1.

A Polish national (‘the applicant’), who had been a former prisoner in Auschwitz, brought a civil claim against a German newspaper before the Polish courts for having used the expression ‘Polish extermination camp’ in an online article to refer to a Nazi extermination camp built on the territory of (then) occupied Poland during the Second World War. Although that article had been online for only a few hours before it was corrected, the applicant maintains that the online publication has harmed his national identity and dignity.

2.

Do Polish courts have international jurisdiction to hear such a claim? In the main proceedings, the applicant is not only seeking monetary compensation, but also other remedies: a court order prohibiting the publisher from using the expression ‘Polish extermination camp’ in the future and the publication of an apology.

3.

The present case thus invites the Court to specify, once again, ( 2 ) the criteria for the assessment of international jurisdiction for claims alleging infringements of personality rights committed by means of an online publication. However, the present case does so in a rather particular context: the person alleging an infringement of his personality rights was not named in the publication at issue. Nevertheless, it would appear that as national case-law currently stands, the personality rights of Polish nationals include the protection of their national identity, national dignity, as well as respect for the truth about the history of the Polish nation. Moreover, in cases such as the present one, it seems that those personality rights of Polish survivors of Nazi extermination camps are held to be affected by such statements.

4.

Recitals 15 and 16 of Regulation (EU) No 1215/2012 ( 3 ) state that:

5.

Article 4(1) of Regulation No 1215/2012 states that: ‘Subject to this Regulation, persons domiciled in a Member State shall, whatever their nationality, be sued in the courts of that Member State.’

6.

Article 5(1) of that regulation provides that: ‘Persons domiciled in a Member State may be sued in the courts of another Member State only by virtue of the rules set out in Sections 2 to 7 of this Chapter.’

7.

Article 7, which falls under Section 2 headed ‘Special jurisdiction’, of Chapter II of the same regulation, sets out that:

‘A person domiciled in a Member State may be sued in another Member State:

…

8.

As the Polish Government explained in its written observations, Articles 23 and 24 of the Polish Civil Code protect a wide range of personality rights. Article 23 contains a non-exhaustive list of the dimensions of personality rights that may be protected under that provision. Those provisions do not expressly mention national identity, national dignity, or the right to respect for the truth about the history of the Polish nation. However, the Polish Government cited numerous examples from national case-law confirming that those three dimensions are now included within the scope of the personality rights laid down in Article 23 of the Polish Civil Code, as interpreted by Polish courts. ( 4 )

9.

According to the Polish Government, these rights can be infringed through individual attacks and through statements that concern a larger group of people, including the nation as a whole. In order to bring a claim based on such a statement, the claimant must demonstrate that he or she has been individually affected by the statement in question. Such an impact upon individuals has been confirmed, in particular, in the case of former prisoners of Nazi extermination camps, who are in a position to invoke an infringement of their personality rights vis-à-vis descriptions of these camps using the adjective ‘Polish’. According to the national case-law, the use of that adjective places the guilt for the very existence of the extermination camps upon the shoulders of a group of persons who were themselves prisoners of such camps, thus suggesting that the victims were in fact the perpetrators.

10.

Therefore, as confirmed in essence by the referring court while setting out its reasons for requesting a preliminary ruling, ( 5 ) in accordance with national case-law, the personality rights of Polish nationals include the protection of their national identity, national dignity, and the right to respect for the truth about the history of the Polish nation. In this regard, it is considered that the rights of Polish survivors of Nazi extermination camps are infringed by incorrect statements concerning the Nazi extermination camps. As such, those individuals are considered to be entitled to bring an action under national law in a case such as the one in the main proceedings.

11.

The applicant in the main proceedings is a Polish national living in Warsaw. He was a prisoner in Auschwitz during the Second World War. At present, he is involved in activities aimed at preserving the memory of the victims of crimes committed by Nazi Germany against Polish nationals during the Second World War. These activities include, inter alia, participation in meetings with an educational purpose.

12.

The defendant in the main proceedings, Mittelbayerischer Verlag KG, is a legal person established in Regensburg (Germany). It publishes a regional online newspaper in German on the www.mittelbayerische.de website, which is of course also accessible from other countries, including Poland.

13.

On 15 April 2017, an article entitled ‘Ein Kämpfer und sein zweites Leben’ (A Fighter and his Second Life) was published on that website. It describes the fate of Mr Israel Offman, a Jewish Holocaust survivor who was born in Częstochowa (Poland). He was incarcerated in camps in Bliżyn and Auschwitz-Birkenau, Sachsenhausen and Dachau, and worked as a forced labourer in Leonberg and Plattling. After the Second World War, he eventually settled permanently in Germany. The article begins with a story about how, in 1961, after the birth of Mr Offman’s third child, a girl, a registrar in Niederbayern (Lower Bavaria, Germany) refused to register the name that the parents had chosen for their daughter – Faya. The registrar stated that that name sounded too foreign and could not be pronounced in German. The article explains that the parents wanted to name their daughter Faya because that was the name of Mr Offman’s sister, who, as stated in the original text of the article, ‘was murdered in the Polish extermination camp of Treblinka’.

