Harry M. Flechtner [*]
1. OPENING COMMENTS
It is an honour to participate in a project that pays tribute to Professor Peter Schlechtnem, and his many invaluable contributions to legal scholarship, particularly in the field of international commercial law. I had the extraordinary good fortune of meeting Professor Schlechtriem soon after I started to work on the UN Convention on Contracts for the International Sale of Goods ('the CISC' or 'the Convention'). He treated me with remarkable kindness and collegiality (as all who knew him will find unsurprising). He listened patiently to my naïve thoughts on the Convention (although I had no credentials to justify his interest), and generously shared his sophisticated insights. My good fortune continued, for I had several more opportunities to work with Professor Schlechtriem, and to benefit not only from the workings of his remarkable mind but also from his benevolent spirit.
I believe that the rich output of scholarship that has focused on the CISG and related topics is due, in significant part, to the fact that the community of scholars interested in the topic is so supportive, friendly, and collegial (as well as, of course, extraordinarily bright and articulate); simply put, it is fun to be around this group. I am convinced that this is due largely to the fact that this community was established by people like Professor Schlechtriem, who embodied these qualities, and who managed to be both great and good. Professor Schlechtriem's ideas and his character permeate the world in which I have been lucky enough to spend much of my professional life. The most fitting memorial to him will be the continuation of this scholarly community. [page 91]
While it is an extraordinary honour to be involved in this project honouring Professor Schlechtriem, that has not stopped me from -- and I choose the word advisedly -- cheating in my choice of topics. In this paper I address an extremely broad and important subject: the scope of the CISG -- i.e., what issues does the CISG govern when it applies to a transaction? I have no intention, however, to attempt a comprehensive treatment of the topic; rather this paper ranges over a motley collection of issues, with only the vaguest suggestion of a unifying theme.
I will address generally limits to the Convention's scope of application, distinguishing between express scope limitations (found in Arts. 4 & 5 CISG) and what I (rather inarticulately) call 'less articulated' limits. I will also discuss two specific scope issues that have appeared in decisions applying the CISG: 1) whether requirements imposed by U.S. domestic sales law on attempts to disclaim 'implied warranties' apply to attempts to derogate from the seller's obligations under Arts. 35(2)(a) & (b) CISG; and 2) whether burden of proof issues, arising under the CISG that are not expressly addressed therein, are governed by the general principles of the CISG. As part of this last topic, I will defend the use of the distinction between substantive law and procedural law in defining the scope of the CISG; I argue that the distinction can be appropriately used to address the burden of proof question, as well as another scope issue that I have analysed previously (and which I therefore do not address at length here) -- whether the CISG provides for the recovery of damages to cover the costs of attorneys employed to litigate a claim under the CISG. I conclude with an undoubtedly overblown (but naively sincere) argument that properly recognising the limits of the Convention's scope is critical to its success, and to the success of future attempts to create uniform international commercial law.
2. EXPRESS SCOPE LIMITATIONS: THE 'VALIDITY' EXCLUSION
Article 4 CISG provides, 'except as otherwise expressly provided in this Convention, it is not concerned with: (a) the validity of the contract or of any of its provisions or of any usage' (emphasis added). Thus, contrary to the general impression of its scope, the Convention in fact governs questions of validity -- but only if a provision expressly addresses the question. For example, Art. 11 CISG eliminates form requirements for contracts governed by the Convention; unless the transaction involves a State that has [page 92] made an Art. 96 CISG reservation (see Art. 12 CISG), domestic rules requiring that a contract be in writing in order to be valid, are pre-empted by the CISG.
The effect of the validity exclusion in Art. 4(a) CISG, therefore, is not to eliminate validity issues from the reach of the Convention, but to take such issues out of the gap-filling procedures in Art. 7(2) CISG. That subsection provides:
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 the rules of private international law.
This two-step gap-filling process (requiring, first, reference to the CISG general principles, and only in the absence of relevant general principles permitting reference to non-uniform law applicable under rules of private international law ('PIL')  applies only if a question is 'governed by this Convention' but is 'not expressly settled in it'; in other words, Art. 7(2) CISG applies only to so-called 'internal gaps' in the CISG. Under Art. 4(a) CISG however, 'validity' issues are governed by the Convention only if they are expressly addressed in it. Consequently, gaps concerning questions of validity are beyond the scope of the CISG: they constitute so-called 'external gaps that must be resolved by direct reference to the non-Convention law applicable under PIL rules, and are not to be settled by reference to the Convention's general principles. Unless a question of validity is expressly addressed in the Convention, therefore, it continues to be governed by applicable domestic law.
Contrast this approach with that in Art. 5 CISG, which provides: 'This Convention does not apply to liability of the seller for death or personal injury caused by the goods to any person.' There is no exception here for express provisions of the Convention, no suggestion that particular Convention rules may in fact expressly address issues relating to a seller's liability for death or personal injury. Article 5 CISG goes beyond Art. 4 CISG in removing issues from the scope of the Convention: it does not merely indicate that the issues it identifies must be treated as 'external gaps' to the extent the CISG does not expressly provide for them; rather, it states that express terms of the [page 93] CISG should not be interpreted to address a seller's liability for death or personal injury.
Although validity issues in CISG transactions are, except where expressly provided for, subject to applicable domestic law, the question of what constitutes a matter of 'validity' that is excluded from the scope of the Convention by Art. 4(a) is governed by the CISG itself. Concerning the meaning of the term 'validity' under the CISG, Professor Schlechtriem opined:
if a contract is rendered void ab initio, either retroactively by a legal act of the state or of the parties such as avoidance for mistake or revocation of one's consent under special provisions protecting certain persons such as consumers, or by a 'resolutive' condition (i.e., a condition subsequent) or a denial of approval of relevant authorities, the respective rule or provision is a rule that goes to validity and therefore is governed by domestic law and not by the CISG.
A US court, adapting a test proposed by Professor Hartnell, stated that 'validity' encompasses 'any issue by which the domestic law would render the contract void, voidable or unenforceable'.
