61995C0106

1 The questions referred to the Court for a preliminary ruling by the Bundesgerichtshof (Federal Court of Justice) are concerned with the interpretation of Articles 5(1) and 17 of the Convention of 27 September 1968 on jurisdiction and the enforcement of judgments in civil and commercial matters (`the Convention'), as amended by the 1978 Accession Convention.

More specifically, the national court wishes to establish whether an `abstract' agreement on the place of performance - that is to say, an agreement concluded, not to establish the place at which the person liable actually has to perform his obligations, but simply and exclusively to designate the courts having jurisdiction - is valid under Article 5(1) of the Convention and hence whether the courts for the place of performance thus identified have jurisdiction to rule on disputes relating to the obligation in question. If that question is answered in the negative, the Court is asked whether in this case the `abstract' agreement fulfils the requirements of Article 17 of the Convention, that is to say, whether a jurisdiction clause has been validly concluded.

2 I would first hark back to the origins of these proceedings. By a time charterparty concluded orally, the inland waterway company Mainschiffahrts-Genossenschaft e.G. (`MSG'), having its seat in Germany (Würzburg), made an inland waterway vessel available to the undertaking Les Gravières Rhénanes SARL (`the defendant'), having its registered office in France. The vessel was used on the Rhine between 1 June 1989 and 10 February 1991 to transport gravel from the place of loading to the place of unloading, both of which were located in France. (1) During unloading operations, the vessel was damaged by the unloading equipment used by the defendant on its responsibility. The damage, which was made good after the contract had expired, was assessed at different levels by the respective assessors appointed by the parties.

What is at issue in these proceedings is the sum of DM 197 284, namely the difference between the amount paid by the defendant's insurers and the amount claimed by MSG. With a view to obtaining that sum, MSG brought an action for damages for breach of the charterparty before the Schiffahrtsgericht (Maritime Court) Würzburg, the competent inland waterways court.

3 MSG claims that the German courts have jurisdiction on the ground that the defendant did not express any objections to a commercial letter of confirmation containing a pre-printed reference to Würzburg - the place at which MSG itself had its principal place of business - as the place of performance and the forum having jurisdiction or to the invoices, including a similar reference, issued by it, which the defendant paid without challenging them in any way.

By interlocutory judgment, the Schiffahrtsgericht declared the application admissible. On appeal, the Oberlandesgericht (Higher Regional Court) Nürnberg upheld the defendant's claim that it could be sued only in the French courts, and dismissed the application as inadmissible for want of jurisdiction.

4 The plaintiff appealed on a point of law against that judgment to the Bundesgerichtshof, in which it claims that the judgment of the court of first instance should be upheld. In the order for reference, the Bundesgerichtshof found, inter alia, that `the contractual obligations under the time charterparty which was concluded were to be performed in France; this is where the main element of the performance of the contract took place: it was not only there that the defendant, as the party determining the use of the vessel, had its seat, but it was there that the vessel was almost invariably loaded and always unloaded, with the result that it was not necessary to have an agreement deviating from Article 5 of the Brussels Convention in order to safeguard the - by all means - legitimate interest of the plaintiff in having a single place of performance'.

After finding that all the - principal and ancillary - contractual obligations could be performed only in France, the Bundesgerichtshof therefore concluded that the determination of the place of performance was fictitious and that `the reference in [the plaintiff's] general conditions of business to Würzburg as the place of performance was intended solely to enable legal disputes to be brought before the courts of the place at which its principal place of business was located'. The decision had therefore been taken solely in order to give jurisdiction to the plaintiff's courts and hence to avoid complying with the formal requirements imposed by Article 17 of the Brussels Convention. Nevertheless, as the Bundesgerichtshof points out, under German law `an effective agreement on the place of performance was made'.

5 In view of those circumstances, the Bundesgerichtshof deemed it necessary to refer the following questions to the Court in order to resolve the dispute pending before it:

`1. Is an oral agreement on the place of performance (Brussels Convention, Article 5) to be recognized even if it is not intended to fix the place at which the person liable has to perform the obligations incumbent on him, but is intended solely to establish - informally - that the courts for a particular place are to have jurisdiction (a so-called "abstract" agreement on the place of performance)?

2. In the event that the Court of Justice should answer question 1 in the negative:

(a) Can an agreement conferring jurisdiction in international trade or commerce in accordance with the third hypothesis mentioned in the second sentence of the first paragraph of Article 17 of the 1978 version of the Brussels Convention also be concluded by one party's not contradicting a commercial letter of confirmation containing a pre-printed reference to the courts of the consignors' place of business having sole jurisdiction or must there have been in every case prior consensus with regard to the content of the letter of confirmation?

(b) Is it sufficient in order for there to be an agreement conferring jurisdiction within the meaning of the aforesaid provision if the invoices sent by one party all contain a reference to the courts of the carrier's place of business having sole jurisdiction and to the conditions of the bill of lading used by the carrier which also stipulate the courts of the same place as having jurisdiction, and the other party invariably paid the invoices without objecting, or is prior consensus also required in this respect?'

6 What therefore has to be established in the first place is whether an abstract agreement on the place of performance - which simply conceals a jurisdiction clause, which, as such, would be subject to formal requirements - is to be considered valid under Article 5(1). If that question is answered in the negative, it has to be determined whether in the case in point the agreement on the place of performance may be regarded as a validly concluded jurisdiction clause under Article 17.

