Are You Networked Yet? On Dialogues in European Judicial Networks

This article critically examines the functioning of European judicial networks as one modality of transnational dialogues between judges. In order to provide a conceptual framework, we first explore the meaning of the concepts of 'network', 'dialogue' and 'constitutional pluralism' and find that existing channels of communication between national courts and the CJEU hardly qualify as real dialogues. Then follows an in-depth examination of the practical workings of a number of well-established and active European judicial networks, which bring together judges who are more or less at the same level in their domestic judicial hierarchy and have similar functions in their respective legal systems. This involves determining who participates in judicial networks; how and to what end dialogues are conducted within these forums; and why judges may be motivated to participate in such networks. Using these insights, we argue that networks require a change in legal thinking, which is facilitated in the EU context by the constitutionalism pluralism discourse. Further, judicial networks have the potential to increase the dialogic qualities of the relationship between the CJEU and national courts, while they may at the same time affect the balance within that relationship in favour of the national courts.


Introduction
It is axiomatic that judges in different jurisdictions 'look different and behave differently': 1 training, staffing, jurisdiction, powers, impact and stature differ a great deal. Nevertheless, judges relate better or more naturally to their brethren in other countries than to other domestic bodies or institutions, which they do not want to intrude in judicial business. This is especially so for those judges sharing a common position or role in their legal system. 2 This is captured by the idea of 'transnational judicial communities': judges from different legal systems share common beliefs, values and a self-perception and understanding of their role in the legal system and in society. 3 Judges may share a common interest in the intrinsic value of legal concepts and in the quality of the legal argument and so transnational dialogues may also flow naturally. 4 Doctrine has caught on. In recent years, there have been articles dealing with various modalities of such transnational dialogues. Some authors have written about judicial communities and informal, occasional contacts between judiciaries. 5 Others have explored the 'virtual' or 'indirect' transnational dialogue through case law, 6 with part of the literature documenting how such dialogues take place between the CJEU and the European Court of Human Rights, 7 between national courts and the CJEU, or between European and international courts. 8 In this article, another modality for judicial dialogue is examined: we focus on the practical workings of more or less institutionalized horizontal European networks. The results of this examination are used to reflect on what the rise of judicial networks may mean for the conceptualization of the working relationship between the CJEU and national courts in the European Union, and for the conceptualization of the European legal order as a composite legal order itself.
The article is organized in the following manner. Section 2 explores the meaning of three central concepts: 'network' , 'dialogue' and 'constitutional pluralism' . Clarity as to how these notions are understood is important for both empirical and normative reasons. Section 3 surveys the landscape of existing horizontal European networks, that is, transnational networks which bring together judges who are more or less at the same level and have similar functions in their respective legal systems. The first part of the section explores who participates in the networks and how and to what end dialogues are conducted within these forums. The second part of this section discusses the incentives for judges to participate in networks. In Section 4, the normative implications of judicial networks for the more 'vertical' relationship between national judiciaries and the CJEU are considered.

Some definitional issues
This section considers the definition of three core concepts that will be used in the remainder of this paper, namely 'network' , 'dialogue' and 'constitutional pluralism' . These will be discussed sequentially.
Networks are a well-established part of the social sciences vernacular. Tanja Börzel has offered the following general definition: 'a set of relatively stable relationships which are of non-hierarchical and interdependent nature linking a variety of actors, who share common interests with regard to a policy and who exchange resources to pursue these shared interests, acknowledging that cooperation is the best way to achieve common goals ' . 9 For social scientists, networks are employed as an analytical tool to describe why and how cooperation or interaction between various actors takes place. Thus, within the context of the EU, we observe that some authors have characterized the European Union itself as a 'network form of organization' , a system of 'governance without government' whereby ongoing negotiations take place in policy networks made up of public and private actors at different levels. 10 Others have used 'networks' to explain the workings of European institutions such as the Commission. 11 They focus on the relationship between the Commission and national and transnational interest groups in various policy areas and the influence of these sectoral networks on policy outcomes and changes in European policy-making. Yet 13 much of global governance is conducted through purely transnational networks of governmental officials, and transjudicial communications and global networks of courts are very much included. 14 Networks in the social science sense are thus primarily an explanatory concept -they are used to make sense of real-life events, with a particular focus on explaining how and why actors work together and what relational norms govern their dealings. Consequently, the networks that social scientists write about often tend to be rather informal structures, exhibiting a low degree of institutionalization. The transnational 'judicial communities' mentioned in the introduction to this paper can be considered to be networks in the way that the social sciences use that concept.
