Presented in "Celebrating Success: 25 Years United Nations Convention on Contracts for the International Sale of Goods" (Collation of Papers at UNCITRAL -- SIAC Conference 22-23 September 2005, Singapore), published and copyright by the Singapore International Arbitration Centre at 28-38. Reproduced with permission of the SIAC.
Filip De Ly [*]
The United Nations Convention on Contracts for the International Sale of Goods ('CISG') is not a monolithic system; from the outset it envisaged coexistence with other sources of law as well as with private self-regulation (see parts II and III below). Also, its provisions on its international and substantive scope of application assume that it is not a self-contained system but rather that it interacts and leaves room for rules from other origins (see part IV below). Furthermore, the CISG does not intend to deal with all its gaps but leaves some gaps to be filled by domestic law (see part V below). Finally, the CISG reservations also imply openness for various rules (see part VI below). Developments since its adoption, however, also suggest new directions in terms of growing maturity and contributions to domestic law as well as to further instruments of international uniform law (see part VII below). These elements will be discussed below.
The central proposition of this paper at the occasion of the CISG's [25th] anniversary will, thus, tend to show that the CISG endorses an eclectic model in the field of uniform law.
II. THEORETICAL OBSERVATION
Traditionally, sales law was governed by domestic law in different jurisdictions. This led to many potential conflicts in international sales contracts which needed to be co-ordinated by conflict rules. This domestic conflict model corresponded to a Westphalian world view where international sales transactions were to be dealt with by domestic courts resorting to domestic conflict rules determining whether local or foreign domestic sales law was to be applied. The disadvantages of such a model such as high transactions costs, no level playing field or forum shopping are too well known and need not be explained further in this paper. It may be noted that this model has become even more cumbersome after decolonisation as of the 1950s. It just no longer provides optimal solutions with diversity of sales law in some 200 countries worldwide. [page 28]
Within this model, some drawbacks may be overcome. For instance, some risks relating to forum shopping may be tackled by concluding conflict Conventions such as the 1955 Hague Sales Conflicts Convention. However, these solutions only provide incomplete answers to the disadvantages mentioned above. They may tackle forum shopping but still involve high transaction costs related to differences in substantive laws and still do not provide for a common platform for the business community overcoming psychological and cross-cultural differences.
An alternative model is, of course, a uniform substantive law model. The International Institute for the Unification of Private Law ('UNIDROIT') first started this in sales law at the end of the 1920s, ultimately leading to the Uniform Law on the International Sale of Goods ('ULIS') and the Uniform Law on the Formation of Contracts for the International Sale of Goods ('ULFIS') in 1964. Uniform sales law reduces transaction costs, provides a common platform for international operators in Contracting States and reduces some of the disadvantages associated with forum shopping. The ULIS and ULFIS were very radical in their relationship to domestic law. They excluded domestic law altogether for those matters dealt with and imposed an autonomous interpretation and application (Article 17 of the ULIS) with no room for the traditional model to intervene by means of local conflict rules pointing to domestic law. However, this radical move and departure from the traditional model failed not only because the ULIS and ULFIS remained a continental European civilian approach to uniform sales law raising little appetite outside European civil law countries  but also for States adopting the uniform sales law because the exclusion of conflict of laws and the revolutionary assault on the traditional model proved to be unworkable. Not all issues covered by the uniform sales laws could be settled within these laws' systems.
Thus, an intermediate model developed at the occasion of the preparation and negotiation of the CISG under which a compromise solution was found stretching the uniform substantive model as far as possible while retaining a modest place for domestic law and conflict rules. The CISG did not intend to claim a monopoly over international sales law but gave way to a limited extent to other sources of law (domestic sales law) and to a different methodology (conflict of laws rather than uniform substantive law). A striking example of this balancing act may be found in Article 7(2) of the CISG. Within the CISG system, one is thus faced with a compromise between unity and diversity where unity prevails but leaves a [page 29] limited space for competing legal rules. This compromise solution, thus, reflects an eclectic model consisting of a bulk of uniform substantive law provisions complemented by conflict rules and domestic law.
The non-monopolistic pretensions of the CISG is also evidenced by Article 90 of the CISG which gives way to Contracting States' other international obligations regarding sales law but - apart from the ULIS and ULFIS where the mutually exclusive relationship was dealt with in Article 100 of the CISG and the Scandinavian position dealt with below under the Article 92 and 94 reservations -- no longer seems to raise practical problems.