The Commission has suggested reversing the order of the preliminary questions on the ground that, if only on logical grounds, it should be determined whether there is a valid jurisdiction clause within the meaning of Article 17 before any finding is made as to whether there is a valid agreement on the place of performance within the meaning of Article 5(1). This is precisely because Article 17 provides for exclusive jurisdiction, which therefore prevails over every other jurisdiction, even special jurisdiction under Article 5(1). In addition - the Commission further maintains - if there were to be a validly concluded jurisdiction clause this would mean that the Court would have no need to answer a question on the validity of an abstract agreement on the place of performance which is undoubtedly more complex and thornier (sic!).

7 For my part, I consider that the order of the questions as they have been put by the national court should be respected. Over and above the different (or allegedly different) degree of difficulty of the two questions - a criterion which, to my mind, cannot or should not influence the Court's decision - I consider that the Bundesgerichtshof's decision to place the question on the interpretation of Article 5(1) first is anything but a matter of chance. Whilst an affirmative answer to that question would mean that the German courts certainly have jurisdiction, it is possible that the Court's answer with regard to Article 17 may require further findings of fact on the part of the national court in order to establish which courts have jurisdiction, which it is not competent to the Bundesgerichtshof to make. Consequently, should the Court decide to answer the second question only, this might make a second reference to the Court necessary with regard to the same dispute.

In addition, there is the fact that the question of abstract clauses on the place of performance certainly should not be underestimated and, as will be seen, has been tackled and resolved by the various national courts in different ways. In those circumstances, clarification from the Court seems necessary. I shall therefore consider the questions in the order in which they have been put.

The first question

8 Article 5(1) of the Convention, which is the subject of the national court's first question, provides that `a person domiciled in a Contracting State may, in another Contracting State, be sued: in matters relating to a contract, in the courts for the place of performance of the obligation in question; ...'.

Next, it should be observed that Article 5(1), which appears in Section 2, `Special jurisdiction', provides for an exception to the general rule contained in Article 2 that the courts having jurisdiction should be those of the State in which the defendant is domiciled. This derogation in favour of the place of performance of the obligation in question is justified, as the Jenard Report points out, by the fact that there is `a close connecting factor between the dispute and the court with jurisdiction to resolve it'. (2)

9 The same reasoning is to be found in the relevant case-law of the Court, which has repeatedly held that that freedom of choice on the part of the plaintiff under Article 5 `was introduced in view of the existence in certain well-defined cases of a particularly close relationship between a dispute and the court which may be most conveniently called upon to take cognizance of the matter'. (3) In other words, the reason why the courts of the place of performance should have jurisdiction is that there is a direct and, above all, objective connection between the dispute and the courts having jurisdiction to entertain it. It is consequently the physical proximity of those courts to the relationship at issue which, at least in the intention of the authors of the Convention, justifies their having jurisdiction.

The jurisdictional criterion set out in Article 5(1), however, is not, or at least not directly, the `proximity' of the courts to the dispute, but the place of `performance of the obligation in question'. (4)

10 It should next be recalled that the Court has consistently held, except in relation to employment contracts, (5) that for the purposes of determining the place of performance within the meaning of Article 5(1) `the obligation to be taken into account is that which corresponds to the contractual right on which the plaintiff's action is based'. (6) The place of performance of that obligation is determined by referring to the national law. The Court has in fact stated that for this purpose the court seised must `determine in accordance with its own rules of conflict of laws what is the law applicable to the legal relationship in question and define in accordance with that law the place of performance of the contractual obligation in question'. (7)

In brief, Article 5(1) must be interpreted as meaning that the courts of the place of performance of the contractual obligation actually relied on in the proceedings have jurisdiction; the place of performance of that obligation is to be determined on the basis of the substantive law applicable thereto under the provisions of the private international law of the court seised.

11 Having said this, the place of performance may also be determined contractually as between the parties. This may be inferred from Zelger v Salinitri, (8) in which the Court specifically considered the relationship between Article 5(1) and Article 17. In that case, the Court was called upon to decide whether the validity of an agreement between the parties on the place of performance of the obligation in question was or was not dependent upon compliance with the formal requirements laid down by Article 17.

Starting from the premiss that `the jurisdiction of the court for the place of performance (provided for in Article 5(1)) and that of the selected court (provided for in Article 17) are two distinct concepts and only agreements selecting a court are subject to the requirements of form prescribed by Article 17 of the Convention' (paragraph 4 of the judgment), the Court reached the conclusion that `if the parties to the contract are permitted by the law applicable to the contract, subject to any conditions imposed by that law, to specify the place of performance of an obligation without satisfying any special condition of form, an agreement on the place of performance of the obligation is sufficient to found jurisdiction in that place within the meaning of Article 5(1) of the Convention' (paragraph 5).

12 In accordance with Zelger v Salinitri it is therefore enough, in order for a clause specifying the place of performance to be valid under Article 5(1), that `the place of performance of a contractual obligation has been specified by the parties in a clause which is valid according to the national law applicable to the contract' (paragraph 6).

If that solution were to be applied to the case now before the Court, it would have to be held that the agreement on the place of performance - in so far as it was valid under the German law applicable to the contract - was also valid for the purposes and effects of Article 5(1). However, as the Bundesgerichtshof itself observes in the order for reference, `it cannot, however, be inferred without question from the judgment in Zelger v Salinitri whether oral agreements on the place of performance which are not concerned with fixing the place at which the person liable has to carry out the performance due from him, but are intended instead solely to determine the courts of a particular place as having jurisdiction, without the requirements of Article 17 of the Brussels Convention having to be fulfilled (so-called abstract agreements on the place of performance), must also be recognized'.