14.

As the referring court notes, it is a historical fact that the camp in Treblinka was a German Nazi extermination camp built within the territory of occupied Poland during the Second World War.

15.

It follows from the order for reference that the original phrase ‘Polish extermination camp of Treblinka’ remained on the website for only a few hours, apparently from 5.00 on 15 April 2017, when the entire article was published online, until around 13.40 that same day, when, following an email intervention by the Polish Consulate in Munich (Germany), the phrase in question was replaced with the following: ‘was murdered by the Nazis in the German Nazi extermination camp of Treblinka in occupied Poland’. A footnote to the article contains a short explanation that the phrase ‘Polish extermination camp of Treblinka’ was originally used in the text but was subsequently corrected.

16.

On 27 November 2017, the applicant filed a civil action with the Sąd Okręgowy w Warszawie (Regional Court, Warsaw, Poland). With his action, he submitted a printout of the corrected version of the disputed publication. He did not specify the circumstances in which he became aware of the publication. The applicant requested that his personality rights, in particular his national identity and national dignity, be protected by:

17.

To justify the international jurisdiction of the Polish courts, the applicant relied on the judgment of this Court in eDate. ( 6 ) The defendant filed a motion for dismissal of the action on the ground that the Polish courts lack jurisdiction. It emphasised that, unlike the situation in eDate, the online article that formed the basis for the action did not directly concern the applicant. The defendant also drew attention to its regional profile and readership range, since its reporting covers the Oberpfalz (Upper Palatinate, Bavaria, Germany) and focuses primarily on regional news. By way of example, the heading ‘Germany and the World’ is only in fourth place on the page menu. The defendant also pointed out that the website exists solely in German.

18.

By order of 5 April 2019, the Sąd Okręgowy w Warszawie (Regional Court, Warsaw) rejected the defendant’s motion for dismissal of the action. That court considered that the conditions laid down in Article 7(2) of Regulation No 1215/2012 were satisfied. It pointed out that between 15 April 2017 and 29 November 2018, there had been more than 32000 visits from Poland to the defendant’s website. Poland was ranked 14th out of 25 countries of origin in terms of visitors to the website. Accordingly, the defendant could have foreseen that the publication could reach readers in other countries, including Poland. It could also have foreseen that the online publication of an article containing the phrase ‘Polish extermination camps’ might be noticed by Polish readers. In view of the online availability of the publication in Poland, and given its content, Poland could be considered as the place where personality rights may be infringed.

19.

On 25 April 2019, the defendant brought an appeal against the first-instance decision before the referring court, the Sąd Apelacyjny w Warszawie (Court of Appeal, Warsaw, Poland). The defendant submitted that Article 7(2) of Regulation No 1215/2012 was applied incorrectly and claimed that it was not reasonably possible to foresee that court proceedings would be brought in Poland. The defendant submitted that, if the content of an article concerns a person other than the applicant or does not concern any particular person, the defendant is objectively unable to predict the court before which it may be sued. The defendant pointed out that the content of the article in question is so distant from Poland that this objectively excludes any reasonable foreseeability of court proceedings in Poland.

20.

By a procedural order, the Sąd Apelacyjny w Warszawie (Court of Appeal, Warsaw) asked the applicant’s legal counsel to describe the factual circumstances of the claim more precisely. The applicant was also invited to state whether he has a command of the German language, when (before or after the correction of the disputed phrase) and how (whether directly online or through third parties) he became aware of the publication. However, the referring court states that those questions remained unanswered.

21.

The referring court has doubts as to whether the requirement of reasonable predictability of jurisdiction, set out in recitals 15 and 16 of Regulation No 1215/2012, cover a situation such as the one in the main proceedings. That court admits that the expression ‘Polish extermination camp’ is likely to elicit a negative reaction in Poland. It may create the false impression with some readers that the extermination camps had been built by Polish nationals and that they were responsible for the crimes committed therein, thus causing offence to those Polish nationals who were themselves incarcerated in extermination camps or whose relatives were killed by the Nazi occupiers during the Second World War. However, the referring court also wonders whether the specific circumstances of the present case may lead to the conclusion that the defendant could have reasonably foreseen that, in connection with the article at issue, an action for the protection of personality rights might be brought against it before a Polish court even though the applicant was not described, directly or indirectly, in the text of the article.

22.

Under those circumstances, the Sąd Apelacyjny w Warszawie (Court of Appeal, Warsaw) decided to stay the proceedings and refer the following questions to the Court of Justice for a preliminary ruling:

23.

Written observations were submitted by the applicant, the defendant, the Polish Government, as well as the European Commission.

24.