To a limited extent, this may be different under some recent Conventions such as the 1988 UNIDROIT Factoring Convention, the UNIDROIT Cape Town Convention or the UNCITRAL  Assignment of Receivables Convention for instance in relation to the coexistence of sales law and retention of title clauses.
In this respect, one qualification may seem to be in order regarding European Union directives. On the one hand, the EU consumer directives (product liability, consumer warranties, unfair contract terms, distant selling, door-to-door sales) hardly seem to affect the CISG because consumer sales are excluded from the CISG's scope of application. On the other hand, the influence comes from more recent directives in the field of collection of claims, electronic signatures and e-commerce  which contain relevant provisions for B2B transactions such as interest rates, digital signatures and e-commerce contract formation which may interact with the CISG.
Finally, it must be noted that, even within the United Nations system, the CISG does not have monopolistic ambitions. The CISG does coexist with the 1974 United Nations Sales Limitation Convention  which was not incorporated into the CISG but only adapted in 1980 to co-ordinate with the CISG. More recently (July 2005), UNCITRAL concluded its 38th Session by adopting a Draft [page 30] Convention on the Use of Electronic Communications in International Contracting to be submitted to the United Nations General Assembly for final adoption in the fall of 2005 in New York. The official text of the draft Convention, incorporating the latest changes agreed at UNCITRAL's summer Session in Vienna, is due to be released in September 2005 as an annex to UNCITRAL's report to the General Assembly (document A/60/17). Upon entry into force, this newest UNCITRAL Convention will coexist with the CISG in relation to the issues covered by the new Convention in relation to the State Parties to the new Convention. However, even prior to entry into force or in relation to non-Contracting States, the new Convention may constitute persuasive authority regarding the issues falling within its scope of application and, thus, may complement the CISG.
Another aspect of the non-monopolistic claims of uniform substantive sales law relates to the relevance of self-regulation recognised by the ULIS and CISG . For instance, the parties under Article 6 of the CISG are free to exclude the CISG in whole or in part. As international commercial sales are, by and large, governed by party autonomy, the CISG basically provides default rules. This implies that international sales law can only be understood properly if sales contract practice is also analysed encompassing research into general conditions, standard or model contracts, the provisions of tailor-made sales contracts and the extent and reasons why parties do or do not exclude the CISG. These elements are often ignored for a number of reasons, including the blindfold of prevailing legal philosophical conceptions for law creation by private practice. If one were to accept that law can also be created by subjects other than States (lex mercatoria), it would [page 31] be easy to recognise that international sales law does not only live within the CISG (supplemented to a certain extent by domestic law) but also in the sales contracts themselves which have either excluded the CISG in whole or in part or supplemented its rules with provisions that correspond to certain problems and needs encountered in practice.
As to commercial custom or usage as a source of international sales law, the CISG also recognises any such source in Article 9, which again proves the non-monopolistic ambitions of the CISG. However, it may be noted that custom or usage -- except for closed business environments -- is increasingly losing relevance in view of rapidly changing business environments (globalisation, technological changes) and that the course of dealing is in this respect much more important.
IV. CISG'S SCOPE OF APPLICATION
After the ULIS's attempt to give a definition of international sales on the basis of objective criteria such as the cross-border flow of goods, the CISG returned, for reasons of predictability, to the subjective parameter of a party's place of business. Places of business in different Contracting States, then, leads to the application of the CISG. Any such definition may be under- or over-inclusive. A sale between two Italian companies where the goods sold are to be manufactured and then imported into Italy would not fall under the CISG. On the contrary, a sale between a US and an Italian company regarding shoes manufactured in Sicily and delivered in Florence would fall within the ambit of the CISG, although the shoes would never leave Italy prior to delivery. Definitional criteria may thus have elements of arbitrariness, but the CISG's place of business test has the advantage of certainty and ease of determination and would encompass an overwhelming number of international sales contracts. Consequently, the CISG cannot be blamed for an imprecise definition of its territorial scope of application or for covering an overwhelming majority of international sales transactions. It is further submitted that the limited number of sales which under an economic definition of internationality would be characterised as international but do not meet the CISG's internationality test, is a small price to be paid for achieving uniformity. In any event, the exclusion of any such sales was clearly envisaged by the CISG's drafters and again demonstrates the CISG's non-monopolistic ambitions. [page 32]
In practice, the international character of the sales contract is by and large readily determinable. The major problem in decided cases relates to the question whether liaison offices and representative offices may constitute a place of business for CISG Article 1 purposes. The majority view seems to be that they are not, based on the argument that a place of business requires an establishment, some duration and authorisation to act on the part of any such place. This interpretation is CISG friendly because otherwise the sales contract would be characterised as a domestic transaction between a local company and the local office of a foreign trader, subject to domestic sales law. The question is, however, whether any such view does not mix up requirements for branches of foreign companies (subject to all kinds of tax, administrative and corporate requirements) and places of business. The better view seems to be that liaison or representative offices are places of business for CISG Article 1 purposes but that the application of the CISG will be decided on the basis of Article 10(a). Under any such perspective, liaison or representative offices are places of business for CISG purposes but the application of the criteria of Article 10(a) will determine whether the liaison or representative office or any other establishment or place of business is to be retained to determine whether the CISG is applicable by virtue of Article 1(1)(a).