13 I would start by observing that the solution adopted by the Court in Zelger v Salinitri, which I concur with in principle, did indeed leave the issue now before the Court open. This is evidenced by the fact that that judgment does not make any mention of the problem of possible circumvention of Article 17 by means of fictitious determinations of the place of performance, even though that problem was discussed during the proceedings and tackled by the Advocate General in his Opinion. (9) In any case, suffice it to say that, unlike in the present case, what was in issue in Zelger v Salinitri was not an `abstract' agreement: the place of performance specified by the parties was in fact the same as that determined by the law.

In addition, in that case, while stating that provided that an agreement between the parties on the place of performance within the meaning of Article 5(1) is in accordance with the applicable national law, it does not have to comply with the formal requirements imposed by Article 17, the Court did not confine itself to pointing out that the provisions in question have different functions and therefore operate on different levels. (10) It stressed in particular that whilst jurisdiction determined by virtue of Article 5 is justified by `the existence of a direct link between the dispute and the court called upon to take cognizance of it' (paragraph 3), exclusive jurisdiction under Article 17 `dispenses with any objective connection between the legal relationship in dispute and the court designated' (paragraph 4).

14 It is undisputed that the place of performance indicated in the `abstract' agreement which is at issue in these proceedings dispenses with an objective connection between the legal relationship at issue and the court designated. Moreover, according to the national court, in this case the only purpose of the agreement on the place of performance is precisely that of designating the courts having jurisdiction. Since this is the way matters stand, it is unquestionable that the agreement at issue tends to come within the ratio legis of Article 17 and should therefore come within its scope, if only logically. Consequently, it should be declined validity under Article 5.

This solution, however, would seem to come up against a not inconsiderable number of obstacles, in view, among other things, of the fact that the Court's case-law has established unequivocally: (a) that the place of performance should be determined by reference to national law (Tessili v Dunlop); (b) that national law may provide that the parties can determine the place of performance, in which case it is unnecessary to comply with the formal requirements of Article 17 (Zelger v Salinitri); and lastly (c) that the place of performance thus identified may dispense with any objective connection with the dispute (Custom Made Commercial (11)). The obvious conclusion to which reading those three judgments together brings one is precisely that possible fictitious determinations of the place of performance are invariably and in any event valid under Article 5(1). This, moreover, was the proposition put forward by the Commission during these proceedings, a proposition which - if the judgments in question are read in isolation from their context - leads one to a conclusion which, albeit on the face of it unobjectionable and certainly simpler, completely fails to take account the function played by Article 17 within the system of the Brussels Convention.

15 In the final analysis, on the basis of this proposition, once it is conceded that the parties may locate the constituent elements of the contract as and where they wish - which allegedly may be inferred from the abovementioned judgments - the inevitable consequence would be that the parties themselves are entitled to `utilize' the determination of the place of performance also in order to choose a forum other than the one which would normally have jurisdiction (under Article 2 or under Article 5 if the place of performance were to be determined by virtue of the law). In other words, there would be no limit to the parties' will, apart from any laid down by national law itself, either as a result of the imposition of formal requirements or as a result of case-law referring to institutions such as fraud on a statute. (12) Otherwise, agreements on the place of performance would have to be recognized as being valid invariably and in any event.

That interpretation of the aforementioned three judgments, which is admittedly based only on syllogisms, is to my mind somewhat over-simplistic. In the first place, whilst it is true that the place of performance is determined by reference to the substantive law, this does not automatically mean that the Community Court is precluded from laying down limits designed to prevent some other provision of the Convention, in this case Article 17, from being circumvented. Secondly, it seems to me that, in view of the reference which it makes to the direct connection between the dispute and the court called upon to decide it, Zelger v Salinitri should instead be interpreted as meaning that the parties may indeed decide to locate the performance of the contract at a place other than the one which would be fixed by the law, but that place must at all events be the actual place of performance: only then are the parties not obliged to comply with the formal requirements set out in Article 17. Lastly, for present purposes I consider that the Court's dictum in Custom Made Commercial that `under Article 5(1), in matters relating to a contract, a defendant may be sued in the courts for the place of performance of the obligation in question, even where the court thus designated is not that which has the closest connection with the dispute has no bearing on the matter'. (13) What was at issue in Custom Made Commercial was the possibility of excluding reference to the lex causae as the criterion for determining the place of performance: this was simply because, if that criterion were to be applied, the court which would have jurisdiction would be the one less connected with the relationship at issue. (14)

16 In the case now before the Court, in contrast, it is undisputed that the agreement in question is not confined to conferring jurisdiction on the courts for a place which has no connection with the case, except for the fact that it is the plaintiff's domicile, but is designed simply to evade the formal requirements laid down by Article 17. The issue raised by these proceedings therefore goes considerably further than whether jurisdiction under Article 5 must invariably and in any event be vested in the courts for the place which has the greater physical connection with the case in question.

In the final analysis, here the question is that of establishing whether, where the contractual obligation cannot in fact be performed in the place agreed as the place of performance either because that place is at odds with the very nature of the contract or because it is not consistent with geographical fact, Article 5(1) nevertheless remains a valid jurisdictional criterion.

17 To allow such a solution would manifestly mean that the parties are in fact entitled, by means of a fictitious determination of the place of performance, to circumvent the formal obstacles placed in the way of jurisdiction clauses. Should the Court accept this solution, it would therefore be giving its blessing to an alternative use of Article 5(1), in that the parties would be authorized to designate a court other than the one which generally has jurisdiction by certainly simpler means than those provided for by Article 17, and this would result in both the wording and the purpose of that provision being breached or at least devalued. It is hardly necessary to point out in this connection that the formal requirements imposed by Article 17 are not aims in themselves, but seek to protect the weaker party to the contract; the aim pursued by them, or rather by means of them, is therefore to avoid jurisdiction clauses incorporated in the contract by only one party passing unobserved.