This Opinion is structured as follows. I start by recalling the current state of the (case-) law relating to international jurisdiction for damages allegedly caused to an individual’s personality rights under Article 7(2) of Regulation No 1215/2012 (A). I shall then turn to the particular features of the present case (B). Next, I will explain why the issue of whether a person has been named can hardly be a conclusive yardstick for determining whether any damage could have occurred in a given jurisdiction (C). Then, I shall turn to the issue of foreseeability of the forum (D). Finally, I conclude by suggesting that while EU law, as it currently stands, is indeed rather accommodating in terms of international jurisdiction in cases where an individual’s reputation is damaged online, there are nonetheless some limits to the substance of such claims that a national court is obliged to respect (E).

25.

The Court’s case-law on what is now Article 7(2) of Regulation No 1215/2012 has always emphasised the need to preserve the availability of the international jurisdiction for the damage claims of the alleged victim. Thus, the Court has consistently understood that provision as referring both to the place of the event giving rise to the damage, and, to the place where the damage occurred. ( 7 )

26.

With regard to infringements of personality rights, the part referring to the ‘place where the harmful event occurred’ has gradually established two different grounds of jurisdiction.

27.

In Shevill, with regard to printed publications, the Court associated this place to any Member State ‘in which the [allegedly harmful] publication was distributed and where the victim claims to have suffered injury to his reputation’, with the courts of this Member State having jurisdiction ‘solely in respect of the harm caused in the State of the court seised’. ( 8 )

28.

As a result, the so-called ‘mosaic approach’ was born. ( 9 ) It is based on the understanding that an allegedly defamatory publication produces ‘its harmful effects upon the victim’ in all places ‘where the publication is distributed, when the victim is known in those places’ and that, consequently, the courts in these Member States ‘are territorially the best placed to assess the libel committed in that State and to determine the extent of the corresponding damage’. ( 10 )

29.

With the Court’s judgment in eDate, Shevill‘moved’ online. Indeed, in eDate, the Court acknowledged ‘the difficulties in giving effect, within the context of the internet, to the criterion relating to the occurrence of damage which is derived from Shevill’ and ‘the serious nature of the harm which may be suffered by the holder of a personality right who establishes that information injurious to that right is available on a world-wide basis’. ( 11 ) Therefore, the Court decided to adapt its interpretation of Article 7(2) of Regulation No 1215/2012 ‘in such a way that a person who has suffered an infringement of a personality right by means of the internet may bring an action in one forum in respect of all of the damage caused, depending on the place in which the damage caused in the European Union by that infringement occurred’. ( 12 )

30.

This is how the ‘centre-of-interests’ ground of jurisdiction was born. The Court stated that ‘the impact which material placed online is liable to have on an individual’s personality rights might best be assessed by the court of the place where the alleged victim has his centre of interests’. ( 13 )

31.

However, and rather importantly, eDate’s centre of interests did not replace Shevill’s ‘mosaic’ with regard to online publications. It has been added alongside it, as an alternative. Accordingly, the victim of an alleged online infringement of personality rights may choose between seising (i) the court of the Member State where his or her centre of interests is based in respect of all the damage caused, or (ii) the courts of one or several Member States in respect of the part of the damage that has been caused there. ( 14 )

32.

In my Opinion in Bolagsupplysningen, ( 15 ) I sought to convince the Court that such an approach is inappropriate. In addition to the significant practical problems created by extending the mosaic approach to online publications, this choice cannot be justified by any of the principled arguments that are put forward in its favour. In particular, it hardly improves the situation of either of the two parties – except for situations of harassment – but rather drastically reduces, if not excludes altogether, the chances for the defendant to predict where he or she might be sued.

33.

In its judgment in Bolagsupplysningen, the Court decided to maintain that dual approach. ( 16 ) At the same time, however, the Court clarified that only the courts of the Member State in which the claimant’s centre of interests is located may hear claims for the rectification or removal of the allegedly harmful online content. Indeed, as ‘a single and indivisible application’ such a claim can only be brought ‘before a court with jurisdiction to rule on the entirety of an application’. ( 17 )

34.

It would therefore appear that following Bolagsupplysningen, courts with ‘mosaic’, limited jurisdiction can only award ‘divisible’ remedies, whereas courts with ‘centre of interests’, full jurisdiction can award both, ‘divisible’ and ‘indivisible’ remedies. ( 18 )

35.

The Court did not specifically set out which types of remedies fall within which category. However, I would assume that ‘divisible’ remedies essentially mean monetary compensation (and potentially other remedies which are capable of being split along jurisdictional lines), whereas ‘indivisible’ remedies are all those that may be realistically awarded and executed only once, since once executed, they have an impact everywhere (rectification or removal of the information).

36.

Finally, the case-law to date has been rather vague on the exact conditions that must be met so that ‘mosaic’ jurisdiction or ‘centre of interests’ jurisdiction can be established. On the one hand, for ‘mosaic’ jurisdiction, the Court’s case-law mentions two requirements: the content complained of must have been made available in the Member State in question and the claimant’s personality rights must have been potentially harmed in that Member State. ( 19 ) On the other hand, ‘centre of interests’ jurisdiction clearly requires more. However, perhaps due to the fact that it has been introduced and confirmed in cases where the particularly close connecting factor between the dispute and the courts of the place where the damage occurred ( 20 ) effectively led to the ‘victim’s centre of interests’ being located at the victim’s domicile, ( 21 ) the case-law does not provide much detailed criteria in terms of specific factors or other elements that ought to be taken into account.