2. Substantive scope of application
Article 2 of the CISG is a rather straightforward provision and in practice does not often lead to protracted litigation. There may be some interpretation issues but these are by and large easy to solve. Parties in commercial practice have hardly had problems identifying, for instance, that oil and gas contracts come within the CISG's scope or that distribution, franchising or share purchase agreements ('SPA's') regarding international acquisitions do not. More difficult interpretation issues primarily arise in relation to software  and, more importantly in practice, regarding goods to be delivered, installed and maintained in the context of construction or infrastructure projects. [page 33]
By excluding some categories of sales from its scope for mostly evident reasons, the CISG is clear as to the categories of sales covered and excluded. This confirms again the CISG's eclectic model leaving room for domestic law regarding these excluded contracts.
The CISG's eclecticism is also confirmed by Article 7(2) of the CISG which is a compromise between the autonomy of uniform sales law solutions as embedded in the CISG and the need to resort to a certain extent to domestic law. Article 7 of the CISG applies to matters governed by the CISG and does not cover issues expressly excluded from the CISG, such as validity questions, property law issues or product liability which are settled by domestic law by and large applicable by virtue of a conflict rule.
For matters governed by the CISG, Article 7(2) refers to the general principles on which the CISG is based or, absent any such principles, to the law applicable by virtue of rules of private international law. This requires a two-prong approach: first, one needs to determine whether a matter is governed by the CISG and, second, whether there is a general CISG principle on which a solution to the problem at hand may be based. Both elements seem simple at first sight but raise delicate problems as to their application in practice. As to the first test, one may agree that issues regarding formation of sales and rights and obligations of the parties are covered by the Convention. This would imply that issues such as the applicable interest rate, hardship or the recovery of attorney's fees  are covered by the CISG and fall within the scope of the gap-filling provision of Article 7(2), but more doubts exist regarding penalty clauses or standard terms  where jurisdictions have different approaches as to treating these issues as validity questions or subject to subsequent judicial review.
An example of these difficult problems arose in the Gran Canaria Tomatoes case decided by the Dutch Supreme Court on 28 January 2005. The facts of this case related to a sale of tomato plants by a Dutch seller to a Belgian buyer. The plants were imported by the Dutch company from a seller in Gran Canaria and expert evidence had shown that the plants were infected by a bacteria. There were [page 34] no express warranties obtained from the seller in Gran Canaria and the Dutch seller relied on its general conditions of sale containing an exemption clause. The trial courts had decided that, under the CISG, there was an agreement between the parties as to the application of the seller's general conditions and that the Dutch seller was exempted from liability. The Belgian buyer attempted to avoid the application of the CISG by invoking Article 8(2) of the Rome Convention on the Law Applicable to Contractual Obligations protecting a party against acceptance by silence if any such protection is afforded by the law of its place of habitual residence. The Dutch Supreme Court held that the Rome Convention was not applicable because the matter regarding the formation of the sales contract was governed by the CISG. The gap as to the application of standard conditions, in the opinion of the court, was to be decided by the CISG's rules on contract formation and, thus, the acceptance of the seller's general conditions was governed solely by the CISG, pre-empting conflict rules. The Gran Canaria Tomatoes case shows that Article 7(2) of the CISG not only may pre-empt domestic rules regarding general conditions but also protective conflict rules.
The CISG reservations also contribute to the eclectic model that has emerged out of the CISG and to that model developing further away from unity and towards greater diversity and divergence. However, not all reservations are important in practice. These general propositions will now further be commented upon in relation to the various reservations.