More recently, the law has also taken an interest in networks. For both law and social science, the core feature of a network is the same: the existence of a series of interlocking and non-hierarchical relationships between the various actors. Both disciplines also use networks in an instrumentalist fashion. The way they do so, however, differs. Social science relies on the network concept to relate and interpret a social reality. Conversely, the law designs structures that it often also expressly refers to as 'networks' in which the relevant actors are connected through a series of interrelated legal rights and obligations. The law thus actively creates networks that exhibit a higher degree of institutionalization than the social science version.
It should be clear, then, that the social science and the legal approaches are not mutually exclusive. Indeed, and applied to the idea of judicial dialogues, formal judicial networks (especially those set up by the judiciaries themselves) commonly build on pre-existing judicial communities and these judicial communities are in turn reinforced by the well-functioning of these formal networks. In the end, the relationships between the network participants will no longer be grounded only in formal legal rules that mandate interaction, but also based on informal connections, which themselves are grounded in the mutual trust and the development of common behavioural standards so typical of judicial communities.
Reflecting on the concept of 'dialogue' , we now consider some insights from the broader 'theory of institutional dialogue' as developed in Canada 15 in the context of constitutional conversations between courts and legislatures, by authors like Peter Hogg, 16 Allison Bushell, 17 Kent Roach 18 and Luc Tremblay. 19 According to this theory, courts and legislatures in Canada participate in a dialogue regarding the determination of the proper balance between constitutional principles and public policies and, therefore, there is good reason to think of judicial review as democratically legitimate. The dialogue paradigm is thus used as a middle way between judicial supremacy on the one hand, and legislative supremacy on the other. It is an ongoing dialogue because the judiciary does not necessarily have the last word with respect to constitutional matters and policies; the legislature would almost always have the power to reverse, modify, or make void a judicial decision nullifying legislation. Judicial review, accordingly, is not a veto on the politics of the nation, but rather the beginning of a dialogue as to how best to reconcile individual rights with the accomplishment of social and economic policies for the benefit of the community as a whole.
In the same context, Luc Tremblay has made a useful distinction between dialogue as conversation and dialogue as deliberation, in order to indicate the limits of the theory of institutional dialogue. 20 Highly critical of the use of the dialogue paradigm to sustain the democratic legitimacy of judicial review, Tremblay points to the need to pay more attention to the nature of the dialogue, in order for the concept to have any real meaning in the discussion. It is a useful distinction to be made also in the context of judicial dialogue in the context of the relationship between the European Court of Justice and national courts. According to Tremblay, 'in a general sense, a dialogue assumes that two or more persons, recognized as equal partners, exchange words, ideas, opinions, feelings, emotions, intentions, desires, judgments, and experiences together within a shared space of inter-subjective meanings' . 21 Now, dialogue as conversation implies a conversation with no real goal in mind, no specific purpose. There is equality, no hierarchy between the participants, a free exchange of ideas and each of the participants may take something home from the conversation, but the aim is not to achieve a common goal, a decision, or even agreement. It assumes that partners encounter each other in a shared world through a common language and common grammar. This presupposes some form of cooperation. An example of this form of dialogue is the case of the casual evening at home, where one invites friends for drinks and dinner and one friend mentions that she is taking up tango dancing and will rent a villa in Tuscany. This provokes new interests in my mind. I too may consider tango dancing (would that be something for me?) and going to Tuscany for the summer. But there is nothing which and no one who says that I should do so because she is going to. And I may decide to take up Italian lessons and go to Spain for the summer.