37.

It is within such a legal context that the referring court enquires, first, whether there can be a centre of interests for a claimant who was not individually identified by the publication at issue. I prefer using that expression rather than the turn of phrase employed by the referring court, which states that ‘the online publication cited … does not contain information relating directly or indirectly to that particular natural person’. Whether or not an applicant has been individually identified in a publication, be it (i) by name, or (ii) by his or her unique personal characteristics or circumstances, ( 22 ) should be a relatively objective factual assessment. The extent to which the information published relates to a particular natural person is a matter of (a rather subjective) assessment taking place on a scale, which already includes elements of evaluation of the information provided, and which soon crosses over into an evaluation of the merits of the case, that is to say the extent to which that person has in fact been affected. ( 23 )

38.

By its second question, the referring court asks about the relevant elements which are to be taken into account when assessing whether there is a sufficient connecting factor for asserting jurisdiction based on Article 7(2) of Regulation No 1215/2012.

39.

In view of the nature of the remedies sought, the referring court enquires specifically about ‘centre of interests’ jurisdiction. Indeed, before the referring court, ( 24 ) the applicant seeks three distinct remedies: monetary compensation, the prohibition of any future use of the contested term by the defendant, and a public apology addressed to the applicant. It is not specified whether the requested monetary compensation is intended to represent the totality of the alleged damage, or only the damage which incurred within the jurisdiction of the Polish courts. However, and in any case, the two other remedies look very much like indivisible remedies – that is to say those that may be awarded only by a court having jurisdiction on the basis of the centre of interests of the claimant. ( 25 )

40.

Nevertheless, it is clear that, on the basis of established case-law of the Court set out in the preceding section of this Opinion, in parallel to the potential ‘centre of interests’ jurisdiction, ‘mosaic’ jurisdiction still remains.

41.

That parallelism has rather important implications. Unless the Court wishes to radically revisit, in the context of the present case, its case-law on ‘centre of interests’ jurisdiction, or ‘mosaic’ jurisdiction, or above all their ongoing parallel existence, set out in eDate and confirmed recently in Bolagsupplysningen, ( 26 ) or, while doing so, perhaps taking a closer look at Bier in the context of online delicts altogether, ( 27 ) it is clear that, irrespective of whether there would ultimately be a centre of interests for an (allegedly) harmed party in a case such as the present one, there will always be ‘mosaic’ jurisdiction, allowing any national court to assert jurisdiction with regard to the portion of the divisible remedies (certainly monetary compensation) relating to the part of the damage that has been caused in its territory.

42.

The present case thus demonstrates, once again, why the ongoing parallelism of both types of jurisdiction poses problems, on at least two levels. On the one hand, it is a somewhat futile exercise to entertain passionate debates on where exactly the centre of interests of a harmed party might be, in a context in which ‘mosaic’ jurisdiction will, in any event, be available in potentially 27 Member States. On the other hand, until now, the centre of interests has in fact been equated with the claimant’s domicile on the assumption that this will be the place where the claimant’s reputation will be affected the most. Within that context, it is not clear what space is left for the discussion of any objective connecting factors of the dispute to the forum of a given Member State, since the concept of the centre of interests has been built around the subjective situation of the alleged victim. ( 28 ) Thus, in practical terms, the only task to be carried out is to determine the victim’s place of domicile, and that effective forum actoris may be rebutted only if that victim has very little social interactions in that place.

43.

I am no admirer of the case-law as it currently stands. ( 29 ) In my view, the Court will be obliged to revisit that case-law one day. However, I do not think that the present case is appropriate in that regard, for a rather simple reason: the sticky issue in this case does not concern international jurisdiction, but rather the substance of the claim and of the content and nature of the personality rights apparently recognised under national law. However, the question of who exactly may reasonably claim to be a victim, and what is the scope of the (personality) rights which may have been infringed by an alleged defamatory online publication, is primarily a matter of merits of a claim, set out by national applicable law. It is not (and in previous cases has not been) an issue of jurisdiction under Article 7(2) of Regulation No 1215/2012 (or its predecessors).

44.

For these reasons, I would suggest that the Court adopt a narrow and minimalist approach this case. In essence, the present case is concerned with one issue: with regard to the indivisible remedies, namely the prohibition on the use of a certain statement in the future and the publication of an apology, is the applicability of the criterion of a centre of interests of a party allegedly harmed by online publication precluded by the fact that that person is not named in the publication at issue?

45.

The defendant and the Commission have submitted, in essence, that the victim of an alleged infringement of personality rights should be able to seise the courts of their centre of interests only if they have been named in the publication in question. Both of these interveners put forward a number of arguments on how to distinguish the facts of the present case from those in eDate and Bolagsupplysningen. They warn of the dangerous expansion of international jurisdiction that could be brought about if the ‘centre of interests’ jurisdiction were to be available not only for individuals clearly singled out by a publication, but also for additional undefined members of national, ethnic, religious, or other groups merely referred to indirectly in an online publication.