The complex relationship between the Convention's validity exclusion and applicable domestic law creates a challenging analytical situation. Indeed, these challenges recently produced a decision by a US Federal District (trial) Court in Norfolk Southern Railway Company v. Power Source Supply, Inc. ('Norfolk and Western') that managed to be both plausible and absurd. The case involved a sale of used locomotives by a US railroad corporation, located in the state of Pennsylvania, to a Canadian locomotive broker.
The buyer sent an initial purchase order to the seller, followed by a series of 'revised' purchase orders incorporating changed sales terms. Upon receipt of the 'final' revised [page 94] purchase order, the seller transmitted and the buyer executed a final bill of sale for the locomotives.
The bill of sale included the following language not found in the buyer's purchase orders:
THE EQUIPMENT [is] BEING SOLD ON AN 'AS, WHERE IS' BASIS AND WITH ALL FAULTS. EXCEPT AS SET FORTH HEREIN, THE SELLER MAKES NO REPRESENTATION OR WARRANTY, EITHER EXPRESS OR IMPLIED, AS TO THE CONDITION OF THE EQUIPMENT, INCLUDING WITHOUT LIMITATION ANY IMPLIED WARRANTY OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. [...].
The transaction was governed by the CISG, because the parties were located in different Contracting States and the parties had not agreed to opt out of the CISG. Applying Art. 19 CISG, the court held that the seller's bill of sale (which included the language set out above) constituted a counter-offer that the buyer had accepted. An issue before the court was whether this language excluded the seller's implied obligations under Art. 35(2) CISG concerning conformity of goods.
Citing Art. 4(a) CISG, the court first noted that the validity of the disclaimer language was not governed by the CISG but by applicable domestic (non-uniform) law. It then determined that the language in the seller's bill of sale was sufficient, under the laws of both Canada and Pennsylvania, to exclude the implied quality obligations provided for in Art. 35(2) CISG.
In analysing the seller's disclaimer the court assumed that, under Pennsylvania law (which in this regard is the same in 49 of the 50 US states), the validity of a contract clause excluding a seller's obligations under Art. 35(2)(a) CISG is governed by the rules that address disclaimers of the implied warranty of merchantability provided for in the (US) Uniform Commercial Code ('UCC'). Under § 2-314 UCC an implied warranty of merchantability arises automatically (i.e., without affirmative agreement between the parties) when the seller is a merchant in goods of the kind. The implied warranty of merchantability requires the seller to deliver goods that, inter alia, 'are fit for the ordinary purposes for which goods of that description are used'.
This rule is certainly similar to Art. 35(2)(a) CISG which, unless the parties agree otherwise, requires the seller to deliver goods 'fit for the purposes for which goods of the same description would ordinarily be used'. Thus the analogy that the court appears to have assumed between the implied warranty of merchantability in U.S. domestic sales law and the seller's implied obligation under Art. 35(2)(a) CISG, as well as its finding that a disclaimer addressed to the issue of 'merchantability' is effective to exclude the seller's liability under Art. 35(2)(a) CISG, are both plausible. [page 95]
Under § 2-316(2) UCC, a written clause that attempts to disclaim the implied warranty of merchantability must be 'conspicuous' and to be effective include the word 'merchantability'. The court's analysis assumed that this is a rule of 'validity' which, if U.S. law applied under PIL rules, would have to be met before a contract provision could effectively exclude the seller's obligation under Art. 35(2)(a) CISG. Once again, the court's assumption that § 2-316(2)(a) UCC states a rule of 'validity' seems plausible enough: the rule in § 2-316(2) UCC is mandatory (the parties cannot contract out of its requirements), it is designed to achieve a public policy purpose (to ensure that buyers are fairly apprised that they are giving up an important protection like the implied warranty of merchantability) and it renders a disclaimer that fails to meet its requirements void ab initio.
The Norfolk & Western court's analysis however, leads to a clearly absurd result: under its reasoning, if PIL rules lead to the application of U.S. law, a clause that attempts to derogate from Art. 35(2)(a) CISG is not valid, unless it uses terminology ('merchantability') never employed in the CISG. Why would a competent drafter who was attempting to exclude the obligation imposed by Art. 35(2)(a) CISG -- and who was trying to convey that information to the buyer in the most unmistakable and straightforward way -- ever choose to state that the goods are not warranted to be 'merchantable', as required by § 2-316(2) UCC?
A well-drafted clause designed to exclude the seller's obligations under Art. 35(2)(a) CISG might read along the lines of:
The parties agree that the goods conform to the contract even if they are not fit for the purposes for which goods of the same description would ordinarily be used. The parties intend to derogate from Art. 35(2)(a) CISG.
Such a clause, however, would be invalid under the reasoning of the Norfolk & Western court -- no matter how clear, straightforward and actually-communicative it is -- because it does not employ the term 'merchantability.' Imposing the § 2-316(2) UCC validity requirements in transactions governed by the CISG, as the Norfolk & Western court does, may well have the opposite effect of that intended by those disclaimer rules: by employing the term 'merchantability', which is never used in the CISG, in a disclaimer clause would obscure rather than clarify the intent of the parties, and would darken rather than illuminate the legal effect of the clause. That, to state the obvious, is perverse. [page 96]
Of course the clause before the court in the Norfolk & Western case met the requirements of § 2-316(2) UCC; perhaps, if faced with having to declare a clearly-drafted derogation from Art. 35(2)(a) CISG invalid because it did not use the term 'merchantability,' this court (or some other) would sensibly back away from the precipice.
At any rate, the clause in the case complied with § 2-316(2) UCC because, it seems fair to presume, it was drafted for contracts governed by the UCC and not the CISG. Indeed, it clearly was not competently drafted for a CISG contract: it uses terminology like 'warranty' and 'merchantability' never employed in the CISG, and it fails to invoke proper CISG language or relevant CISG provisions. Although I believe the clause should be deemed effective to exclude the seller's Art. 35(2)(a) CISG obligation, that conclusion requires an inference -- that the term 'warranty of merchantability' in the clause should be understood as the obligation imposed by Art. 35(2)(a) CISG. This inference, however, would probably not be accepted by all. Where the CISG governs a transaction, no competent drafter, whether from the U.S. or elsewhere, is likely to meet the validity requirements imposed by the Norfolk & Western decision unless he or she is specifically aware of the decision. The case creates a trap for the unwary drafter with no compensating benefit.