Dialogue as deliberation is very different: it is aimed at a specific purpose. It aims at reaching a common agreement, at solving problems collectively and determining collectively which opinion or solution is right or best. Conditions for such dialogue are threefold. First, the participants must recognize each other as equal partners, and are equally entitled to put forward their views, to make proposals, to defend particular options and to take part in the final decision. No hierarchy must confer in advance on one or more of the participants the authority to settle the disagreements. Second, deliberative dialogues must consist in a process of rational persuasion, not coercion, with all partners being open to critical analysis of their views, and none imposing their views on the others. Third, this type of dialogue aims at achieving an end result, and hence participants must justify their positions, in order to defend their position and convince the others. Of course, there are situations where no agreement can be found, and partners must agree to disagree, or compromise. These three conditions could be termed equality, rationality and reasoned agreement.
Of course, expertise and voice may vary from one partner to the other, and each may have a different role to play in the deliberation. Yet, in order to be able even to talk about a 'deliberative conversation' , the three conditions of equality, rationality and reasoned agreement are crucial, otherwise the dialogue remains merely 'conversational' . To be sure, there is nothing wrong with conversational dialogues: they can be very pleasant and, as already said, they can lead to common results (albeit not aimed for). But they should not be confused with deliberative dialogues. It is this kind of deliberative dialogue conducted under the three conditions just mentioned which, according to Tremblay, has special normative value and legitimating force, and which can support the view that judicial review is democratically legitimate as part of a dialogue between courts and legislatures.
There is of course something misleading deceptive about presenting the two prototypes of dialogue as set out by Luc Tremblay, since there are many other types of dialogue between these two prototypes. His approach seems to presuppose a certain positive quality to all dialogue and thereby disregards interactions that are unpleasant or downright antagonistic. He also omits the extreme case where there is a refusal to even engage in a dialogue.
Leaving that point aside and instead moving from dialogues between courts and legislatures to those between just courts, it is furthermore debatable whether the courts involved in the European judicial dialogues are truly equal partners. In theory, there is equality with a division of labour: interpretation for the CJEU and application for the national courts. In practice however, a vertical relationship has developed between the two, with the national courts being bound by the CJEU's case law (unless they make a new reference), the distinction between interpretation and application being blurred, and the CJEU developing as a primus inter pares and behaving as such ('the oracle of Luxembourg'). Bearing those caveats in mind, let us take a closer look at the types of dialogue between national courts and the CJEU. 22 The traditional formal avenue for dialogue between the national courts and the CJEU is the preliminary reference procedure found in Article 267 TFEU. 23 Yet, it can be argued that there is no real dialogue going on when applying the standards just described: the national court refers a question and is then virtually removed from the procedure before the CJEU; there is no further conversation, 24 much less deliberation. The national courts steps in again only after the CJEU has given its response to the question asked. Furthermore, answers from the CJEU are known to be sometimes volatile and unpredictable: at times restricted to the case at hand and almost doing the work of the national court, at other times general and almost laconic, often open-ended. There are plenty of examples where the answer of the CJEU was not helpful (not only for the case at hand but also for future cases). 25 Very few national courts make the effort of making a new reference in the same case where the answer has not been helpful. In other words, they do not engage in a true dialogue with the CJEU. Finally, the dialogue through the preliminary reference procedure is easy to avoid for the national courts, by simply not making the reference. It is well known for instance, that quite a few constitutional courts and other highest courts do not make references to Luxembourg, but rather encourage the engagement of the ordinary courts with the CJEU. The German and Austrian constitutional courts have even made it a constitutional obligation for their highest ordinary courts to refer questions to the CJEU, 26 the infringement of which will lead to the quashing of the decision made by the court not making the reference, and the sending of that decision back for that court to make a reference.
Beside the preliminary reference procedure, dialogues between the national courts and the CJEU have been mostly implicit, informal and indirect. They can, for instance, be conducted through case law. A national court may cite CJEU case law, thereby indicating that it accepts and follows the CJEU, but there is no real exchange of opinions, and no deliberation. 27 The CJEU for its part never makes reference to the case law of the national courts of the Member States. The CJEU may have good reasons not to enter into an explicit and detailed debate with each and every national court making the reference in each and every case, and it may be wise to omit it in all cases. But more importantly, the CJEU has no say over national law: it does not interpret or apply national law, and would be acting beyond the province of its own mandate if it did. The necessary bi-directional element, essential for any type of dialogue, is absent.