46.

In terms of the rule that both parties propose the Court to formulate, the suggested approach is essentially the following: since the applicant has not been named in the publication at issue, the publisher could not have reasonably foreseen that the applicant could be affected by the publication. Thus, in order for the ‘centre of interests’ jurisdiction to be established, it is necessary for the (would be) applicant to be named in the publication.

47.

I would agree with both interveners on the first proposition, albeit from a different standpoint: a reasonable degree of foreseeability for the publisher in terms of where its publication might cause damage is certainly a necessary component. It ought to be the outer limit for defining any place where the damage could occur. However, as I shall explain in the following Section (D), that does not mean, or certainly did not mean in the present case-law, that the international jurisdiction would be limited to places which the publisher would have subjectively foreseen. Moreover, there is a difference between objective foreseeability of the harm that a particular statement might cause and foreseeability of the identity of a specific applicant.

48.

However, with regard to the second proposition put forward by the defendant and the Commission, I do not agree that there is a rule according to which in order for there to be a ‘centre of interests’ jurisdiction of an alleged victim, that person would have to be named in the publication at issue.

49.

To start with, it is true that the ‘centre of interests’ jurisdiction is not available under any and all circumstances. There might simply be no such centre in a particular case. However, that is also true of situations in which persons have actually been singled out in a publication by their name. ( 30 )

50.

With regard to the rest, I certainly agree that there is bound to be a natural correlation: if a person has been named or otherwise clearly individually identified in a publication, it is fair to assume that the publisher should have known or at least diligently have assumed ( 31 ) that such a person is likely to have a centre of interests somewhere. ( 32 ) Conversely, the publisher of content which makes no such reference to a particular person will rarely be in such a position. However, does that preclude the possibility that that publisher may still post content online that is very likely, on any reasonable interpretation, to be perceived as causing damage in a rather clear and foreseeable place?

51.

There is no visible line in the sand that could serve as a clear jurisdictional rule in this regard. However, this is logical in view of the nature of the delict at issue: under normal circumstances, an action for defamation requires a case-by-case assessment of a statement viewed in the light of its context and measured in terms of the negative impact it has had on the protected rights of the individual. Indeed, within such an assessment, some statements are far too remote to cause any discernible damage to the reputation of an individual. Others might cut painfully close, even for persons not individually identified. There is a scale, not hermetically sealed boxes. The following examples may help to illustrate that point.

52.

First, imagine an online publication containing a defamatory statement concerning a person’s close relative: his or her husband, spouse, child or parent. For the purposes of this example, let us assume that the defamatory statement concerns the person’s husband. Thus, imagine a defamatory article published about the husband, where he is named, but no one else is referred to by name. The husband decides, for whatever reason, not to bring an action. Nevertheless, it is clear to the people around him, including his spouse, that he has suffered as a result of that statement. If the spouse decided to bring an action in her own name for non-material damage caused to her by the publication of that article, provided that national law allowed for such a person to have standing, could such a person not rely on her own centre of interests?

53.

Second, what about defamatory remarks made about a group of people, who although not named, might, with some effort, be identified (although the publisher may not be in the position to know them all)? Take, for example, a group of people that are linked professionally or culturally (employees of a certain company, members of a society or a club) or geographically (inhabitants of a given place or town). Imagine an offensive statement made in an online publication about such a group of people.

54.

Third, even further along the scale are members of any larger community, bound together by their common characteristics, be it national, ethnic, race, gender, sexual orientation, and so on. What about members of such a community who consider themselves harmed by certain statements made online? Since they are not individually identified by their names, does that preclude online publications from being perceived as causing them harm? Does that mean that such persons cannot, by definition, rely on their centre of interest?

55.

There are two variables common to all of these scenarios. First, there are no well-defined boxes or borders. Instead, there is simply a fluid, continuum of possible ‘degrees of individualisation’ to be assessed in the light of the infinite factual variety of cases, when looking at a given statement assessed in its context with regard to a particular claimant. Second, the issue of whether in any of these cases a person would have standing to bring such an action, or whether such a person could even have the status of ‘victim’, and with regard to which (personality) rights, is a matter of applicable national law.

56.

Such a case-dependent diversity cannot be effectively supplanted by a rule concerning international jurisdiction stating that if a person has not been named in an online publication, or at least not sufficiently individually identified by it, ( 33 ) the jurisdiction normally based on a preliminary assessment as to whether or not any harm occurred or is likely to occur to that person in that jurisdiction, would no longer be available per se. That would represent an instrument far too blunt and inappropriate for a number of cases. One criterion (a person has been named) would be used as a very unreliable proxy for a very different assessment (a particular publication has caused damage within a given jurisdiction).

57.