Exactly where did the court's analysis go wrong? Its premise that applicable domestic law governs the validity of a clause disclaiming the seller's Art. 35(2)(a) CISG obligations is unassailable given Art. 4(a) CISG. And, as discussed above, the requirement in § 2-316(2) UCC that a disclaimer of the implied warranty of merchantability must use the term 'merchantability' appears to state a rule of 'validity' within the (autonomous) meaning of Art. 4(a) CISG. Indeed, in my view there is nothing wrong with the court's analysis (or presumed analysis) of the CISG. It is its application of U.S. domestic law that is, ironically, incorrect.
It is clear beyond dispute, in my view, that the merchantability disclaimer rule in § 2-316(2) UCC was intended to apply only to attempts to disclaim the implied warranty of merchantability provided for in § 2-314 UCC. That is why it requires the specific terminology employed in the UCC provision. That is also why it is inapplicable to other implied quality obligations such as that imposed in Art. 35(2)(a) CISG, even if those other obligations are similar to the implied warranty of merchantability. In short, the reason the Norfolk & Western court is wrong in finding that contract provisions derogating from Art. 35(2)(a) CISG must meet the requirements of § 2-316(2) UCC is because the UCC requirements are simply inapplicable as a matter of U.S. domestic law: those requirements apply only in transactions governed by Art. 2 UCC, not in transactions governed by the CISG. [page 97]
To be as clear as possible, I will give an example of a U.S. domestic validity provision that is addressed to disclaimers of a seller's quality obligations and that I believe is applicable to attempts to derogate from certain parts of Art. 35(2) CISG. Section 108 of the Magnuson Moss Warranty Act  ('the Magnuson Moss Act') -- a US federal consumer protection statute -- provides that a seller who provides a 'written warranty' on a 'consumer product' cannot completely disclaim 'implied warranties' on the product (although it can limit their duration to that of a reasonable written warranty); disclaimers that violate this rule are 'ineffective.' This provision prevents sellers who provide typical written consumer warranties (e.g., warranting against defects for a specified period) from completely eliminating the UCC implied warranty of merchantability, as well as the implied warranty of fitness for particular purpose provided for in § 2-315 UCC. Section 108 Magnuson Moss Act is a mandatory rule designed to promote public policy relating to consumer protection, and constitutes in my view a rule of 'validity' under Art. 4(a) CISG.
If PIL rules point to the application of U.S. law, I believe that § 108 Magnuson Moss Act prevents a seller who makes a 'written warranty' concerning a 'consumer product' from eliminating certain implied obligations under Art. 35(2) CISG. Those obligations include the Art. 35(2)(a) CISG obligation to furnish goods suitable for their ordinary purposes -- which, as noted above, is analogous to the UCC implied warranty of merchantability -- and the obligation under Art. 35(2)(b) CISG to furnish goods suitable for any particular purpose of the buyer that was known to the seller at the time the contract was concluded -- which is analogous to the UCC implied warranty of fitness for particular purpose.
Unlike the disclaimer rule of § 2-316(2) UCC, § 108 Magnuson Moss Act does not mandate the use of UCC-specific terminology in contracts, and does not appear to be limited to UCC warranties. Applying § 108 Magnuson Moss Act to a seller's implied quality obligations under Art. 35(2) CISG, furthermore, would advance rather than impede the purposes of the Act.
Of course, Art. 2(a) CISG provides that the Convention does not apply to consumer sales, which means that most transactions governed by the CISG would not be subject to the requirements of the Magnuson Moss Act. The CISG's exclusion of consumer sales, however, is triggered by the particular buyer's intended use of the goods -- Art. 2(a) CISG provides that the CISG does not apply if the goods are 'bought for [page 98] personal, family or household use'. In contrast to the CISG, the definition of 'consumer product' in § 101 Magnuson Moss Act applies to sales of goods that are 'normally' used for consumer purposes (i.e., products for which such use is not uncommon'), irrespective of the particular buyer's intended use. Thus the CISG would apply to a business's purchase of a car for use by a sales representative (provided the requirements of Art. 1 CISG are met and the CISG is not excluded), even though the transaction would be considered the sale of a 'consumer product' subject to the Magnuson Moss Act.
In my view, § 108 Magnuson Moss Act would limit the power of the parties to derogate from Art. 35(2)(a) CISG or Art. 35(2)(b) CISG in such a transaction, provided the seller gave the buyer a written quality assurance meeting the definition of 'written warranty' in § 101 (6) Magnuson Moss Act.
3. 'LESS-ARTICULATED' SCOPE LIMITATIONS
Article 4(a) and (b) CISG specifies particular subject areas that (except where expressly addressed) are beyond the reach of the Convention -- i.e., validity and property issues. The first sentence of Art. 4 CISG, in contrast, contains a general statement describing matters not governed by the CISG, phrased in terms of an affirmative description of the Convention's scope: 'This Convention governs only the formation of the contract of sale and the rights and obligations of the seller and the buyer arising from such a contract.' Issues not encompassed by the phrase 'the formation of the contract of sale and the rights and obligations of the seller and the buyer arising from such a contract' are declared to be beyond the reach of the Convention. The sentence is so general however, that it is less a rule for determining the reach of the CISG than a nebulous expression of the drafters' intent. Thus, many of the boundaries that distinguish matters governed by the CISG from those that are beyond its scope are vaguely-suggested rather than precisely-described. Identifying these limits on the scope of the CISG involves a difficult inquiry into its intended reach with limited guidance from its text.