It is only too obvious from this point of view that Article 5(1) may not be interpreted in such a way as to enable the outcome to be brought about which, through the imposition of rigid formal requirements, Article 17 set out to avoid. (15) The question would arise as to how making use of that state of affairs could be regarded as correct having regard to Article 5(1) and consistent with the system and aims of the Convention as a whole and in particular with regard to Article 17.

18 Although it recognizes that a fictitious determination of the place of performance may be a means of circumventing the formal requirements laid down by Article 17, the Commission nevertheless considers that this is permitted by the Convention itself. It simply points out in this connection that Article 5(1) does not require the parties to indicate the actual place of performance of the contractual obligations.

That argument goes too far. In fact, the only criterion which can be inferred from the provision in question, which however contains no indication as to how the place of performance may be determined contractually, is that the place must be the place of performance of the obligation in question. Now, it seems reasonable to me to consider that that criterion, far from enabling any place in the world to be chosen, requires there to be a connection with the place at which the obligation in question actually has been or should be performed. In brief, even if the place is different from that fixed under the law, it must still be a place which is connected with, or at least appropriate to, the subject-matter of the contract.

Neither does the argument that such an approach would lead parties no longer to stipulate clauses on the place of performance, which would end up by devoiding Article 5(1) of its substance, seem to me to have any particular importance. I would merely observe in this connection that the place of performance certainly is not destined to disappear simply because agreements on the place of performance in certain circumstances do not have the effect of conferring jurisdiction: the plaintiff may still sue in the courts for the place of performance of the obligation in question under the law.

19 In the final analysis, I consider it absurd even only to conjecture that the Convention might be interpreted in such a way as to enable Article 17 to be circumvented, bearing in mind moreover the importance of Article 17 within the system of the Convention. In view of this, I am unable to agree with the Commission's proposition that this is the price to be paid in order for there to be legal certainty and foreseeability of the courts having jurisdiction. In point of fact, those objectives are certainly not aims in themselves; for example, making the application of the rules on jurisdiction foreseeable means first and foremost reinforcing the legal protection available to persons established in the Community by allowing the plaintiff easily to identify the court before which he may bring an action and the defendant reasonably to foresee the court before which he may be sued. (16) I am unable to see, moreover, in what way legal certainty might be strengthened by a breach of Article 17.

It is true that to superimpose the two provisions without any interpretative review is unacceptable, especially bearing in mind the need to safeguard balance between the parties as sought by Article 17.

20 Unless Zelger v Salinitri is called in question and the place of performance is accordingly determined independently where it is decided upon contractually as between the parties - a solution which is in principle desirable and would have the merit of being linear and unequivocal, (17) but would oblige the parties to comply with particularly strict formal conditions, even where the actual place of performance was designated - the simplest solution is for the Community Court to lay down limits designed simply and solely to obviate circumvention of Article 17. (18)

Fixing such limits does not moreover require particular imagination or effort. In my view, it is sufficient to establish that where the national court finds - as the German court found in this case - that the place of performance determined by the parties has no substantial connection with the subject-matter of the contract and that that place was determined solely in order to introduce a jurisdiction clause without having to comply with the formal requirements laid down by Article 17, the place of performance thus identified cannot be regarded as complying with Article 5(1) of the Convention.

The second question

21 By this question, which is asked in the event that the Court should consider, as I have suggested it should, that agreements on the place of performance having the sole aim of formally designating the court having jurisdiction as a court other than that for the actual place of performance are unlawful, the Bundesgerichtshof seeks to establish whether in such case it may considered that there is a jurisdiction clause under Article 17 of the Convention. More specifically, the question is whether, in international trade or commerce, a jurisdiction clause can be regarded as having been validly concluded even through one party's lack of response to a commercial letter of confirmation [question 2(a)] or through that party's having paid without objection invoices referring to the courts having jurisdiction [question 2(b)], or whether there should in every case have been prior consensus between the parties and that consensus must necessarily have been evidenced in writing.

The question therefore relates to the interpretation of the second sentence of the first paragraph of Article 17, as amended by the 1978 Accession Convention. That convention added to the possible ways of concluding an agreement conferring jurisdiction - `in writing' (first hypothesis) or `evidenced in writing' (second hypothesis) - a third hypothesis to the effect that a jurisdiction clause may be concluded `in international trade or commerce, in a form which accords with practices in that trade or commerce of which the parties are or ought to have been aware'.

22 The Court has not yet had occasion to rule on the interpretation of that provision.(19) Before I even proceed to consider the aims and innovations introduced by that provision and draw conclusions with regard to the case before the Court, I consider it nevertheless appropriate to call to mind the interpretation which the Court gave to the version of Article 17 which was in force before the amendment under discussion, naturally in so far as that interpretation is relevant to the present case.

The Court has consistently held that the provisions of Article 17 must be interpreted restrictively and strictly, given that they constitute an exception to the general principle that the defendant's courts have jurisdiction (Article 2) and to the special jurisdictions provided for in Articles 5 and 6. In particular, the Court has made it clear that by making the validity of the conferment of jurisdiction subject to the `existence of an "agreement" between the parties, Article 17 imposes on the court before which the matter is brought the duty of examining, first, whether the clause conferring jurisdiction upon it was in fact the subject of a consensus between the parties, which must be clearly and precisely demonstrated'. (20)

Thus, for example, the Court denied any validity to a jurisdiction clause set out on the reverse of a letter of confirmation delivered by the seller to the purchaser after the conclusion of an oral agreement on the ground that it did not comply with the formal requirements laid down by Article 17, (21) which is a very similar case to the one at issue in these proceedings. Even in those circumstances the Court held that there was a need for written confirmation on the part of offeree, the only attenuation consisting in the possibility of departing from the requirement for writing in the case of a continuing trading relationship. (22)

23 The new version of Article 17 seems to have been drawn up precisely because of the excessive formalism of such an approach in regard to international trade. Indeed, it is stated in the Schlosser Report (23) that the Court's interpretation of Article 17 `does not cater adequately for the customs and requirements of international trade. In particular, the requirement that the other party to a contract with anyone employing general conditions of trade has to give written confirmation of their inclusion in the contract before any jurisdiction clause in those conditions can be effective is unacceptable in international trade. International trade is heavily dependent on standard conditions which incorporate jurisdiction clauses.'

Seen in that light, the amendment introduced by the 1978 Accession Convention certainly reflects greater (perhaps necessary) attention and sensitivity to the requirements of international trade and, more generally, to the actual workings of the business world. It is obvious that too strict an application of the principles set out in Article 17 would make it virtually impossible for jurisdiction clauses to operate where they were contained in contractual documents which because of their specific characteristics are not signed by one of the parties.

24 Whilst those were the considerations which prompted the amendment of Article 17, the fact remains, as the Schlosser Report itself emphasizes, that `This is however, as should be clearly emphasized, only a relaxation of the formal requirements. It must be proved that a consensus existed on the inclusion in the contract of the general conditions of trade and the particular provisions'. (24)

In the final analysis, the intention behind that amendment would seem to have been to attenuate only the formal requirements laid down by Article 17 in order to bring them into line with the requirements of international trade, yet without losing sight of the aim of actual consensus. The conflict between those two requirements - or in any event the difficulty in reconciling them - is only too obvious. The way in which consensus is manifested, according to the Court's interpretation, is closely connected with proving actual consensus between the parties. (25) Whilst it is therefore true - and how could it be doubted? - that the formal requirements are not an end in themselves but serve precisely to prove that there was genuine consensus between the parties and hence that the agreement on the conferment of jurisdiction is valid, it follows from this, as has been pointed out in academic writings, (26) that any relaxation of the formal requirements is bound to have repercussions on the reality of the consensus.

25 On top of this, it is extremely difficult to establish that there are trade or commercial practices with regard only to formal requirements - these being procedural rules which are liable to be changed by the States at any time (27) - and doing so would therefore devoid the new version of Article 17, as amended by the 1978 Accession Convention, of any meaning. It must therefore be held that the amendment entails not only a relaxation of the ways in which consensus is manifested, but also, for that very reason, less strictness with regard to the formation of the parties' consensus with regard to designating the court having jurisdiction. In other words, actual consensus, which was initially essential and guaranteed only by writing or evidence in writing of an oral agreement, now yields in international trade to a presumption of actual consensus. This is confirmed by the very wording of the provision in question in so far as it provides that the practice in question must be one of which the parties were aware or `ought to have been aware'.

However, the foregoing must not cause it to be overlooked that the essential aim of Article 17 is to protect the weaker party to the contract and, to that end, to avoid jurisdiction clauses incorporated in the contract by one party alone from going unnoticed. It follows that the interpretation of the provision in question must necessarily and rigorously be strict. Furthermore, the interpretation cannot in any event be such as to cause the objective of `actual' consent to be lost sight of, if only through awareness or reasonably imputable knowledge of customary practices in international trade or commerce. To my mind, this is the criterion which should guide the interpretation of the provision in question. (28)

26 Given all these considerations, when and in what way may it be considered that there is a trade or commercial practice with regard to manners of achieving consensus of which the parties are or ought to have been aware?

I would immediately exclude determination of the existence of such a practice by referring to the lex causae or the lex fori, as some commentators have suggested, (29) and hence to the legal system identified thereby. In the first place, it is all too obvious that such a solution would be liable to conflict with the ratio legis of the provision in question: it might have the perverse effect of endorsing as `trade or commercial practices' within the meaning of Article 17 practices, not of international trade, but of only one or more legal systems, with the result that validity would be conferred on jurisdiction clauses based (possibly) on only local practices or on `practices' have not yet taken hold as such. On top of this, the Court has to date interpreted the concept of `agreement' between the parties within the meaning of Article 17 independently. There does not seem to me to be any ground for diverging from this approach where the manners of reaching a consensus are attributable to trade or commercial practices.

27 Consequently, it is appropriate, also with regard to the third hypothesis provided for in Article 17, that the Court itself should indicate the objective evidence and/or conclusive conduct from which it can be inferred that one is in the presence of a trade or commercial practice which may appropriately be `utilized' for the purposes of the valid conclusion of a jurisdiction clause. Seen from that perspective and bearing in mind the importance of consensus on the part of the parties concerned, the concept of trade or commercial practices for the purposes and effects of Article 17 can only be based, as Advocate General Lenz has suggested, on the existence of `a practice ... which is followed generally, continuously and regularly by the circles concerned in transactions which correspond, both objectively and in point of their localization, to the transaction at issue'. (30) In brief, the practice must be such as to lead to the conviction that the parties' conduct is conclusive, that is to say that it implies a meeting of the minds and hence actual consensus as regards the conferment of jurisdiction. From that perspective, the existence of a specific trade or commercial practice in a particular sector should in any event be proved: it is only in those circumstances that the practice acquires legal effectiveness under Article 17. (31)

This interpretation is moreover confirmed by the 1988 Lugano Convention (32) and the Brussels Convention as amended by the 1989 San Sebastian Convention. (33) Apart from the requirements set out in the 1978 version, Article 17 of the Lugano and San Sebastian Conventions requires the trade or commercial usage in question to be `widely known to, and regularly observed by, parties to contracts of the type involved in the particular trade or commerce concerned'. (34) Accordingly, it is not enough for a jurisdiction clause to have been concluded in a form which accords with a usage in the particular trade or commerce concerned of which the parties were or ought to have been aware; the usage must have been `on the one hand, widely known in international trade or commerce, and, on the other, regularly observed by parties to contracts of the type involved in the particular trade or commerce concerned'. (35)

28 In my view, this suggested interpretation of the notion of a trade or commercial usage also enables it more easily to be established in what circumstances it should be considered, at least in principle, that the parties were or ought to have been aware of the usage in question. It is self-evident that it may reasonably be presumed that there was reasonably imputable knowledge of a practice which was continually followed in relation to contracts of the same type in the particular trade or commerce concerned.

Having said this, it is true that there cannot be an absolute presumption of reasonably imputable knowledge, it should be proved by objective evidence. To this end, particular importance attaches to factual circumstances, such as previous contractual dealings between the same parties or even with other parties but with regard to contracts of the same geographical and substantive type. Again, it may be important to ascertain whether or not the relevant usage was recognized by the legal system of the trader against whom the jurisdiction clause is sought to be enforced.

29 Let us return to the case before the Court. Since the contract in question was a charterparty concluded between traders in the relevant sector, there is no doubt that we are in the realm of international trade or commerce and hence within the field of application of the provision in question. I would go on to point out once again that the national court asks whether one party's lack of response to a commercial letter of confirmation containing a pre-printed indication of the courts having jurisdiction constitutes a practice within the meaning of Article 17 and hence should be deemed to constitute acceptance of the conferment of jurisdiction. In the alternative, the national court asks whether that conferment of jurisdiction may be deemed valid as a result of the defendant's having paid all MSG's invoices containing a similar indication of the competent courts.

In view of foregoing considerations it is only too plain that the existence of commercial practices of the type contemplated by the national court's questions with regard to the manner of achieving and manifesting consensus has to be proved in relation to the trade or commerce in question.

30 I therefore cannot accept the proposition put forward by the plaintiff and the German Government in the proceedings to the effect that, under Article 17, lack of response to a commercial letter of confirmation is to be deemed to constitute tacit acquiescence, irrespective as to whether any inquiry has been made whether there exists a practice to that effect in the relevant trade or commerce. (36) That proposition, which relies, on the one hand, of the existence of a presumed, generalized - that is to say, Europe-wide - practice and, on the other, on the fact that the amendment made by the 1978 Accession Convention was adopted precisely in order to avoid the excessive formalism of the case-law with regard to cases of this type, (37) is in fact without any foundation.

Admittedly, it emerges from the Schlosser Report that one of the reasons for the amendment in question was the desire to avoid the requirement for writing with regard to a commercial letter of confirmation. However, this does not mean, or at least not automatically, that the practice in question is recognized invariably and in any event as having the status of a commercial practice within the meaning of Article 17. What is more, the legal significance of silence in international trade is not recognized as broadly and generally as it is in Germany. I would just refer, for example, to Article 18(1) of the Vienna Convention of 11 April 1980 on Contracts for the International Sale of Goods, according to which `A statement made by or other conduct of the offeree indicating assent to an offer is an acceptance. Silence or inactivity does not in itself amount to acceptance.' (38)

31 In the final analysis, in order to establish whether the silence of one party in relation to a commercial letter of confirmation constitutes a manner of acceptance and hence means that a jurisdiction clause has validly been concluded, it is necessary to prove the existence of such a usage on the basis of the criteria set out above.

The national court has not provided the Court with any evidence enabling any view to be given in this regard. In these circumstances, it is therefore for the national court to establish whether in the sector of inland waterway transport on the Rhine there is a consolidated, habitually followed practice in relation to charterparties on the basis of which silence in response to a commercial letter of confirmation is deemed to constitute acceptance of the jurisdiction clause. In the event that it is found that there is such a usage, it will then have to be established whether the party against whom the jurisdiction clause is relied was or ought to have been aware of the existence of such a usage. Whether such reasonably imputable knowledge existed of the usage in question will depend, in turn, on the importance of that usage in the country in which the party against which the clause is relied is habitually resident and on whether or not this was the first contract concluded with that party and/or in that sector.

32 The solution just put forward can also be applied in order to determine whether there was a presumed trade or commercial usage consisting in the fact that the defendant paid without any objection the invoices issued by MSG, all of which contained a statement to the effect that the place at which MSG had its principal place of business was the competent forum. It will therefore be necessary to show also with regard to this hypothesis that, on the basis of the same criteria listed above, there was a commercial usage of which the parties were or ought to have been aware.

33 In the light of the foregoing considerations, I therefore propose that the Court should answer the Bundesgerichtshof's questions as follows:

(1) Article 5(1) of the Brussels Convention of 27 September 1968 must be interpreted as precluding jurisdiction being vested in the courts for the place of performance of a contractual obligation where that place, determined contractually by the parties, has no substantial connection with the subject-matter of the contract and it appears that that determination was made with the sole aim of designating the courts having jurisdiction without complying with the formal requirements laid down by Article 17 of the Convention.

(2) The third hypothesis provided for in the second sentence of the first paragraph of the Brussels Convention of 27 September 1968, in the version as amended by the 1978 Accession Convention, must be interpreted as meaning that a jurisdiction clause may be regarded as having been validly concluded by reason of the silence of one party with regard to a commercial letter of confirmation containing a pre-printed reference to the courts having jurisdiction or by reason of that party's having paid without objection invoices issued by the other party containing a similar reference. To that end, however, it is for the national court to ascertain: (a) whether there exists in the trade or commerce concerned a practice to that effect which is followed continuously and generally and whether that practice is followed in contracts which are similar from both the substantive and geographical points of view to the contract at issue; (b) whether the defendant was or ought to have been aware of that usage, in particular because it was known also in his country of residence and/or having regard to previous contractual relations with the same contracting party and/or in the same sector.

(1) - In this connection, it should be observed that it appears from the order for reference that the places of unloading were located solely in France and that, apart from some negligible exceptions, the places of loading were also in France. That statement has, however, been contested by MSG at the hearing, where it maintained that the places of loading were located chiefly in Germany.

(2) - OJ 1979 C 59, p. 1, in particular at p. 22.

(3) - Case 12/76 Tessili v Dunlop [1976] ECR 1473, paragraph 13.

(4) - But, as the Court itself has stressed, it is precisely that place which `usually constitutes the closest connecting factor between the dispute and the court having jurisdiction over it, and it is this connecting factor which explains why, in contractual matters, it is the court of the place of performance of the obligation which has jurisdiction' (Case 266/85 Shenavai v Kreisler [1987] ECR 239, paragraph 18).

(5) - The case-law has in fact determined, with regard to that type of contract, that the place of performance should be determined, not in accordance with the lex causae, but independently on the basis of the obligation characteristic of the contract (see, in particular, Case 133/81 Ivenel v Schwab [1982] ECR 1981). That case-law has since been expressly enshrined in Article 5(1) of the Brussels Convention, as amended by the 1989 Accession Convention.

(6) - Case 14/76 De Bloos v Bouyer [1976] ECR 1497, paragraph 13.

(7) - Tessili v Dunlop, cited in footnote 3, paragraph 13.

(8) - Case 56/79 Zelger v Salinitri [1980] ECR 89.

(9) - See the Opinion of Advocate General Capotorti in Zelger v Salinitri [1980] ECR 98, in particular at 101 et seq.

(10) - In particular, the Court stressed that, whilst Article 17 has a purely procedural function, Article 5(1) has an essentially substantive function, which is accompanied only indirectly by a procedural effect, an effect which is however independent of the parties' intention.

(11) - Case C-288/92 Custom Made Commercial [1994] ECR I-2913, paragraphs 14 to 21.

(12) - Thus, for example, the Cour d'Appel (Court of Appeal), Liège, by judgment of 12 May 1977 (Journal de Tribunaux 1977, p. 710) denied validity to an agreement on the place of performance on the ground that it was contrary to the geographical reality of the contract. That judgment was subsequently upheld by the Cour de Cassation (Court of Cassation) (judgment of 28 June 1979, Journal des Tribunaux 1979, p. 625, with a note by Vander Elst in Revue critique de jurisprudence belge, 1981, p. 347 et seq.), which referred to the fraudulent nature of the agreement in question. To all intents and purposes, therefore, the Belgian courts have resolved this problem on the basis of national law alone. For the rest, national case-law is fairly varied. For the most part, especially since Zelger v Salinitri, it finds that where there is an agreement on the place of performance there is no need to comply with the formal requirements set out in Article 17 (see to this effect the judgment of 28 March 1979 of the Cour d'Appel, Lyon, in La Semaine juridique, Edition générale 1981, Jurisprudence, No 19519, and the German case-law cited by Schack, `Abstrakte Erfüllungsortvereinbarungen: form- oder sinnlos?', in Praxis des internationalen Privat- und Verfahrensrechts, 1996. p. 247 et seq., footnote 5). The judgment of 27 February 1996 of the French Cour de Cassation (Europe, April 1996, No 171, p. 23) which held that an agreement on the place of performance was invalid on the ground that, first, the obligation in question could be performed only in France and, secondly, that under Article 5(1) of the Brussels Convention the only obligation to be taken into account is that which actually underlies the court proceedings, warrants a separate mention.

(13) - Custom Made Commercial, cited in footnote 11, paragraph 21.

(14) - In other words, whilst it is true that the jurisdiction of the courts for the place of performance is justified by the direct, objective link which normally exists between the relationship at issue and the court seised, the criterion used in Article 5 (place of performance of the obligation in question) may lead to different results, which should otherwise be disregarded.

(15) - In fact, even those commentators who approve of Zelger v Salinitri are in favour of laying down uniform formal requirements both for jurisdiction clauses and for `abstract' agreements on the place of performance. See Kropholler: `Europäisches Zivilprozeßrecht', 1996; Gaudemet-Tallon: Les conventions de Bruxelles et Lugano, Paris, 1993, p. 121; Kaye: Civil Jurisdiction and Enforcement of Foreign Judgments, Abingdon, 1987, p. 524; Lasok, Stone: Conflict of Laws in the European Community, Abingdon, 1987, p. 219; Desantes Real: La Competencia Judicial en la Comunidad Europea, Barcelona, 1986, p. 258 et seq.; Calvo-Caravaca: Comentario al Convenio de Bruselas, Madrid, 1994, p. 90 et seq. For the contrary view, see Geimer: Internationales Prozeßrecht, 1993, paragraph 1491.

(16) - Case 125/92 Mulox [1993] ECR I-4075, paragraph 11.

(17) - This solution is advocated by both Jayme: `The Role of Article 5 in the Scheme of the Convention. Jurisdiction in Matters Relating to Contract', in Civil Jurisdiction and Judgments in Europe. Proceedings of the Colloquium on the Interpretation of the Brussels Convention by the Court of Justice considered in the context of the European Judicial Area, Luxembourg, 11 and 12 March 1991, London, 1993, p. 73 et seq. and Huet: `Note sous l'arrêt de la Cour Zelger', in Journal du droit international, 1980, p. 435 et seq. Both of them suggest that the place of performance should be determined independently, which, without requiring compliance with the formal requirements set out in Article 17, would nevertheless require consensus on the part of the parties to be proven.

(18) - In this regard, I consider that there is some point in stressing that the absence of an independent concept of the place of performance, which would certainly be desirable, cannot be regarded as not permitting the Community Court to lay down limits designed to prevent Article 5 being used as an alternative to Article 17. Whilst it is in fact true that determination of the place of performance has been left to the substantive law, it is also true, as the Court has explained on several occasions, that the choice between an independent interpretation and referring to the rules of conflict of laws is made pragmatically in that `the appropriate choice can only be made in respect of each of the provisions of the Convention to ensure that it is fully effective having regard to the objectives of Article 220 of the Treaty' (Tessili v Dunlop, cited in footnote 3, paragraph 11). Fixing Community limits must be regarded as being all the more necessary in present circumstances where it is a question of securing full harmony and functionality for two provisions, one of which, Article 17, has to date been interpreted independently.

(19) - See, however, the detailed Opinion of Advocate General Lenz in Case C-288/92 Custom Made Commercial [1994] ECR I-2915, in particular at I-2934 et seq.

(20) - Case 24/76 Estasis Salotti [1976] ECR 1831, paragraph 7, and Case 25/76 Segoura [1976] ECR 1851, paragraph 6.

(21) - Segoura, cited in the preceding footnote.

(22) - This possibility was expressly added to Article 17 by the 1989 Accession Convention. The new version of Article 17 provides that a clause conferring jurisdiction can be validly concluded also `in a form which accords with practices which the parties have established between themselves'.

(23) - OJ 1979 C 59, p. 71, point 179.

(24) - My emphasis.

(25) - Indeed, the Court has held that `the purpose of the formal requirements imposed by Article 17 is to ensure that the consensus between the parties is in fact established' [Estasis Salotti, paragraph 7, and Segour, paragraph 6, both cited in footnote 20].

(26) - See, for example, Kohler: `Rigueur et souplesse en droit international privé: les formes prescrites pour une convention attributive de juridiction `dans le commerce international' par l'article 17 de la convention de Bruxelles dans sa nouvelle rédaction', in Diritto del Commercio Internazionale, 1990, p. 611 et seq.

(27) - See Mezger: Travaux du comité français de droit international privé, 1980-1981, p. 15 et seq.

(28) - Moreover, the concept of presumptive knowledge in the sense of reasonably imputable knowledge is not unknown in the Court's case-law. I refer to Case C-214/92 Powell Duffryn [1992] ECR I-1745, in which the Court held with regard to a clause conferring jurisdiction contained in the statutes of a company that `irrespective of how shares are acquired, every person who becomes a shareholder knows, or ought to know, that he is bound by the company's statutes' (paragraph 27; my emphasis).

(29) - See to this effect Rauscher: Zeitschrift für Zivilprozeß 104, 1991, p. 272, especially at p. 292 et seq.

(30) - Opinion in Custom Made Commercial, cited in footnote 19, in particular at I-2939.

(31) - See to this effect, among others, Kaye: Civil Jurisdiction and Enforcement of Foreign Judgments, Abingdon, 1987, p. 1062 et seq.; Huet, Note, in Clunet, 1990, p. 153 et seq.; Stöwe: Gerichtsstandsvereinbarungen nach Handelsgebrauch, 1993, p. 56 et seq.

(32) - OJ 1988 L 319, p. 1.

(33) - OJ 1989 L 285, p. 4.

(34) - These additional requirements were in fact taken over from Article 9(2) of the 1980 Vienna Convention on Contracts for the International Sale of Goods.

(35) - It is worth mentioning in this connection that the amendment arose precisely because of the concern expressed by the EFTA countries in the course of the adoption of the Lugano Convention that one party's lack of response to a commercial letter of confirmation might be deemed to be acceptance. See the Jenard and Möller Report, OJ C 189 1990, points 55 to 59.

(36) - See to this effect, for example, Schmidt: Recht der Internationalen Wirtschaft, 1992, p. 173 et seq.

(37) - The reference is to the Segoura case, cited in footnote 20.

(38) - My emphasis. Likewise, I would call to mind Article 8(2) of the Rome Convention on the law applicable to contractual obligations, which relates to the existence and validity of the parties' consent as to the law applicable to the contract. That provision enables either party to refer to the law of the country in which he has his habitual residence in order to establish that he did not consent `if it appears from the circumstances that it would not be reasonable to determine the effect of his conduct' in accordance with the law theoretically applicable under the Convention. This solution, as the Giuliano and Lagarde Report makes clear, is specifically designed to resolve the question of the significance of the silence of one party with regard to the formation of the contract.