Accordingly, I do not believe that it would be advisable for the Court to embrace a jurisdictional rule such as the one proposed by the defendant and the Commission. Rather than trying to draw a clear distinction between cases in which a party has been individually identified and cases in which they have not, the Court should instead recall the need for the objective foreseeability of the forum and jurisdiction, together with the potential criteria that are to be taken into account in such an assessment, an issue to which I now turn.

58.

The key question under the heading of foreseeability is: what exactly should be foreseeable? The place where the damage occurred or might occur (the second prong of Bier in general) or simply the place where an actual victim has his or her centre of interests (in other words, only a small subset of that overall issue of harm)?

59.

The applicant and the Commission rightly point out that the foreseeability of jurisdiction is a general requirement of Regulation No 1215/2012. It is indeed set out in recitals 15 and 16, which state that any deviation from the principle that jurisdiction is based on the defendant’s domicile, ( 34 ) that is to say any alternative grounds of jurisdiction, must be ‘based on a close connection between the court and the action or in order to facilitate the sound administration of justice. The existence of a close connection should ensure legal certainty and avoid the possibility of the defendant being sued in a court of a Member State which he could not reasonably have foreseen. This is important, particularly in disputes concerning non-contractual obligations arising out of violations of privacy and rights relating to personality, including defamation’.

60.

However, a rather long chain of reasoning is needed in order to move from that proposition to the one suggested in the present case: (i) the defendant would not have been able to foresee the jurisdiction of the Polish courts because the publication at issue did not name the applicant; (ii) thus the defendant would have been completely unaware of his identity; (iii) therefore the defendant was not in the position to know his domicile; and (iv) consequently the defendant could not reasonably have foreseen where the applicant’s centre of interests might lie.

61.

I agree that such a chain could indeed be established if one were to reduce the issue of predictability of the forum to the question of the identity and the domicile of the applicant only. That is, however, not the ‘predictability’ that the quoted recitals seem to aim at. That foreseeability was intended to be ensured by a close connection between the action and the court, that is to say an objective connection between the action and the forum in the interest of the sound administration of justice.

62.

I am bound to agree that, as far as the locating of the centre of interests is concerned, objective predictability of what might best be called ‘the centre of gravity’ of a dispute appears to have been reduced to the determination of the place of the domicile of the victim. ( 35 ) However, I do not think that the Court should continue on this path, by creating additional rules that further deviate from what the actual assessment ought to be. Simply put, the reasonable foreseeability of the centre of gravity of a dispute should not be effectively replaced by the publisher’s knowledge of the place of the victim’s domicile. Again, while fully agreeing with the need for the rules on jurisdiction to be foreseeable, I do not believe that such a reduction fosters the type of foreseeability that Regulation No 1215/2012 had in mind.

63.

First, the assessment of any personality rights involves at least two considerations: not only the factual or social situation of the harmed party and the harm caused to that party (thus the side of the victim) but also, and above all, the nature, content, and scope of a particular statement placed in its proper context (the side of the statement and its scope and impact). It is at the cross-section of both these considerations that the foreseeability of jurisdiction ought to be assessed, not simply vis-à-vis the foreseeability or even the knowledge of the victim’s domicile.

64.

In this dimension, I cannot but again ( 36 ) subscribe to the position taken by Advocate General Cruz Villalón, who indeed emphasised the importance of foreseeability and legal certainty in such a context. ( 37 ) The type of foreseeability was nevertheless intended to be ensured by identifying the (objective) centre of gravity of the dispute, and not merely the (subjective) victim’s centre of interests. The learned Advocate General therefore proposed such a ‘centre of gravity’ test to be composed of two cumulative elements, one focusing on the claimant and the other on the nature of the information at issue. ( 38 ) The courts of a Member State would have jurisdiction only if that were the place of the claimant’s centre of interests and if ‘the information at issue [was] expressed in such a way that it may reasonably be predicted that that information is objectively relevant in [that Member State]’. ( 39 )

65.

It is indeed true that to date the Court seems to have focused exclusively on the first of the two factors (the situation of the claimant), while being entirely silent on the second (the nature and the content of the material allegedly causing harm). However, that silence need not necessarily be interpreted as an express rejection of the relevance of the second criterion. It serves to be mindful that in both eDate and Bolagsupplysningen, the claimants in those cases were named in the publications in question. It might thus have simply been assumed that the information was therefore objectively relevant in the place of the domicile of the claimants, and that there was no need to dwell in detail on the content and nature of the material at issue.

66.

Seen in this light, the starting point is the nature, content, and the scope of the given material, assessed in connection with a specific person, that is, the claimant bringing an action. Could a given statement, viewed in its proper context, be seen as causing damage to that claimant in a specific place? The foreseeability aimed at is the reasonable predictability of a given forum in view of a particular statement. It cannot be reduced to the ex ante knowledge of the identity of a particular claimant and his or her domicile.

67.