Based on the first sentence of Art. 4 CISG, the travaux préparatoires of the Convention, the nature of its provisions, and the views of informed commentators, it is clear that the drafters of the CISG (as well as, presumably, the States that have ratified it) understood that they were addressing substantive sales law. This understanding is, I believe, indispensible for resolving scope questions. For example, it is generally assumed that the Convention's focus on substantive sales law means that it is not [page 99] addressed to questions of delict or tort liability. There are, however, varying understandings of what is encompassed by 'tort' or 'delict,' and of course the distinction between 'tort/delict' and 'contract' is not expressly made in the Convention. This situation can make it extremely difficult to determine whether a particular issue that is addressed in some (or even most) domestic laws by doctrines labelled 'tort' or 'delict' was intended to be governed by the Convention. Hence, the 'contract vs. tort/delict' distinction can be nothing more than a rough guide or 'factor' in determining the scope of the CISG, not an authoritative rule. Nevertheless, the guidance the distinction offers can be critical to determining the reach of the Convention.
The same can be said concerning the distinction between 'substance' and 'procedure'. It is clear beyond dispute that the Convention focuses on substantive sales rules and does not attempt to create a comprehensive procedural system or even a substantial set of procedural rules. There is nothing in the Convention to address a myriad of procedural questions, from the jurisdiction of fora to rules for conducting hearings to the vast majority of evidentiary issues, and many others.
That is not to say however, that the Convention contains nothing relating to issues that might be labelled 'procedural.' Indeed, Art. 11 CISG quite clearly steps into the procedural realm by providing a particular evidentiary rule: a contract governed by the Convention 'can be proved by any means, including witnesses'. This provision, however, does not mean that evidentiary questions or procedural issues in general are matters governed by the CISG, any more than the fact that the CISG expressly [page 100] addresses a question of 'validity' (i.e., the question of formal validity covered by Art. 11 CISG) means that validity issues in general are within the Convention's scope.
Just as with the 'contract vs. tort / delict' distinction, the distinction between substance and procedure is not expressed in the Convention; thus it is not a rule defining the scope of the Convention, but rather a guide or consideration that assists in drawing the boundaries of the CISG. Nevertheless, it is important to take cognisance of this distinction -- for one reason, as discussed below, because the drafters themselves utilised it.
Professor Joseph Lookofsky and I have used the substance/procedure distinction to argue that Art 74 CISG does not permit the recovery of damages for attorney fees incurred by a prevailing claimant in litigating a claim. I do not intend to revisit that issue, concerning which I have already thoroughly expressed myself (the phrase ad nauseam will occur to some readers), but I do wish to address an issue of methodology that has arisen from the attorney fees controversy. Use of the substance/procedure distinction to help resolve this issue -- an approach Professor Lookofsky and I have employed -- has been criticised, in some cases by those who agree with our ultimate conclusion that the Convention does not sanction the recovery of damages to cover attorney fees incurred in litigating the claim. I continue to believe the criticism to be wrong. [page 101]
Since the Convention does not expressly adopt the substance/procedure distinction in defining its scope, the 'procedural' label should not be employed as an authoritative legal test in the manner of the express exclusion of 'validity' issues. Moreover, it would most certainly be wrong -- contrary to the mandates of Art. 7(1) CISG -- to conclude that a matter is beyond the Convention's reach merely because it is labelled a 'procedural' issue in domestic law. But it most surely is not wrong to use the substance/procedure distinction which was employed by the drafters themselves, as a tool to help analyse the intended scope of the CISG. Indeed, the suggestion that one should not use the distinction because it is difficult or 'outmoded' elevates the judgment of commentators over the ideas of those who actually drafted the Convention and (by extension) the intent of the Contracting States.
In this paper I will again draw on the substance/procedure distinction, this time to address a scope issue concerning which I know my position is in the minority, opposed by many of the best minds that have analysed the Convention. The question I will address is whether the CISG itself governs which party bears the burden of proving elements required by Convention provisions. Warren Khoo concluded that the matter was beyond the scope of the CISG. This conclusion was based on evidence from the travaux preparatoires that the drafters (of whom Mr. Khoo was a prominent member) did not intend, in general, to deal with burden of proof issues:
Delegations speaking on the burden of proof [at the Vienna Diplomatic Conference] were all quite definite that it was not the intention to deal in the Convention with any questions concerning the burden of proof. The consensus was that such questions must be left to the court as matters of procedural law.
In addition, the drafters of the CISG rejected a proposal to add language expressly allocating the burden of proof to one provision, because, in the words of an UNCITRAL report, 'it was considered inappropriate for the Convention, which relates to the international sale of goods, to deal with matters of evidence or procedure.' [page 102]
There is also limited case law support for referring to domestic law to determine burden of proof questions.
On the other hand, more recent commentators -- including some of the most eminent and respected CISG authorities -- argue that burden of proof issues are governed by the Convention and that where the CISG fails to allocate the burden of proof expressly, there is an 'internal gap' that should be resolved pursuant to the two-step methodology of Art. 7(2) CISG. The argument is that the CISG contains at least one express allocation of the burden of proof, demonstrating that the matter is within the scope of the CISG and that the 'rule and exception' approach to burden of proof constitutes a general CISG principle applicable to cases not covered by such an express allocation. Substantial case law, particularly more recent case law supports this approach.
The fact that the CISG expressly addresses the issue of burden of proof in one instance (or, in my view perhaps two ) does not establish that it governs that issue generally -- just as the fact that Art. 11 CISG includes a validity rule in Article 11 does not mean that the CISG generally governs validity issues. The drafting history cited above certainly suggests that the drafters considered burden of proof issues beyond the Convention's ken -- a perception that tends to be confirmed by the extremely small number of CISG provisions that expressly address burden of proof questions. That latter fact, furthermore, means that there is a very small data set from which to derive [page 103] general principles for the overwhelming number of burden of proof issues that are not expressly addressed in the CISG.
Even authorities that support the view that the CISG governs burden of proof questions suggest there are significant limits to this aspect of the Convention. Two decisions of the German Bundesgerichtshof illustrate this point. In both, the court asserted that burden of proof questions are within the scope of the Convention; in one however, the court ended up applying domestic burden of proof rules because it could not find adequate resources within the Convention to deal with the specific issue; in the other, the court was forced to discover a general CISG principle on burden of proof for which it could cite scant express support in the CISG, and which (coincidentally?) corresponded to domestic German doctrine that would not necessarily work well in other systems.
In the first of those decisions the Bundesgerichtshof dealt with the burden of proof consequences of a party's admission of liability. The court commented:
[T]he CISG regulates the burden of proof explicitly (e.g., in Art. 79(1)) or tacitly (Art. 2(a)), so that consequently, recourse to the national law is blocked to that extent, and [...] the CISG follows the rule/exception principle. [... T]he burden of proof rules of the CISG [however,] cannot go farther than the scope of its substantive applicability. That scope results from Art. 4(1) CISG; according to that provision, the CISG regulates exclusively the execution of the sales contract and the duties and responsibilities of the buyer and the seller resulting from that contract. The question whether and possibly which evidentiary consequences an actual admission of liability has, is not part of that scope. That question -- just like the meaning of a defective mens rea, an assignment, a set-off, or similar issues -- does not implicate a specific sales-law-related problem, but rather a legal aspect of a general type; there is no intimate relationship to the actual or legal aspects of the international trade in goods, which make up the regulatory subject of the CISG.
Thus, the court adopts a complex approach under which some burden of proof issues are governed by the Convention, others are subject to national law, and the test (or tests?) for distinguishing the two is remarkably challenging to apply: one must determine whether the issue before the decision-maker 'does not implicate a specific sales-law-related problem, but rather a legal aspect of a general type', and whether 'there is no intimate relationship to the actual or legal aspects of the international trade in goods, which make up the regulatory subject of the CISG'. I have my doubts that [page 104] this result adds to the uniformity of international sales law, or to the cause of predictability and consistency in applying the Convention. A simple rule that burden of proof issues are beyond the scope of the Convention except when expressly addressed, might do a better job on both fronts.
In the second decision, the Bundesgerichtshof reaffirmed that the Convention itself addresses burden of proof issues and that it adopts a 'rule and exception' approach. The court, however, reversed the result dictated by that principle by invoking the 'proof-proximity [Beweisnähe]' principle, which relieves a party of an evidentiary burden it would normally bear if the facts at issue 'are exclusively in [the other] party's sphere of responsibility and [ ... ] therefore are, at least theoretically, better known to that party [ ... ]'. The court seems to have applied the Beweisnähe principle as a general principle of the CISG pursuant to Art. 7(2) CISG, although the approach appears to have roots in German domestic law.
The Beweisnähe principle (or exception) for allocating burden of proof, I am sure, makes a great deal of sense within the German court system and, undoubtedly, many others as well. Indeed, the main reason offered by the Bundesgerichtshof in support of finding Beweisnähe among the general principles of the Convention is that it is a 'necessary' corrective to inequities created by the usual burden of proof rule attributed to the CISG's general principles -- the rule-and-exception approach. The Beweisnähe principle, however, makes far less sense in U.S. courts. Under U.S. procedure, if a party who does not bear the burden of proof has superior access to evidence, the matter is generally addressed through a pre-trial discovery process, rather than through reversing the burden. According to the Bundesgerichtshof, however, Beweisnähe is a general CISG principle that must be applied in quite different procedural systems even where the principle is inappropriate.
As the foregoing discussion illustrates, burden of proof rules are often 'embedded' in particular procedural systems and tend to reflect the features of those systems. I know of no evidence that the drafters of the CISG studied or even considered what would be sensible burden of proof principles for the myriad of diverse court systems and arbitral tribunals that exist around the world. The drafters were too busy attempting to come up with workable and acceptable substantive rules that could be interpreted and applied in various procedural environments. To expand the Convention so it reaches burden of proof issues not only violates the drafters' apparent intent, but also disserves the reputation of the CISG by making it appear unsophisticated and inadequate. This is illustrated by the first of the Bundesgerichtshof decisions discussed above, where the court was forced to resort to domestic law for an adequate solution for the burden of proof issue it confronted. [page 105]
The foregoing discussion of the burden of proof question demonstrates how the substance/procedure distinction can -- and should -- be properly used as a tool in determining whether an issue is governed by the CISG. Since the CISG does not expressly provide that matters of procedure are beyond its scope, the question is not whether a particular issue should be characterized as a question of 'procedure' or 'substance' -- a distinction notoriously difficult to make.
Contrast the approach when dealing with an express exclusion from the scope of the CISG, such as the 'validity' exclusion in Art. 4(a) CISG, where the question precisely is whether a matter involves an issue of 'validity' (another notoriously difficult characterisation). Nevertheless, the substance/procedure distinction is useful in clarifying whether the CISG was intended to govern a particular issue.
With regard to burden of proof issues, several facts that have nothing to do with the substance-procedure distinction raise doubts as to whether the CISG governs burden of proof generally -- statements in the travaux suggesting the drafters did not intend to address burden of proof, as well as the extremely small number of CISG provisions (one, perhaps two) that address the issue expressly. The fact that burden of proof is arguably, a 'procedural' matter (burden of proof, like evidentiary rules, is closely associated with the process for determining facts, and burden of proof rules are influenced by differences in procedural systems) suggests an explanation for why the drafters of a convention aimed primarily at substantive issues would not deal with burden of proof. This, in turn, tends to confirm the view that burden of proof is, except where expressly addressed, not within the Convention's scope.
Employing the substance/procedure distinction in the manner suggested by the foregoing discussion is both valid and valuable. As noted earlier, the drafters themselves made use of the distinction in connection with the burden of proof issue.
Similarly, I believe that Professor Lookofsky and I were justified in using the substance/procedure distinction to explain why (and thus to confirm that) the drafters did not address the recovery of litigation attorney fees in the CISG -- i.e., because it probably did not occur to the drafters of the substance-oriented CISG to address an issue that is generally treated as a procedural matter.