When they disagree with the CJEU, national courts will mostly be less explicit, and restrict themselves to 'sending dark signals' to the CJEU. This too can hardly be called a dialogue, but rather consists 22 For a more elaborate discussion of the relationship between the CJEU and national courts, also with a focus on the role played by judicial networks, see M. in positing a diktat versus the other court's diktat. 28 These messages are most often heard at times of Treaty change, when they are concerned with the system as such and do not address concrete cases. Examples in point are the Maastricht Urteil 29 and the recent Lisbon Urteil, 30 or the decision of the French Constitutional Council (Conseil constitutionnel) on the Constitutional Treaty. 31 The CJEU, for its part, sometimes does the same, and engages with the national courts by sending signals or even by holding its silence. 32 However, these silent judgments, as they have been labelled, are difficult to interpret and hardly qualify as an open exchange of views. 'Silent dialogues' may cause confusion, not only for those participating in them, but also for the wider audience of these courts, namely parties, Member States, institutions and the general public.
The 'judicial dialogue' between the CJEU and national courts is also conducted through the extrajudicial writing of judges, advocates general and their clerks who at the same time may be professors and legal scholars. 33 Members of the CJEU have for instance been active in spreading the message of the autonomous nature of EU law, the constitutionalisation of the Treaties (through the case law of the CJEU on which they themselves served) and the development of the CJEU into a constitutional court. They have insisted that the relationship of the CJEU with national courts is one of cooperation, and that the method of European law, despite the European legal order's autonomy, is comparative methodology, anchoring European law in the European legal traditions, and building on an existing ius commune. 34 The message is clear: the primacy of EU law and the final authority of the CJEU do not present a threat to national constitutions and national courts, since the CJEU will base the general principles of European law on the common traditions of the Member States and the national courts are invited to participate. National court judges too have defended the position of their courts in extrajudicial writing, explaining and defending the position of their courts. 35 Finally, dialogue among courts operates through those courts' personnel, through the people who operate the system, people of flesh and blood. Members of national courts visit the CJEU in Luxembourg and vice versa, and judges meet at academic conferences. The European Court of Human Rights organises symposiums, known as 'Dialogues between Judges' , on a yearly basis, at the occasion of the opening of the judicial year in January, when it invites leading judicial figures from across Europe to attend a seminar on issues of common interest. 36 These personal encounters between European and national judges foster trust and respect between the respective courts, offer an occasion to clarify and discuss their case law and allow for an exchange of ideas on legal issues of common interest. 37 To conclude this point on dialogue and overlooking the various channels for dialogue and measuring them against the standards proposed by Luc Tremblay, it would seem that the traditional narrative based on judicial dialogue must be put in perspective. The preliminary reference procedure and case law, the formal and official channels of communication available to courts, seem to score relatively low when seen in the light of the standards of both conversational and deliberative dialogue. These channels of communication hardly qualify as real dialogues -they seem more like a series of monologues. Extrajudicial exchanges, both in writing and via personal contacts, score much higher as they allow for a free exchange of ideas and arguments on a more equal footing and even give room for deliberation on common issues.
Finally, it is important to bear in mind the conceptual context against which these judicial dialogues take place, which we would describe as the rise of the concept of constitutional pluralism. Legal thinking about the relationship between the European and national legal orders has changed fundamentally, and is no longer set in terms of dichotomy, hierarchy, conflict and (cold) war. 38 Rather, the focus these days is on common principles, on the development of a common constitutional space, combined with mutual respect for identity and diversity of the constituent parts. 39 There is a clear move away from hierarchy to heterachy, from conflict to mutual respect and deference, and the seeds are sown for a real dialogue. The insistence of the German Constitutional Court (Bundesverfassungsgericht), in its Maastricht Urteil, on its relationship of cooperation (Kooperationsverhaltnis) with the Court of Justice, was perceived by commentators as insincere, because the debate was still set in terms of final authority and hierarchy. The dialogue paradigm does not really work in such an environment.