Second, if one were to accept that it is primarily for national law to define the content of personality rights and any possible impact upon them, and that what matters for a statement to cause harm is not simply the knowledge of the identity of the claimant, but more importantly the content and the nature of the statement, then it is conceivable that a ‘geographically determined’ statement might cause (primarily) offence in the ‘geographically determined’ place to which it refers. If someone posts a statement online alleging that all inhabitants of town X are car thieves or that the only job women of Y excel at is prostitution, would it come as a surprise if the inhabitants of town X or women of Y take issue with such statements? ( 40 )

68.

Third, the issue of any subjective intent of the publisher with regard to the actual publication causing offence is connected to that element. However, it is not necessary that the website on which the infringement occurred is specifically directed towards the public in the Member State in question. Whereas such a subjective intention is required for instance under Article 17(1)(c) of Regulation No 1215/2012, ( 41 ) it would create evidentiary problems if applied outside the area of consumer contracts. It would also be incompatible, as the Polish Government rightly points out, with the different wording of Article 7(2) and Article 17(1), respectively. However, and in any case, it would make little sense in view of the nature of the delict in question in this case. In fact, the Court has already come to a similar conclusion when it refused to apply a similar criterion as a counterweight to ‘mosaic’ jurisdiction in cases concerning the infringement of intellectual property rights. ( 42 )

69.

I would also caution against introducing, in essence, ‘a criterion of intent’ to online torts. ( 43 ) The subjective intent of the publisher at the time of publication, if indeed discernable, may be used as an indication only. It is, however, not conclusive. Instead, what matters is whether, as deduced from a range of objective ‘items of evidence’, it could reasonably have been foreseen that the information published online would be ‘newsworthy’ in a specific territory, thereby encouraging readers in that territory to access it. Such criteria could include matters such as the subject matter of the publication, the top-level domain of the website, its language, the section in which the content was published, the keywords supplied to search engines, or the website access log. ( 44 )

70.

However, since those considerations apply to the impact side of Bier, that is to say, where the damage occurred, it is indeed logical that they focus on the objective, subsequent impact of a given publication from the point of view of the public, rather than being primarily concerned with the original and rather subjective intentions of a publisher. It is from this perspective that, in line with recital 16 of Regulation No 1215/2012, a clear objective connection between the action and the forum ought to be assessed, which then justifies the seising of jurisdiction, as a counterweight to the virtually unlimited geographical reach of online content. ( 45 )

71.

Fourth, with that understanding of foreseeability in mind, some guidance might be offered to assist the referring court with regard to the criteria listed in its second question. Indeed, the first two criteria mentioned by the referring court are of some relevance for that assessment: the intended public and the language of the publication. However, in contrast to what the defendant maintains, and in line with the arguments already outlined above in this section, the audience is an objective one. In other words, the public likely to be interested in such information and likely to have access thereto, but not the public subjectively defined by any business plans, the type and number of headings on a website, or the subjective perception of the publisher.

72.

By contrast, it is difficult to see how the other two criteria mentioned by the referring court would relate to reaching a decision on the issue of international jurisdiction, rather than elements pertaining to the merits of such a claim. This is particularly true of the suggestion to take into account ‘the period during which the online information in question remained accessible to the public’ or ‘the individual circumstances of the [applicant]’. Both of these elements are certainly important for the assessment of the damage that the publication has caused to the reputation of the alleged harmed party. However, their role, if any, in so far as it concerns jurisdiction, is much more limited because Article 7(2) of Regulation No 1215/2012 merely requires that the damage for which the claimant seeks relief has occurred in the forum.

73.

Fifth and finally, all those issues are ultimately for the referring court to assess. However, at the level of international jurisdiction, the issue of foreseeability ought to be properly characterised as enquiring as to whether a particular statement, in view of its nature, context and scope, could have caused harm to a given claimant within the given territory. It thus relates clearly to foreseeability and predictability of the given forum. It should not be reduced to the question of whether a particular publisher knew or could have known the domicile of a possible victim at the time the material was uploaded online.

74.

Assessed in this light and on the basis of the facts presented before this Court, it is indeed difficult to suggest that it would have been wholly unforeseeable to a publisher in Germany, posting online the phrase ‘the Polish extermination camp of Treblinka’, that somebody in Poland could take issues with such a statement. It was thus perhaps not inconceivable that ‘the place where the damage occurred’ as a result of that statement could be located within that territory, especially in view of the fact that that statement was published in a language that is widely understood beyond its national territory. Within that logic, while it is ultimately for the national court to examine all those issues, it is difficult to see how jurisdiction under Article 7(2) of Regulation No 1215/2012 could be axiomatically excluded.

75.

All things considered, it was perhaps not entirely unforeseeable that if harm were to occur following the online publication of a statement such as that at issue in the main proceedings, it could be territorially determined by the geographical or national reference contained therein. In the context of the present case, it is rather surprising that, in view of the statement in question, its nature, duration, and the context in which it was posted online, the individual personality rights of any person could reasonably have been infringed by that statement, or, for that matter, individually affected.

76.

However, those issues in fact relate to the applicable national law and have not yet been examined by the referring court. They concern the question of whether, in the present case, the applicant could successfully claim to have been actually harmed by the use of the expression ‘Polish extermination camp’. This is not so much a question of international jurisdiction as it is a question of substance or the potential impact on that substance, if assessed separately or as a preliminary issue in conjunction with standing.