I have discussed two specific scope issues in this paper:
1. whether the requirements for effective disclaimers of implied warranties in § 2-316(2) UCC constitute rules of 'validity' applicable to contract provisions derogating from Art. 35(2)(a) or (b) CISG; and [page 106]
2. whether burden of proof is a matter 'governed by' the Convention, so that burden of proof questions not expressly addressed in the CISG are to be resolved by reference to its general principles pursuant to Art. 7(2) CISG.
I care very much that the first issue be resolved along the lines I suggest i.e., that the § 2-316(2) UCC requirements be found inapplicable to contract provisions attempting to exclude a seller's obligations under Art. 35(2)(a) or (b) CISG. Incorrect resolution of this issue is likely to have serious and undesirable practical consequences, and to harm the cause of international uniform sales law.
I care far less passionately whether my views on burden of proof (i.e., that the matter is generally beyond the scope of the Convention) prevail. The 'rule and exception' approach which, proponents posit, is the primary burden of proof rule embodied in the Convention's general principles is likely also to be found in the burden of proof rules of most domestic laws; as a consequence, in most cases it probably matters little whether my views are accepted or rejected. Were I a judge or arbitrator facing the burden of proof issue I might well follow the 'agnostic' position adopted in some decisions, and refuse to rule whether the Convention governs burden of proof since the result would likely be the same under domestic law. Given my position's lack of practical impact and the substantial authority arrayed against me, as a judge or arbitrator I might be willing to abandon my position in the name of promoting uniformity in interpreting the CISG.
I nevertheless inflict on you this partially-academic exercise in defining the Convention's scope because I believe it is critical to develop sound approaches to scope questions. Applying the Convention to issues for which it was not designed threatens undesirable results, and it undermines the delicate political consensus that made the CISG possible and that will be needed for future attempts to create uniform commercial law. Take, as an example, the question whether Art. 74 CISG provides for the recovery of attorney fees incurred in litigating a claim under the CISG (a dead horse that I earlier said I would not again beat -- but (to mix metaphors) I am drawn like a moth to the flame). Article 74 CISG quite clearly is not designed for this purpose; it contains none of the features that loser-pays systems have often found advisable (such as determining the amount of recoverable attorney fees by general formula rather than by reference to the amount of fees actually incurred); furthermore, it requires absurd manipulations in order to avoid the untenable result that, since recovery of attorney fees would be allowed as damages for breach, only successful claimants -- not those prevailing in defending against a claim -- could recover attorney costs. [page 107]
Imposing such a poor rule, never intended for this purpose on the United States, which generally requires each side to bear its own attorney costs, hardly advances the cause of a loser-pays approach. It only tends to undermine the authority of the Convention, and strengthens the hand of those who argue against the wisdom of uniform international law initiatives.
It has been suggested that doubts regarding the Convention's reach should be resolved in favour of applying uniform international law  ('favor conventionis'), although caution in the use of this approach has also -- very sensibly -- been urged. Particular caution is required when the question involves the scope of the Convention (determining which issues the CISG governs when it applies to a transaction) as opposed to its sphere of application (determining the transactions to which it applies). Extending the Convention's reach to issues that the drafters did not consider part of the CISG project, and for which there was, as a consequence, no opportunity for proper debate, impedes rather than advances its purposes. [page 108]
* Professor of Law, University of Pittsburgh School of Law. A.B. 1973, Harvard College; M.A. 1975, Harvard University; J.D. 1981, Harvard Law School.
1. For a comprehensive survey of specific matters beyond the scope of the CISG, see Ferrari, F., "Scope of application: Articles 4-5" in Ferrari, F., Flechtner, H. and Brand, R.A. (eds), The Draft UNCITRAL Digest and Beyond: Cases, Analyses and Unresolved Issues in the U.N. Sales Convention, 2004, Sellier at pp. 106-113.
2. See infra fn 23.
3. Art. 4(b) CISG states that, except as otherwise expressly provided, the Convention is not concerned with 'the effect which the contract may have on the property in the goods sold.'
4. For discussion of this point see Ferrari, F., "Scope of application," supra fn 1, at pp. 99-100.
5. Art. 12 CISG states that, 'Any provision of article 11, article 29 or Part II of this Convention that allows a contract of sale or its modification or termination by agreement or any offer, acceptance or other indication of intention to be made in any form other than in writing does not apply where any party has his place of business in a Contracting State which has made a declaration under article 96 of this Convention. The parties may not derogate from or vary the effect or this article'.
6. See <http://cisgw3.law.pace.edu/cisg/biblio/ho90.html> - 61
7. See Honnold, J., Uniform Law for International Sales under the 1980 United Nations Convention, 4th ed, edited and updated by Flechtner, H.M., forthcoming 2009, Kluwer, § 64. These conclusions also apply to issues relating to 'the effect which the contract may have on the property in the goods sold,' which are expressly excluded by Art. 4(b) CISG from the scope of the CISG, unless expressly addressed therein.
8. See Honnold, J., Uniform Law for International Sales, supra fn 6, § 65; Schlechtriem, P., "Art. 4" ¶ 7 in Schlechtriem, P. and Schwenzer, I. (eds), Commentary on the UN Convention on the International Sale of Goods (CISC), 2nd (English) ed, 2005 ('What amounts to invalidity has to be analysed, however, "autonomously," i.e. as a concept of the CISG interpreted according to the guidelines of Art. 7(1). In order to preserve or achieve a uniform application of the term "validity," it is not the words used in domestic law and their interpretation under domestic law, but the functions of the respective rules and provisions that are decisive (footnote omitted)); Ferrari, F., "Scope of application," supra fn 1, at p. 100.
9. Schlechtriem, P., "Art. 4" ¶ 7, in Commentary on the UN Convention, supra fn 7.
10. Geneva Pharmaceuticals Tech. Corp. v. Barr Labs. Inc., 201 F. Supp. 2d at pp. 236, 282-83 (US District Court for the Southern District of New York, 10 May 2002, available at: <http:// cisgw3.law.pace.edu/cases/020510u1.html>), quoting Hartnell, H., "Rousing the Sleeping Dog: The Validity Exception to the Convention on Contracts for the International Sale of Goods" (1993) 18 Yale Journal of International Law 1, at p. 45.
11. 2008 WL 2884102: US District Court for the Western District of Pennsylvania, 25 July 2008, available at: <http://cisgw3.law.pace.edu/cases/080725u1.html>.
12. Under § 2-316(3)(a) UCC, the phrase 'as is' (or equivalent language) eliminates implied warranties, including the implied warranty of merchantability, even if the requirements for an effective disclaimer under § 2-316(2) UCC are not met. The Norfolk & Western court ignored § 2-316(3)(a) UCC, and the 'as is' language in the seller's disclaimer. Because discussion of this rule would not impact the points made herein, this paper will do the same.
13. See Honnold, J., Uniform Law for International Sales, supra fn 6, § 230.
14. 15 U.S.C. § 2308.
15. This is defined in § 101(6) Magnuson Moss Act.
16. This is defined in § 101(1) Magnuson Moss Act.
17. Section 108 Magnuson Moss Act limits a seller's right to disclaim an 'implied warranty,' which is defined in § 101(7) of the Act as 'an implied warranty arising under State Law.' The meaning of the phrase 'State Law' is addressed in § 101(15) Magnuson Moss Act, but it has repeatedly been held that the CISG is part of the law of each state of the United States. Thus, the phrase 'implied warranty' in the Magnuson Moss Act, I would argue, encompasses the obligations provided for in Art. 35(2)(a) & (b) CISG.
18. 16 CFR § 700.1(a), a regulation promulgated by the Federal Trade Commission to implement the provisions of the Magnuson-Moss Act.
19. For an argument that the first sentence of Art. 4 CISG both overstates and understates the actual reach of the CISG see Ferrari, F., "Scope of application," supra fn 1, at p. 97.
20. See Ferrari, F., "Scope of application," supra fn 1, at pp. 103-105; Honnold, J., Uniform Law for International Sales, supra fn 6, § 70.1; Schlechtriem, P., "Art. 4" ¶¶ 5 & 23a, in Commentary on the UN Convention, supra fn 7. For a very recent US case holding that tort claims are beyond the scope of the CISG see Sky Cast, Inc v. Global Direct Distribution, LLC, 2008 WL 754734 (US District Court for the Eastern District of Kentucky, 18 March 2008), available at: <http://cisgw3.law.pace.edu/cases/080318u1.html>.
21. See Honnold, J., Uniform Law for International Sales, supra fn 6, §§ 71-73; Ferrari, F., "The Interaction between the United Nations Convention on Contracts for the International Sale of Goods and Domestic Remedies (Rescission for Mistake and Remedies in Torts)" (2007) 71 Rabels Zeitschrift für ausländisches und Internationales Privatrecht 52; Lookofsky, J., Understanding the CISC in the USA: A Compact Guide to the 1980 United Nations Convention on Contracts for the International Sale of Goods, 2nd ed., 2004, § 2.6; Lookofsky, J., CISC Case Commentary on Preemption in Geneva Pharmaceuticals and Stawski, April 2004, available at: <http://www.cisg.law.pace.edu/cisg/biblio/lookofsky8.html>; Lookofsky, J., "In Dubio Pro Conventione? Some Thoughts about Opt-outs, Computer Programs and Preemption under the 1980 Vienna Sales Convention (CISG)" (2003) 13 Duke Journal of Comparative and International Law 263, also available at: <http://law.duke.edu/journals/djcil/articles/djcill3p0263.htm>; Lookofsky, J.M., "Loose Ends and Contorts in International Sales: Problems in the Harmonization of Private Law Rules" (1991) 39 American Journal of Comparative Law 403.
22. For statements that the Convention does not, in general, govern procedural issues, see, e.g., Schlechtriem, P. and Schwenzer I., "Art. 4" ¶¶ 19 and 22, in Commentary on the UN Convention, supra fn 7; Ferrari, F., "Scope of application," supra fn 1, at p. 107; cases cited in UNCITRAL Digest of CISG Article 4 case law ¶ 13 nn. 41 & 42, available at: <http://www.uncitral.org/uncitral/en/case_law/digests/cisg.html>.
23. See the text accompanying fns 23-25 infra.
24. See Lookofsky, J. and Flechtner, H., "Viva Zapata! American Procedure and CISG Substance in a U.S. Circuit Court of Appeal" (2003) 7 Vindobona Journal of International Commercial Law and Arbitration 93 (Vindobona Journal); Lookofsky, J. and Flechtner, H., "Zapata Retold: Attorneys' Fees Are (Still) Not Governed by the CISG" (2006-07) 26 Journal of Law and Commerce 1. Lookofsky, J., "Zapata Hermanos v. Hearthside Baking [case commentary]" (2002) 6 Vindobona Journal 27; Flechtner, H.M., "Recovering Attorneys' Fees as Damages under the U.N. Sales Convention (CISG): The Role of Case Law in the New International Commercial Practice, with Comments on Zapata Hermanos v. Hearthside Baking" (2002) 22 Northwestern Journal of International Law and Business 121, at pp. 126-133. For other authority supporting this result (but not necessarily employing the substance/procedure distinction) see, Stoll, H. and Gruber, G., "Art. 74" ¶ 20 in Commentary on the UN Convention, supra fn 7; CISG Advisory Council Opinion No. 6, Calculation of Damages under CISG Article 74 ¶¶ 5.1 through 5.4, available at: <http://www.cisg.law.pace.edu/cisg/CISG-AC-op6.html>; Gotanda, J.Y., "Awarding Damages under the United Nations Convention on the International Sale of Goods" (2005) 37 Georgetown Journal of International Law 95, 129ff.; Schlechtriem, P., "Legal Costs as Damages in the Application of UN Sales Law" (2006-07) 26 Journal of Law and Commerce 71. For case law in accord see UNCITRAL Digest of Art. 74 CISG case law ¶ 26, available at: <http://www.uncitral.org/uncitral/en/case_law/digests/cisg.html>.
25. See Felemegas, J., "An Interpretation of Article 74 CISG by the U.S. Circuit Court of Appeals" (2003) 15 Pace International Law Review 91 (2003); Zeller, B., "Interpretation of Article 74 -- Zapata Hermanos v. Hearthside Baking -- Where Next?" (2004) Nordic Journal of Commercial Law 1; Diener, K.W., "Recovering Attorneys' Fees under CISG: An Interpretation of Article 74" (2008) Nordic Journal of Commercial Law (Issue 1).
26. Schlechtriem, P., "Legal Costs as Damages in the Application of UN Sales Law" (2006-07) 26 Journal of Law and Commerce 71; Gotanda, J.Y., "Awarding Damages under the United Nations Convention on the International Sale of Goods" (2005) 37 Georgetown Journal of International Law 95, 129ff.; CISG Advisory Council Opinion No. 6, Calculation of Damages under CISC Article 74 ¶¶ 5.1 through 5.4, available at: <http://www.cisg.law.pace.edu/cisg/CISG-AC-op6.html>.
27. See the text accompanying fns 27-29 infra.
28. Khoo, W., "Article 2" ¶¶ 3.2 & 3.3 in Bianca, CM. and Bonell, M.J. (eds), Commentary on the International Sales Law: The 1980 Vienna Sales Convention, 1987.
29. Khoo, W., "Article 2" ¶¶ 3.2, in Commentary on the International Sales Law, supra fn 26 (citing United Nations Conference on Contracts for the International Sale of Goods (Vienna, 10 March -- 11 April 1980), Official Records, New York 1981, at pp. 295-298).
30. UNCITRAL, Report of the Committee of the Whole relating to the draft Convention on the International Sales of Goods, 1977, ¶¶ 177-178, reprinted in Honnold, J. (ed), Documentary History of the Uniform Law for International Sales, 1989, at p. 330.
31. See Bezirksgericht Sanne, Switzerland, 20 February 1997, CLOUT case No. 261, available at: <http://cisgw3.law.pace.edu/cases/970220s1.html>; ICC Court of Arbitration case No. 6653, 26 March 1993, CLOUT case No. 103, available at: <http://cisgw3.law.pace.edu/cases/936653i1.html>.
32. Schlechtriem, P., "Art. 4" ¶¶ 5 and 22, in Commentary on the UN Convention, supra fn 7; Ferrari, F., "Scope of application," supra fn 1, at pp. 110-113; Ferrari, F., "Burden of Proof under the United Nations Convention on Contracts for the International Sale of Goods (CISG)" (2000-2001) Pace Review of the Convention on Contracts for the International Sale of Goods 1; Magnus, U., "General Principles of UN-Sales Law," available at: <http://www.cisg.law.pace.edu/cisg/biblio/magnus.html> (English translation of German text originally appearing in (1995) 59 Rabels Zeitschrift für ausländisches und Internationales Privatrecht 469), text accompanying nn. 86-91; Kruisinga, S., (Non-)conformity in the 1980 UN Convention on Contracts for the International Sale of Goods: A Uniform Concept?, 2004, at pp. 157-86.
33. See Tribunale Vigevano, Italy, 12 July 2000, CLOUT case No. 378, available at: <http://cisgw3.law.pace.edu/cases/000712i3.html>; Tribunale Pavia, Italy, 29 December 1999, CLOUT case No. 380, available at: <http://cisgw3.law.pace.edu/cases/991229i3.html>; Handelsgericht Zurich, Switzerland, 10 February 1999, CLOUT case No. 331, available at: <http://cisgw3.law.pace.edu/cases/990210s1.html>; Handelsgericht Zurich, Switzerland, 26 April 1995, CLOUT case No. 196, available at: <http://cisgw3.law.pace.edu/cases/950426s1.html>; Handelsgericht Zurich, Switzerland, 9 September 1993, CLOUT case No. 97, available at: <http://cisgw3.law.pace.edu/cases/930909s1.html>
34. The phrasing of Art. 35(2)(b) CISG, in my view, indicates that the seller bears the burden of proving that the buyer has not relied on the seller's skill and judgment. See Flechtner, H.M., "Moving Through Tradition Towards Universalism under the U.N. Sales Convention (CISG): Notice of Lack of Conformity (Article 39) and Burden of Proof in the Bundesgerichtshof Opinion of 30 June 2004" in Erauw, J., Tomljenovi , V. and Volken, P. (eds), Liber Memorialis Professor Petar Sarcevic, 2006, 457 at p. 467 n. 33.
35. Bundesgerichtshof, Germany, 9 January 2002, English translation available at: <http://cisgw3.law.pace.edu/cases/020109g1.html>.
36. Bundesgerichtshof, Germany, 30 June 2004, English translation available at: <http://cisgw3.law.pace.edu/cases/040630g1.html>.
38. Ibid. The discussion of this decision herein reflects the analysis in Flechtner, H. M., "Moving Through Tradition", supra fn 33, at p. 457.
39. Ferrari, F., "Burden of Proof", supra fn 31, at pp. 6-7.
40. See the text accompanying supra fns 27-29.
41. As Professor Ferrari has pointed out, those who argue that burden of proof issues are outside the scope of the Convention have been far from unanimous in their approach to the issue of which domestic law should govern the question. See Ferrari, F., "Scope of application," supra fn 1, at p. 110. Fortunately, the choice of 'Selected Topics' as my subject allows me blithely to avoid this issue, which I freely admit is beyond my expertise.
42. See Tribunale d'Appello Lugano, Switzerland, 15 January 1998, CLOUT case No. 253, English translation available at: <http://cisgw3.law.pace.edu/cases/980115s1.html>.
43. See Ferrari, F., "Burden of Proof", supra fn 31, at p. 8.
44. See Lookofsky, J., "In Dubio Pro Conventione," supra fn 20, at pp. 264-65.