1. Article 92 reservation -- The Nordic reservation favouring Nordic sales contract formation rules
Notwithstanding the complications raised by such reservation, it must be noted that the reservation is inapplicable in non-Contracting States and non-Reservation States and that even in the Reservation States (Denmark, Finland, Norway and Sweden) it will only apply if the law applicable by virtue of the reservation forum's conflict rule is Danish, Finnish, Norwegian or Swedish. The effects of the Article 92 reservation reduce uniformity and are to be regretted, albeit that in practice one seems to be able to live with it.
In this respect, one may add that the national committees of the International Chamber of Commerce in Denmark, Finland, Norway and Sweden have proposed to the ICC's Commercial Law and Practice Commission in May 2004 to endorse their initiative to have the Nordic Article 92 reservation abolished. The CLP [page 35] Commission has approved this proposal at its meeting in Rome on 2 May 2004. It was argued in support of this proposal that:
(1) the reservation was enacted in view of uncertain prospects of the CISG's success but can now be withdrawn since the CISG has received wide acceptance worldwide and Scandinavian countries should not find themselves in an isolated position;
(2) the reservation creates undue uncertainty in key areas of international sales regarding offer and acceptance where choice of law provisions are seldom used at this pre-contractual stage to solve problems;
(3) the reservation no longer meets the needs of businessmen in B2B transactions to conclude sales electronically or by other means of communication; and
(4) the Scandinavian rules on contract formation are not superior to those of Part II of the CISG.
2. Article 93 reservation -- States with various territorial units
Notwithstanding the complications to which this reservation may give rise, the question remains whether Article 93 is relevant and important in practice (i.e. whether and if so which Contracting States have used the Article 93 reservation to limit the scope of the CISG to only some and not other of their territorial units). The only example that comes to mind is in the Kingdom of The Netherlands where The Netherlands and Aruba are bound by the CISG but not The Netherlands Antilles. Further research indicates that Article 93 may also be relevant for Australia (regarding the Christmas Islands, the Cocos (Keeling) Islands and the Ashmore and Cartier Islands). Consequently, the Article 93 reservation hardly affects the CISG's scope and relevance.
3. Article 94 reservation -- The Inter-Scandinavian reservation
Under this reservation, parties having their places of business in Article 94 Reservation States (Denmark, Finland, Norway and Sweden) would not have their international sales contracts governed by the CISG but by domestic law if any such contractual dispute came before courts in any of these Reservation States. The Article 94 reservation again reduces the impact of the CISG but may be abolished soon because the underlying rationale (i.e. application of Nordic Council uniform regional law) has disappeared with Denmark not having joined the amendment to that law. Furthermore, the ICC proposal discussed above in relation to Article 92 equally and with the same reasons applies to the withdrawal of the Article 94 reservation. [page 36]
4. Article 95 reservation -- Exclusion of the CISG's indirect application
It is acknowledged that this reservation is probably the most complex one in view of the three variations of it that have been used. The first variation is the mere reservation against the extension of the CISG to sales where one of the parties has its place of business in a non-Contracting State but is faced with the application of the CISG by virtue of a conflict rule of the court having jurisdiction leading to the application of the law of a Contracting State. For instance, the United States, the People's Republic of China, Singapore, the Czech Republic and Slovakia have used this mere reservation variation. A second variation is the German variation where German courts will not apply Article 1(1)(b) of the CISG in sales where Article 95 Reservation States are involved. A third one is the Dutch variation. Article 2 of the Dutch Implementing CISG Act dated 18 December 1991 requests foreign judges in Article 95 Reservation States not to apply the Dutch Civil Code provisions on sales (Book 7, Title 1 of the Civil Code) but rather the CISG, if Dutch law were to be applicable by virtue of the local conflict rule. This suggestion is of course not binding on foreign courts but by enacting this Dutch solution the legislator has indicated that under Dutch law it prefers a solution which enhances uniformity rather than one that relies on local Dutch law.
5. Article 96 reservation -- Form requirement
Some Contracting States (Argentina, Belarus, Chile, Hungary, Latvia, Lituania, Russia and Ukraine) have made use of the Article 96 reservation to enforce their local requirements that a sales contract is to be concluded in writing to be valid. Traders may be faced and surprised by the operation of this reservation affecting the validity of their transactions. Notwithstanding the fact that this reservation and its use may be regretted, it is embedded in the CISG itself confirming its eclectic model.
VII. CONTRIBUTIONS OF THE CISG TO UNIFORM AND DOMESTIC SALES AND CONTRACT LAW
As the previous sections have tended to demonstrate, the CISG did not envisage a complete harmonisation of international sales law. Rather, it took a realistic stand as to compromising the perceived need for unification with the constraints [page 37] stemming from different factors which made the unification effort difficult and overall unfeasible if not limited in scope. Diversity was, thus, inherent in the effort in the first place and the CISG did not raise false hopes or unwarranted expectations.
Now that this year, the 25th anniversary of the CISG is being celebrated and the CISG has matured somewhat, one may wonder whether the CISG has so far achieved more than an incomplete unification as envisaged. The answer in my opinion is undoubtedly positive.
First, the CISG is indirectly influencing outdated domestic sales law. A case in point is France. A recent dissertation has shown how CISG concepts may influence the French 1804 domestic sales law and help to interpret some of its provisions. Further examples are Germany, Russia and China.
Second, the CISG had an important impact on other uniform law Conventions adopted by UNCITRAL or UNIDROIT  as well as on the UNIDROIT Principles of International Commercial Contracts and the Principles of European Contract Law.
Finally, as the CISG is subject to further scholarly analyses, applications in case law, collections of articles and cases as well as case law digesting, the CISG matures into a more reliable and more self-contained system where uniform patterns of interpretation and application prosper, notwithstanding the intended gaps left by its drafters. [page 38]
* Filip De Ly is a Professor of Law at Erasmus University Rotterdam. Professor De Ly studied at Ghent Law School (Belgium) and obtained a LLM degree from Harvard Law School in 1983. He worked for the US law firm of Cleary, Gottlieb, Steen & Hamilton in Brussels from 1983-1986. He is a Belgian citizen who has been working and living in The Netherlands for the past 17 years. Professor De Ly is frequently retained as counsel, expert witness or arbitrator in international commercial arbitrations and often gives legal opinions with regard to international business law problems. He has acted in more than 70 arbitrations in the capacity of expert witness, sole arbitrator, co-arbitrator or chairman of the arbitral tribunal. He has written five books (including International business law and lex mercatoria, Amsterdam, North Holland 1992) and co-authored/edited four books including a book on The place of arbitration. He published some 75 articles in the field of international business law, international commercial arbitration, conflict of laws, comparative law and EC business law. He has also published 14 case notes and 48 reports and book reviews.
1. Contracting States: Denmark, Finland, France, Italy, Norway, Niger, Switzerland, Sweden (see <http://www.hcch.net> consulted 19 June 2005).
2. Contracting States were Belgium, Germany, Gambia, Israel, Italy, Luxemburg, The Netherlands, San Marino, the United Kingdom (only as an opt-in). As of today, the ULIS and ULFIS only apply in Gambia, San Marino and the United Kingdom.
3. United Nations Commission on International Trade Law.
4. For a more detailed view, see Torsello, M, Common features of uniform commercial law conventions, A comparative study beyond the 1980 Uniform Sales Law, Munich, Sellier (2004).
5. See Directive 2000/31 on e-commerce dated 8 June 2000, OJ L 178, 17 July 2000, 1-16; Directive 1999/93 dated 13 December 1999 on common rules regarding electronic signatures, OJ L 13, 19 January 2000, 12; Directive 2000/35 on combatting late payments in commercial transactions dated 29 June 2000, OJ L 200, 8 August 2000, 35-38.
6. The 1974 Convention does not have the same ratification record as the CISG since it applies only in some 25 States, primarily in Central and Eastern Europe and Latin America (see <http://www.uncitral.org>). It may be mentioned that it also applies in the United States and in Western Europe only in Norway.
7. For the latter, see De Ly, F, 'The relevance of the Vienna Convention for International Sales Contracts, Should we stop contracting it out?' 4 Business Law International 241-249 (2003); De Ly, F, 'Opting out: Some Observations at the occasion of CISG's 25th Anniversary', in Quo vadis CISG?, Ferrari, F (Ed), Brussels/Paris, Bruylant/FEC (2005), 25-42. In both publications, the central proposition is that there is no empirical evidence for the statement that the CISG is often excluded in practice.
8. See De Ly, F, International business law and lex mercatoria, Amsterdam, North Holland (1992); De Ly, F, 'Lex mercatoria (New Law Merchant): Globalization and international self-regulation', in Rules and Networks, The legal culture of global business transactions, Appelbaum, R, Felstiner, B and Gessner, V (Ed.), Hart, Oxford (2001), 159-188; also published in 14 Diritto del commercio internazionale, 2000, No 3, 555-590; De Ly, F, 'Emerging new perspectives regarding lex mercatoria in an era of increasing globalization', in Festschrift für Otto Sandrock, Heidelberg, Verlag Recht und Wirtschaft (2000), 179-204. On lex mercatoria and the European Union, see De Ly, F, 'Lex mercatoria and unification of law in the European Union', in Towards a European Civil Code, The Hague, Kluwer Law International (1998, 2nd ed), 41-53. On lex mercatoria and the UNIDROIT Principles, see De Ly, F, 'National Report: The Netherlands: An Interim Report regarding the Application of the UNIDROIT Principles of International Commercial Contracts in The Netherlands', in A new Approach to International Commercial Contracts: the UNIDROIT Principles of International Commercial Contracts, Bonell, MJ (Ed), The Hague, Kluwer Law International (1999), 203-235; De Ly, F, 'Choice of law clauses, UNIDROIT Principles of International Commercial Contracts and Article 3 Rome Convention, The lex mercatoria before domestic courts or arbitration privilege?', Etudes offertes à Bathélemy Mercadal, Paris, Editions Francis Lefebvre (2002), 133-145.
9. See De Ly, F, International business law and lex mercatoria, Amsterdam, North Holland (1992), 134-164. For different views, see Bernstein, L, 'Merchant law in a merchant court: rethinking the Code's search for immanent business norms' (1996)144 U Penn LR 1765-1821; Gillette, C, 'Harmony and stasis in trade usages for internationales sales' (1999)39 Va J Int Law 707-741.
10. Digest of Case Law CISG, Article 1, para 13. The leading case is Fauba France FDIS GC Electronique v Fujitsu Microelectronik GmbH (Paris Court of Appeal, 22 April 1992), Case law on UNCITRAL texts ('CLOUT') Abstract No 158. The appeals decision was affirmed on other grounds by the French Supreme Court on 4 January 1995. See also, ICC Arbitration Case No 7531 of 1994 [English text, <http://cisgw3.law.pace.edu/cases/947531i1.html>], CLOUT Abstract No 304.
11. Bazinas, S, 'Uniformity in the interpretation and the application of CISG: the role of CLOUT and the Digest', Vienna Conference, 14-16 March 2005. The majority view seems to hold that software is a tangible good under the CISG, Article 1 and, thus, only escapes from the CISG under Article 3(2) when, as in tailor-made software or standard software which is extensively modified to fit the customer's needs, the preponderant part of the obligations of the supplier consists in the supply of labour or other services. This view may be criticised to the extent that it ignores the mixed nature of standard software contracts including licensing and servicing elements as well as the question whether software is tangible at all.
12. See CISG-Advisory Council Opinion No 4, 'Contracts for the Sale of Goods to Be Manufactured or Produced and Mixed Contracts (Article 3 CISG)', 24 October 2004. Rapporteur: Professor Pilar Perales Viscasillas, Universidad Carlos III de Madrid.
13. See Lookofsky, J, 'Walking the Article 7(2) tightrope between CISG and domestic law', Vienna Conference, 15-16 March 2005.
14. As to the latter, see Kröll, S, 'The CISG's scope of application: Articles 4, 5', Vienna Conference, 15-16 March 2005.
15. Case C03/290 with the opinion of Advocate-General Strikwerda, available at <http://www.rechtspraak.nl>.
16. See ICC Document 460/563, 1 May 2004.
17. There is some doubt whether the Czechoslovakian reservation survived that country's split.
18. It is unclear whether the People's Republic of China's declaration at the time of accession qualifies as an Article 96 reservation. Also, it may be mentioned that Estonia withdrew the reservation on 9 March 2004 (see <http://www.uncitral.org>).
19. Lamazerolles, E, Les apports de la Convention de Vienne au droit interne de la vente, Paris, LGDJ (2003), 464.
20. See Basedow, J, 'Towards a universal doctrine of breach of contract, The impact of CISG, Commercial Law Theory and CISG', Firenze Seminar New York University School of Law, 15-18 October 2004, International Review of Law and Economics (2005) (forthcoming).
21. See, for instance, Torsello, M, Common features of uniform commercial law conventions, A comparative study beyond the 1980 Uniform Sales Law, Munich, Sellier (2004).