77.

Admittedly, it is not always easy to maintain a clear separation between the issues related to jurisdiction and those pertaining to the merits of an action. Moreover, in the specific context of online publications and infringements of personality rights, international jurisdiction will, in all likelihood, also determine the applicable law. Unlike in other areas of delict/tort law, there is no harmonised conflict-of-laws rule for disputes relating to infringements of personality rights. ( 46 ) Accordingly, each competent court will apply its own domestic choice of laws rule, which is likely to lead to it applying its own substantive law, particularly because the harm or the relevant part thereof are, at that stage, believed to have occurred within its territory.

78.

That does not mean, however, that ‘anything goes’ at the level of applicable law as far as potential infringement of personality rights by online publication in the internal market is concerned. In lieu of a conclusion, I would like to mention two points which could potentially be relevant for the referring court in this regard.

79.

First, it may be useful to recall that Article 3 of Directive 2000/31/EC ( 47 )‘precludes, subject to derogations authorised in accordance with the conditions set out in Article 3(4), a provider of an electronic commerce service from being made subject to stricter requirements than those provided for by the substantive law in force in the Member State in which that service provider is established’. ( 48 )

80.

In its judgment in Papasavvas, the Court clarified that Article 3 of that directive also applies to news outlets, even if they create their revenue purely through advertisements, and equally to national rules of civil liability for defamation relating to their activity. ( 49 )

81.

Contrary to the submissions of the defendant, the fact that Directive 2000/31 appears to be applicable in cases such as the present one is not an argument against the jurisdiction of the Polish courts. However, it clearly limits any court in the Union having jurisdiction under Article 7(2) of Regulation No 1215/2012 with regard to a provider established in another Member State in terms of the substantive outcome of such a case. Any such court, including Polish courts, must therefore verify that the application of their national law on defamation does not subject the defendant to stricter requirements than those in force in the place where the defendant is established ( 50 ) (in casu, Germany), provided that no exception is applicable under Article 3 of Directive 2000/31. ( 51 )

82.

Second, any decision handed down by a court having jurisdiction under Article 7(2) of Regulation No 1215/2012, that is to say, by definition, in a situation in which the defendant is not domiciled on that territory, would have to be recognised in another Member State, typically in the Member State where the defendant is domiciled, in order to be enforceable against that defendant. In such a case, a defendant could argue that such recognition would be manifestly contrary to the public policy of that Member State, under Article 45(1)(a) of Regulation No 1215/2012.

83.

In such a situation, it would, in my view, be entirely conceivable for a court of the requested Member State to refuse to recognise such a decision, for a very simple reason. The applicable national (case-) law used in order to arrive at such a decision deviates considerably from what could be considered to be part of the common European understanding of personality rights.

84.

There certainly is diversity in the Member States on these issues. However, the defining element of personality rights is that they are, indeed, personal: they are to be assessed individually, contextually, with regard to a specific person and his or her dignity, resulting in a case-by-case assessment of individual concern and harm. The national (case-) law applicable in the main proceedings ( 52 ) appears to be replacing that common understanding with what seems to be a nationalistic and generalised legal presumption, or rather fiction: the personality of a Polish national is composed of his or her national identity, national dignity, or right to respect for the truth about the history of the Polish nation, while a Polish survivor of the Nazi extermination camps is affected by statements such as the one at issue in the main proceedings. ( 53 ) The past collective wrongs of one nation are declared to amount to present and future collective harm to another nation, with affiliation to a nation being apparently more important than the individual.

85.

In addition, there may naturally be further reservations that flow from the public policy of the requested Member State, such as the appropriate balance between the freedom of expression and the protection of personality rights, or proportionality in terms of remedies provided for. ( 54 )

86.

However, the bottom line of this section, as well as of this entire Opinion, is that any reservations to the content of any such decision pertain to the merits of such a case, and not really to the issue of international jurisdiction. I understand why it may be tempting, in the context of the present case, to establish a rule according to which only persons named in an online publication are ever capable of relying on a ‘centre of interests’ jurisdiction. However, as I sought to explain in the previous sections of this Opinion, such a seemingly ‘easy fix’ approach would only create future problems in an area of law that is already under strain.

87.

That said, as I nonetheless also sought to illustrate by this closing section, there is in fact no practical need to create any such ad hoc rules. The issues relating to the problematic substance of any such national decision may validly and precisely be addressed there: at the level of the merits of such a claim, either by the national court deciding on the original claim if that court were satisfied that Directive 2000/31 were applicable to the case before it, and/or, potentially, by a court in any other Member State where the recognition of such a decision were to be sought, triggering the public policy exception provided for in Article 45(1)(a) of Regulation No 1215/2012.

88.

I propose that the Court answer the questions referred for a preliminary ruling by the Sąd Apelacyjny w Warszawie (Court of Appeal, Warsaw, Poland) as follows: