicit on the method to be used when applying the Code. In § 6 it lays down principles of interpretation:
["A law, when applied, shall not be given an interpretation other than that desired from the proper meaning of words taken in their context and the clear intent of the legislator." English translation by Professor Willem Vis]
§ 7 concerns gap-filling:
["If a case cannot be decided with reference to the words of the law, regard shall be had to analogous cases explicitly dealt with in the Code and to the policies underlying other kindred laws. If the case still remains in doubt, it shall with careful consideration of the surrounding circumstances, be decided according to the principles of natural law." English translation by Professor Willem Vis]
Here there is a clear hierarchy established: first the text itself, then the analogy, and then natural law. By comparison with ALR, we find a more pronounced confidence in the principles of natural law, to be deduced primarily from the text of the Code itself and ultimately from reason.
The Austrian provisions are particularly interesting because of the fact that their ideas and wording, with only slight differences, reappear as late as in Art. 12 of the Dispozioni preliminari of the Italian Codice civile of 1942. The main difference is that instead of the reference to principles of natural law, there is a reference to the general principles of the legal order.
Bürgerliches Gesetzbuch für das Königreich Sachsen, from 1863, deserves notice as an important forerunner of the German Bürgerliches Gesetzbuch (BGB), which was finally enacted in 1896. It contains a comparatively great number of provisions on the interpretation of statutes (§§ 21-28). Besides going into more detail on interpretation in the strict sense than the Codes mentioned previously, it also contains a provision on the use of analogy (§ 25). More important, from an ideological point of view, is the provision on the force of customs (Gewohnheiten, § 28). Customs are stated not to have any power to change statutes nor to introduce new rules with statutory power. However, customs can be taken into account when parties are free to decide for themselves and it can be assumed that the parties concerned would observe these customs under similar circumstances.
The provision on customs points to an issue that came into the limelight when BGB was prepared. As has appeared, the great codes based on notions of natural law give prominence to analogy when the code itself is not explicit. On the other hand, at the time when work on BGB started, the historical school had for a long time been dominant in Germany, although the enthusiasm for it had waned. The historical school regarded custom as the ultimate source of law. The provision of the Saxon Code just mentioned points, however, to a certain ambiguity in the attitude to custom. When preparing a new general code for Germany, which could -- rightly -- be expected to be immensely important for a long time to come, it had to be decided what the source of guidance should be when the code did not explicitly regulate a matter. Should custom or analogy be the chief supplementary source?
The answer to such a question might seem uncertain. Legal positivism was dominant in Germany during the latter part of the 19th century, and it could be regarded as the successor of the historical school. It was therefore possible that custom should be preferred as a means for filling gaps, in fulfillment of the tradition of the historical school. However, the provisions proposed by the first commission in its draft for a new code are explicit in preferring analogy:
§ 2. Gewohnheitsrechtliche Rechtsnormen gelten nur insoweit, als das Gesetz auf Gewohnheitsrecht verweist.
["1. If the code does not set forth a provision in respect of a certain (legal) relationship, the provisions laid down for analogous relationships shall be applied accordingly. In the absence of such provisions, the principles derived from the spirit of the legal order shall prevail.
2. Customary norms shall be applicable only to the extent that the Code refers to customary law." English translation by Professor Willem Vis]
The report accompanying the first draft discusses the role of custom in detail. While the importance of customary law in the absence of codification is fully admitted, it is stated that the transition from customary law to statute law essentially constitutes progress. A number of reasons are given as to why custom should not be allowed to change statutory law. Only for special issues, particularly in the field of commercial law, is the prevalence of usages over statutes admitted. In this field, there was a precedent in the importance attached to usages in the Allgemeines deutsches Handelsgesetzbuch (ADHGB) from the 1860's.
Having disposed of custom in this fashion, the report goes on to discuss interpretation and analogy. Rules on interpretation are stated to be superfluous. It is said that it is the task of legal science, not of legislation, to discuss interpretation. The legislator who tries to prescribe how interpretation should be done does not regulate ("disponirt nicht") but teaches ("unterweist").
As has appeared, the draft contains a provision giving full force to the use of analogy. The first commission was strongly in favour of analogy as a means for gap-filling. It was pointed out that no statute can be complete in the sense that it deals explicitly with every possible case. The reliance on principles of natural law -- as indicated in the Austrian ABGB -- should not be allowed. Decisions must, according to the first commission, be taken not freely but on the basis of evaluations found in the statute.[l2]
The result may seem a bit paradoxical. By rejecting the ideas underlying the natural law codes, the commission arrives at proposing a rule that closely resembles those found in these codes. Analogy lies closer at hand to a positivist attitude than reliance on custom does
The second commission, like the first, paid more attention to custom than to analogy in the record of its deliberations. The attitude of rejection of custom seems to have been less strict, and in the end the result was that the provision against the application of custom was struck out entirely. The same fate happened to the provision on analogy. The majority of the second commission pronounced itself in favour of the contents of the rule proposed by the first commission, but it did not consider an explicit provision necessary. Interpretation and analogy were said to be closely connected, and if there was no rule on the one subject, rules on the other subject should also be omitted. The risk that considerations of natural law would be applied was not considered serious, and anyhow the proposed provision was not found to provide any protection against it.
The result was thus that no provision on either the force of custom or the use of analogy was introduced into the BGB. However, the impression that can be gathered is that the use of analogy is considered not only fully legitimate but also important for the filling of gaps in the Code.
Let us now return to the Conventions on international sales. We find a striking similarity between the provisions on gap-filling in the earlier codes and the first draft of the BGB on the one hand and those in the Conventions on the other hand. But the background and the function are entirely different.
In the first place, the role of custom, in the form of the binding power of usages, is an issue that has become separated entirely from the general problem of gap-filling. Usages are given preference before the rules of the Convention themselves, to a degree that is stated in detail in Art. 9 of ULIS and Art. 9 of the U.N. Convention. This matter will not be discussed further here.
In the second place, the function of the rules concerning gap-filling is entirely different, in spite of the similarity of wording. Whereas the wording of § 1 of the first draft of the BGB aimed at counteracting the supposed risk of applying natural law principles, the function of the rules in the Conventions is to draw a line against the application of rules in national legal systems. The ideological element has disappeared, but the practical remains.
As indicated in the beginning of this essay, there is a considerable difference between the rules of ULIS and those of the U.N. Convention. Whereas ULIS refers to analogy as the sole means of filling gaps within the area it governs -- in contrast to those of third party rights, restrictions on competition and similar matters, which lie entirely outside the scope of the Law -- the U.N. Convention allows the application of rules indicated by private international law as a means of gap-filling when no general principles capable of forming a basis for analogy can be found.
The importance of this distinction appears when one examines the -- not very numerous -- cases, mostly from lower courts, in which Art. 17 of ULIS has been the basis of decisions. It is by no means obvious what the result will be of applying the general principles on which ULIS is based, or, to put it in another way, where one should seek guidance when there is no explicit provision in ULIS.
In a couple of cases in which Art. 17 has been involved, reference has been made to a general principle ruling that contracts should be performed according to good faith. Such a principle is perhaps not pronounced in ULIS in the explicit fashion that is the case with the American Uniform Commercial Code (in particular in § 1-102), but there can hardly be any doubt that such a principle underlies ULIS as well. The only point of doubt is whether it should be necessary to invoke Art. 17 in order to justify its application.
A little more doubt can be entertained as to the principle of "abuse of right" being part of ULIS. The fact that a number of rules in ULIS refer to the acts of "reasonable men" seems to be a shaky ground for such a principle, especially for those whose legal systems do not make use (or not much use) of the principle of abuse of rights. However, this does not mean that the results arrived at should be criticized, only that a reasonable interpretation of the relevant rules might be sufficient.
Even more doubt arises when ULIS is supposed to contain general principles for which there is not even implicit support in its text, such as those referring to the incorporation of standard forms into a contract.
The more special cases of application of Art. 17 that have occurred illustrate the difficulties in finding support for the solution of all problems.
In a case decided by the German Bundesgerichtshof [Supreme Court] the dispute concerned the place where payment should be effected when a contract had been avoided and the price was to be repaid. The Bundesgerichtshof relied on analogy from the provisions on fulfillment of contracts involving transportation (Art. 59, Para. 1; and Art. 19, Para. 2) and found that they pointed to the seller's place of business as being the relevant point for all performance of such contracts. That place should accordingly, in the Court's opinion, also be considered the place for effecting repayment of the price. The Court rejected the view, suggested in legal writings, that returning performance on the avoidance of a contract should be considered the reverse of the original performance, and that accordingly the buyer's place of business should be the place of fulfillment in this case. An alternative way of stating the same view would be that according to Art. 59, Para. 1, the place for paying a money debt is the creditor's place of business. Whichever of these latter expressions one prefers, the analogy leads in a direction opposite the one chosen by the Bundesgerichtshof.
In a case decided by a Netherlands court, the question was on which of several outstanding debts of a debtor, payment should be considered to have been effected, when the debtor himself had made no indication. The court stated that the principle underlying ULIS was that payment should be considered to refer to the oldest outstanding debt. Here it must be asked on what provision of ULIS the court based the analogy.
A somewhat similar question was also presented to a Netherlands court. According to the court, payment should be referred to interest on debts with whose payment the debtor was in arrears, since Art. 83 prescribes that interest is to be paid when there is a delay in the payment of the price. However, the main objective of Art. 83 is to prescribe the rate of interest (i.e. the official discount rate plus 1 %), and it is difficult to see how this rule can give any guidance when deciding what kind of debt the payment should be applied to.
An issue on which it seems easier to deduce a conclusion by analogy concerns the currency in which payment is to be effected. Since the place of payment is the seller's place of business, it seems to be a good analogy that the currency in which payment should be made is that of this place.
The cases dealing with the application of Art. 17, although not numerous, illustrate the difficulties in applying the principles of such a statute as ULIS to each and every emerging question which prima facie falls within its scope. Although Art. 17 is intended to counteract "the homeward trend", there is a considerable risk that courts, familiar with their own law and aware of the fact that on many points it coincides with ULIS, will assume that there is a similar resemblance on points where this can neither be proved nor disproved.
The attitude inherent in Art. 7 of the U.N. Convention therefore seems more sensible, since it will encourage more openness in the discussion of the relevant issues  In so far as the U.N. Convention is less devoted to details and more to laying down general principles than ULIS was, it should also be easier to establish the field in which analogy is justified.
It is not possible here to try to delimit the field of legitimate analogy in the U.N. Convention. Only an attempt to structure the problem can be made.
There is an area in which the question of using analogy does not arise, because the use of general principles has made it superfluous. A clear example is that of deciding what is a fundamental breach. Art. 25 contains a general provision on that subject, which is applicable in all cases where the issue arises. Problems may arise as to the interpretation of Art. 25, especially since it appears to have been designed chiefly with breaches of contract by the seller in mind and may be less suitable for delays in payment, but these are different from the possibility of using it by analogy. Art. 74, concerning the computation of damages, seems to belong to the same category.
There are other provisions regarding which we can ask where interpretation stops and analogy begins. Art 28, which deals with the obligations of performance of a contract, in my opinion belongs to this group. According to this Article a court is not bound to enter a judgement for specific performance unless the court would do so under its own law in respect of similar contracts of sale not governed by the Convention. Should this principle be applied with regard to the payment of the price, the return of goods once delivered when a contract is avoided, the repair of goods, etc.? Some such questions would perhaps best be regarded as belonging to the direct application of Art. 28, others as rather falling within the scope of possible analogy and still others as even outside analogy. In any case, the questions cannot be answered without careful analysts. Some might argue that Art. 28 contains a rule of a type for which analogy should not be considered; its reference to the lex fori makes it unique. In my opinion this view is not well founded. It would lead to either a very wide use of what is supposed to be the literal meaning of the article or a lack of analysis of the reasons underlying the provision.
There are a number of provisions which are obviously suitable points of departure for analogies. If the question is asked whether it is possible under the U.N. Convention to declare a contract avoided when there is a breach of contract, without considering further circumstances, the answer undoubtedly seems to be no, with the important exceptions that the terms of the contract or the influence of usages might lead to an affirmative answer in a number of cases. There is no single provision that can be cited in support of such a contention, but the body of rules dealing with avoidance give clear guidance. In most cases only a fundamental breach gives rise to a right to declare a contract avoided. But then a new question arises, which also should be decided by the use of analogy. Can a Nachfrist procedure take the place of the test of fundamentality? According to the Convention this is not always the case, as appears, e.g., from Art. 49, which gives the buyer the right to declare a contract avoided by the Nachfrist procedure when the breach is delay but not when it is non-conformity of the goods. However, it would not be correct to draw the conclusion that the Nachfrist procedure is at the buyer's disposal only for delay in a limited sense. We must leave open, until the relevant arguments have been examined thoroughly, the question of whether, e.g., sending goods to an incorrect destination on a f.o.b. sale should be considered equivalent to delay.
There are a number of other subjects on which it seems impossible to derive any guidance from the U.N. Convention, for the simple reason that it does not deal with any subject even remotely similar to them. Such is the case, e.g., with the right to the proceeds from goods produced during the period between the making of the contract and its completion. What point in time, or what circumstance, should be decisive for drawing the line between the seller's and the buyer's rights? Since the Scandinavian Sale of Goods Acts -- both the old and the new ones  -- contain explicit rules on this subject, it is easy for a Scandinavian to see that these rules are independent of anything that is governed by the U.N. Convention. In my opinion there are cases which under ULIS were decided by reference to the general principles underlying it (cf. supra under 3) that should be judged similarly, in particular those dealing with the issue on which particular debt a payment should be considered to be effected.
Even if we attempt to distinguish between the types of situations now mentioned, there remains a considerable field of uncertainty. There are two different questions: should analogy be attempted, and if analogy is attempted, which analogy should be chosen? It seems necessary to answer these questions in the same spirit as the U.N. Convention prescribes for interpretation in a limited sense (Art. 7, Para. 1). Regard is to be had to the Convention's international character, the need to promote uniformity in its application and the observance of good faith in international trade. Sticking to national traditions is as little justified in this field as elsewhere in the application of the U.N. Convention.
The conclusion from the preceding survey is not only that the application of provisions prescribing the use of analogy for filling gaps is found to differ, in spite of the similarity in the wording, when we compare the old l9th-century provisions with the new provisions in the Conventions on international sales. It can also be concluded that the use of analogy is more complicated and calls for more analysis of policy than appears in the discussion on what rules of method a general code should contain. It will no doubt be a long time until even the most practical questions arising under the U.N. Convention are answered with any certainty.
1. Cf. J. Honnold, Uniform Law for International Sales, 1982, pp. 116 ff.
2. Concerning exceptions to the rules see J. Kropholler, Der "Ausschluss" des IPR im Einheitlichen Kaufgesetz, Rabels Zeitschrift 1974 pp. 372 ff.
3. Cf. Dölle/Wahl, Einheitskaufrecht, 1976, EKG § 17 Rdnr. 8.
4. Cf. References in Dölle/Wahl, op. cit., EKG § 8 Rdnr. 9-11.
5. Cf. Dölle/Wahl, op. cit., EKG § 17 Rdnr. 41.
6. Further provisions of the same type are cited by Dölle/Wahl, op. Cit., § 8 Rdnr. 29-45. See also M.J. Bonell, in Bianca & Bonell, Commentary on the International Sales Convention, 1987, pp. 76 ff.
7. The issue is presented historically in the report of the first commission 1887; see Mugdan, Die gesammten Materialien zum Bürgerlichen Gesetzbuch für das Deutsche Reich, Bd I, 1899, pp. 360 ff.
8. See for the general development F. Wieacker, Privatrechtgeschichte der Neuzeit, 2. Aufl. 1967, pp. 456 ff., and for the special development in the preparation of the pertinent sections of the BGB H. H. Jakobs, Wissenschaft und Gesetzbegung im bürgerlichen Recht, 1983, especially pp. 128 ff.
9. See Mugdan, op. cit., pp. 364 f.
10. See Mugdan, op. cit., pp. 365.
11. Mugdan, op. cit., pp. 365 ff.
12. Mugdan, op. cit., p. 366.
13. Cf. Jakobs, op. cit., p. 130.
14. See Mugdan, op. cit., pp. 568 ff.
15. See Mugdan, op. cit., p. 568.
16. These decisions, many of them not reported before, have been gathered in the collection Internationale Rechtsprechung zu EKG und EAG (P. Schlechtriem & U. Magnus eds.), 1987, pp. 182 ff. Altogether 14 cases are mentioned.
17. Some of the cases reported state simply that the rules of EKG do not refer to a certain matter or use the reference to Art. 17 as a means of strengthening a conclusion that refers principally to the interpretation of some other rule in the EKG. See [case numbers] 3, 5, 9, 11, 12, 13.
18. See no. 2 (Hof Amsterdam 5.1.1978), about the concept of "redelijkheid", pp. 182 ff, no. 7 (Oberlandesgericht Düsseldorf, 20.1.1983), p. 186.
19. See no. 14 (Oberlandesgericht Karlsruhe 25.7.1986), p. 190, pp. 286 ff.
20. See no. 10 (Oberlandesgericht Köln 24.10.1984), pp. 181, 187.
21. BGHZ 78, 257 (22.10.1980), no. 1, pp. 182, 351 ff.
22. See in particular Dölle/Weitnauer, Einheitskaufrecht EKG § 78 Rdnr. 6. Cf., regarding Art. 7 of the U.N. Convention, Bonell in Bianca & Bonell, op. cit. supra n. 6, p. 79.
23. See no. 4 (RB Alkmaar 27.5.1982), p. 184.
24. See no. 6 (Hof Amsterdam 4.11.1982). pp. 185 f.
25. See no. 8 (Oberlandesgericht Koblenz 21.1.1983), pp. 186, 342 ff.
26. Regarding the preparation of Art. 7 see in particular Bonell, in Bianca & Bonell, op. cit. supra n. 6, pp. 66 ff.
27. The subject is dealt with at length by Bonell, in Bianca & Bonell, op. cit. supra n. 6, pp. 78 ff.
28. See Art. 49, Para. 1 (a), Art.51, Para. 1, Art. 64, Para. 1 (a), Arts. 70, 72, 73.
29. Secs. 18-20 of the old Acts, Secs. 79-81 of the New Finnish Act (with which the Acts of the other Scandinavian states will probably be similar).
30. Cf., e.g., P. Volken, The Vienna Convention: Scope, Interpretation and Gap-filling, in: International Sale of Goods, Dubrovnik Lectures (Sarcevic & Volken eds.), 1985, pp. 19 ff. (pp. 42 ff.) with further references.
Go to Database Directory || Go to CISG Table of Contents