Today, the conceptual framework has changed, and it has remarkably done so alongside the rise and fall of the Constitutional Treaty. The Treaty has failed and faded, but the concept of constitutional pluralism is still with us today, and it offers a much better setting for a fertile dialogue. The turn to constitutional pluralism, which provides an alternative to conclusiveness and finality of one legal order over the other in the absence of consensus over the ultimate authority, allows for the development of a real dialogue between the courts. At the same time, it is exactly this dialogue between the national and European courts which confirms and enhances pluralist conceptions of the European Union constitutional order.

Mapping the field
Having laid the conceptual groundwork, this section seeks to provide an overview of horizontal European networks, that is to say transnational networks which bring together judges who are more or less at the same level and have similar functions in their respective legal systems. The world of judicial networks is ever expanding, leaving writers with the somewhat daunting task of devising ways to present this reality to the unknown reader. One approach would be to proceed chronologically. Alternatively, it is possible to think of distinctions that could be made between networks depending on, for instance, whether they are more general or specialist in nature. We have chosen to discard these options and instead make use of four simple questions that paint the most accurate picture of the operation of judicial networks and their impact in actual life. First, for what aims have these networks been set up? Second, how do networks go about achieving these aims -how do dialogues take place? Third, what topics are considered suitable for judicial dialogue? Fourth, why do judges decide to create and participate in transnational networks? We cannot pretend to be exhaustive in covering all the judicial networks currently in existence in this section, if only because the proliferation of networks is taking place at an increasingly rapid pace. We shall therefore focus primarily on the most well-established and the most active networks.

Aims of judicial networks
When considering the aims pursued by judicial networks, it is helpful to maintain a broad distinction between those created by the European legislature and those set up at the instigation of the judges themselves. Starting with the former, three prominent examples are the European Judicial Network, Eurojust and the European Judicial Network in Civil and Commercial Matters. The European Judicial Network (EJN) 41 has its origins in Joint Action 98/428/JHA. 42 Its aim is to facilitate judicial cooperation in the fight against serious crimes such as corruption, drug-trafficking or terrorism. To that effect, Member States are called upon to appoint contact points to the EJN from among the central authorities responsible for international cooperation as well as the judicial and prosecuting authorities active in this field. 43 These contact points act as intermediaries between the competent local authorities by enabling direct transnational contacts where appropriate; providing information concerning the judicial and procedural system of other Member States; and offering practical assistance in preparing and implementing requests for cross-border judicial assistance (such as the execution of a European Arrest Warrant). 44 Established in 2002, the mandate of Eurojust 45 is to enhance cooperation between the competent authorities responsible for investigation and prosecution of cross-border and organized crime. 46 It comprises one national member per state which may, but need not, be a judge: a Member State can also send a prosecutor or a policy officer as its representative. 47 More precisely, Eurojust's objectives are to stimulate and improve the coordination between the competent authorities of the Member States; to improve cooperation between these authorities by facilitating the execution of international mutual legal assistance and the implementation of extradition requests; and to offer other support that will render the investigations and prosecutions by the competent authorities more effective. 48 As can be inferred from its name, the European Judicial Network in Civil and Commercial Matters (EJNCCM) 49 is charged with facilitating judicial cooperation between the Member States in civil and commercial matters. 50 It must do so, in particular, through the development of two detailed information systems: one intended for internal use by its members and the other available to the public at large as a useful point of reference when individuals become embroiled in litigation with a cross-border impact. 51 The EJNCCM's members fall into four categories: contact points designated by the Member States; central bodies active in this field pursuant to Community or international legal instruments; liaison magistrates; and other judicial authorities involved in transnational cooperation in civil and commercial matters. 52 The contact points will assist the other members in ensuring sound cross-border cooperation inter alia by supplying them with the information needed to prepare operable requests for cooperation; seeking solutions to difficulties arising on the occasion of a request for judicial cooperation; and facilitating coordination of the processing of such requests. 53 In sum, we see that the networks set up by the European legislature aim to facilitate and enhance judicial cooperation in the relevant field, in order to improve the functioning of the European judicial system, and to increase mutual trust, which is necessary for systems of mutual recognition to work. Networks which are created by the judiciaries themselves, start from pre-existing mutual trust and a mutual (self-)perception among participants of belonging to the same transnational judicial community. The mandates of these networks vary.
First, there are networks that seek to improve knowledge of other legal systems. For example, the Network of the Presidents of the Supreme Judicial Courts of the EU (NPSJC) 54 aims to foster debate and promote knowledge of other legal systems among its members, so as to 'promote and develop a common legal culture' . 55 The European Judges and Prosecutors Association (EJPA), 56 established in March 2004 by trainee judges and prosecutors from the French National Judiciary School, similarly aspires to develop the knowledge of its members of the legal systems of the Member States to enable meaningful legal cooperation on a daily basis. 57 Finally, there is the Association of European Administrative Judges (AEAJ). 58 An apex organization grouping national associations of administrative courts, 59 the AEAJ encourages improvements in the availability of legal redress for individuals vis-à-vis public authorities and promotes the legality of administrative acts, while respecting the legal cultures in the various national legal systems. As such, this network contributes to the dissemination of information on legal redress in administrative matters by means of an extensive exchange of pertinent case law and legislation.
Second, a significant number of European networks have been set up with the aim of sharing practical experiences and promoting the exchange of ideas. Such exchanges can relate to substantive as well as institutional issues and may be carried out with a view to establishing best practices or a common legal culture. A good example is the Association of the Councils of State and Supreme Administrative Jurisdictions of the European Union. 60 This network was established in 1998 61 to provide its members with practical know-how on exercising their national mandate, with special attention to matters concerning European law. 62 The Conference of European Constitutional Courts (CECC) 63 is another example of a network whose principal function is to share experiences. The CECC brings together judicial bodies that exercise constitutional jurisdiction, 64 in particular reviewing the conformity of legislation with the constitution, and it is charged with providing its members with savoir-faire as regards constitutional practice and case law in relation to selected themes. 65 A more recent example is the Association of European Competition Law Judges (AECLJ). 66 Established in 2001, the AECLJ groups together national judges working in the area of European competition law. 67 It provides them with a vehicle for exchanging views and experiences, which should culminate in 'best practices' for the swift and correct application of European competition rules. A final network worthy of mention here is the European Judicial Training Network (EJTN). 68 While not so much in the business of actually exchanging information, the EJTN carries out important groundwork for successful transnational dialogues by designing programmes and methods for judicial training that foster cross-border judicial cooperation and mutual learning.
Third, there are also networks that aim to advance the interests of their members at the national and/ or European level. Thus, the AEAJ is not only involved in improving knowledge of other legal systems, another of its objectives is to 'promote the professional interests of administrative judges at national and European level vis-à-vis the institutions of the European Union and the Council of Europe' . 69 The same can be said about the NPSJC. In addition to disseminating information about the legal systems of the Member States, this network further sees itself as providing 'a forum through which European Institutions are given an opportunity to request the opinions of Supreme Courts' . 70 Also included in this category is the European Network of Councils for the Judiciary (ENCJ). 71 The ENCJ is not s trictly speaking a judicial network, as it is not composed of judges but of national institutions which are independent of the legislative and the executive and which must support the judiciary in the independent delivery of justice. 72 It has accorded itself the task of acting as a mediator between the European institutions and the national judiciaries. It is also available to provide expertise, experience and suggestions to the EU institutions within the European area of freedom, security and justice.

How do dialogues take place?
The preceding discussion has cast some light on the types of mandates of judicial networks. This subsection takes the logical next step and explains how the networks go about achieving their aims.
A first common approach is the organization of regular face-to-face meetings. For instance, the EJN and the EJNCCM both organize periodic meetings between the contact points during which they are further acquainted with the particularities of each contact point's legal system and debate how to overcome remaining practical and legal obstacles to judicial cooperation. 73 Consider further the CECC: its statute stipulates that it must hold triennial specialized congresses for members to actually share experiences as regards constitutional practice and jurisprudence in relation to a particular theme. 74 Other networks that frequently organize conferences, symposia and seminars are the Association of the Councils of State and Supreme Administrative Jurisdictions of the European Union, the NPSJC, the EJPA and the AECLJ. Finally, some networks also provide their members with the possibility of conducting stages or training. To give an example, in 2010 the Association of the Councils of State and Supreme Administrative Jurisdictions of the European Union organized a working visit of two weeks for 12 judges to a court of another Member State. The invited judges participated fully in the daily activities of the host institution: they attended hearings, deliberated on the verdict and assisted in writing the final judgment. At the end of the training, the participating judges wrote a report about their visit. Stages are often carried out under the auspices of the EJTN, which has moreover been given exclusive responsibility by the EU institutions for implementing the Exchange Programme for judicial authorities.

The English Competition Appeal Tribunal (CAT) played a leading role in establishing the AECLJ (and has also provided its secretariat) and
it is thus no surprise that the president of the CAT, Sir Christopher Bellamy, was elected as the Association's first president. The second approach to judicial interaction relies on the use of sophisticated IT facilities. To be clear, this way of communication can be, and often is, combined with face-to-face meetings. The Association of the Councils of State and Supreme Administrative Jurisdictions of the European Union arguably has the most elaborate information system, which comprises three components. The DEC-NAT databank is a freely accessible, searchable database which holds files on around 20,000 national decisions, translated into English to enhance their usefulness for consultation and reference purposes. 75 Each file contains basic information regarding the case (such as the names of the parties, applicable legal provisions and relevant doctrinal works) as well as an analysis which clarifies the aim of the judgment. The second component is the JURIFAST rapid information system for case law, which contains preliminary references, the CJEU's reply and the final decision by the national court as well as other national decisions applying European law. The information contained in JURIFAST is entered into it directly by the Association's members, with the General Secretariat responsible for translations of descriptions and summaries. This database too is freely accessible to the public in French and English. In addition to these public components, the Association in early 2005 set up an intranet, which allows member courts -both judges and staff -to enter into direct contact and ask concrete questions. This mode of interaction enjoys considerable popularity: according to the Association's website, by 1 November 2010, 234 members had registered and had posted 467 messages on 155 discussion topics. Another network that ought to be mentioned in this respect is the NPSJC, which has developed a common portal of case law that allows its members (including the CJEU) to search all the national case law databases. An online translation tool greatly enhances the usability of the portal. The general public has access to the 'lite version' of this portal, which unfortunately excludes translations of the available judgments. Other forms of IT-based interaction employed by various networks include their website (which often contains a special closed-off section for members), regular newsletters, chatting lists and the EJPA even has its own blog! 76 It should finally be mentioned that there is (much) collaboration between the EU institutions and networks created by the judiciaries. This can take the form of providing financial support. Thus, the Commission regularly co-finances the conferences, seminars and training projects organized by the AECLJ. Alternatively, the networks maintain regular contact with the institutions and ask for their participation in their activities. For instance, the Association of the Councils of State and Supreme Administrative Jurisdictions of the European Union is involved with the European Parliament, in particular the Committee on Citizens' Rights, Justice and Home Affairs, and with the Commission, above all with the Commissioners for Justice and the Legal Service. This collaboration is mutually beneficial for both parties: the quantity, quality and pertinence of the work carried out by the networks will be enhanced and the EU institutions are faced with a few representative interlocutors (as opposed to myriad national courts) and kept informed about the current challenges facing national courts in dispensing justice.

What topics are discussed with peers?
The answer to the question posed in the heading could be 'what topics are not?' . Indeed, a very wide variety of issues is considered suitable for dialogue among courts. Understandably, judges like to share thoughts about institutional and organizational issues -with judicial independence and ensuring efficient and high quality justice ranking particularly high on this agenda. For instance, the 2011 congress of the CECC dealt with the general theme 'constitutional justice: functions and relationships with other public authorities' and among the topics that were discussed was the resolution of organic disputes by this type of court and the enforcement of judgments rendered. Similarly, the NPSJC dedicated its 2008 colloquium to the motivation of judgments and precedents and during their 2010 gathering the judges focused their attention on 'practical aspects of independence of justice' . 77 Since we are discussing European judicial

Incentives to participate in judicial networks
The preceding subsections have all taken as their starting point the fact that judges cooperate with each other in judicial networks and have inquired into the practicalities of this cooperation. In this section, we will take a step back and identify a number of the incentives that (may) prompt judges to participate in transnational networks in the first place. The reasons that led to transnational judicial interaction may go some way towards explaining the success (or lack thereof) of networks. We may expect judicial attitudes to differ depending on whether there is a strongly felt practical need for sharing experiences and exchanging information or whether the network has been put in place primarily because judges did not want to be left behind in the face of other courts uniting. Further, the topics considered suitable for judicial dialogues are in large part decided with reference to the incentives of judges to participate in networks. Two more general comments ought to be borne in mind. First, obviously, the reasons for engaging in judicial dialogues are not (necessarily) shared among all network members, or among networks for that matter. Second, networks -including the judicial variety -tend to be dynamic structures. As networks evolve, so too may the incentives to join them: the motives that spurred on the founding network members may not be the same as those that persuade latecomers to apply for membership. It is possible to maintain a broad distinction between what we shall call 'practical' incentives and 'authority' incentives. It will be common, however, for judges to be influenced by a mix of both.
'Practical' incentives are concerned with the pressures of globalization for the actors involved in the judicial business. Participating in networks is not so much a matter of choice but of necessity to make sure judges are still able to get the job done properly in a changing environment. The internationalized nature of litigation makes knowledge about other legal systems a prerequisite to being able to dispense justice in an individual case. These considerations take on an extra dimension within the EU, with its quest to establish a truly internal market and the ever-expanding reach of EU rules. Judges dealing with commercial law, administrative law, consumer law and increasingly also civil and criminal law need to cooperate with out-of-state counterparts simply to make principles such as mutual recognition work. As we have seen, this is precisely the reason for the European legislature to take the lead in establishing judicial networks, as happened for instance with the creation of the EJN and its role in the execution of the European Arrest Warrant.
' Authority' incentives denote membership of judicial networks for the purposes of borrowing or enhancing the participating court's authority and legitimacy. This can take a number of forms. Judges may be motivated by a desire to exercise judicial leadership in a changing institutional setting. In particular,courts that enjoy a certain status within their national legal order could be interested in carrying over this domestic stature and position -and perhaps even exporting domestic solutionsinto the international arena. This strategy acquires added prominence for judges involved in networks that expressly seek to cooperate with the European institutions: there may be opportunities to influence the European agenda in accordance with domestic interests. Alternatively, judges may perceive network membership as a helpful tool in strengthening their legitimacy vis-à-vis various interlocutors. They then use their participation in a transnational network to build credibility and authority outside their own legal system, which is subsequently imported back into that national system. For instance, within the context of EU law the principles of direct effect and supremacy have increased the prominence and relevance of ordinary courts. Highest courts may perceive this as a threat to their authority and resort to judicial networking with peers in other Member States to reinforce their domestic reputation and standing within the domestic legal order. 78 Alternatively, courts could be motivated to enhance their position in relation to the executive and the legislature. This incentive may for instance persuade courts in emerging democracies to find 'strength in numbers' to defend their newly acquired independence. Separation of powers' considerations have also influenced the creation of the CECC: it is in the nature of the work of its members that relationships with the other branches may be tense or politically charged and awareness of how other courts have dealt with similar challenges may be reassuring. Finally, judgesnotably those in countries in transition -may take part in judicial networks for reasons of international profiling and credibility: participation is then an instrument to reassure foreign investors that the country adheres to the rule of law and that property rights, legitimate expectations and the like will receive proper judicial protection.
Some practical factors should also be mentioned that, incentives aside, are relevant in explaining the creation and success of networks. A first such factor is language: the networks discussed in the previous subsections all conduct (the majority of) their activities in English, meaning that for a judge to effectively participate in, and benefit from, those networks she or he must be proficient in English. This explains the (continued) existence of judicial networks with a language, as opposed to more functional, dimension: the Association des Cours Constitutionnelles ayant en Partage l'Usage du Français, bringing together constitutional courts from French-speaking jurisdictions around the world, is just one example. 79 Secondly, we should remember that judges are also citizens and, as such, representatives of their countries. This can give rise to a more political dialogue, whereby the conversations that take place are not just about quality of legal reasoning, but about what judges want -for instance, advocating a certain way of dealing with a legal issue. A third factor is the prominence of personalities. Specific persons may play a leadership role in a network, for instance because they are personally convinced of, and committed to, the benefits of transnational judicial collaboration or because they are great judicial minds, thinking of innovative solutions to both old and novel issues. Having such personalities on board may challenge the traditional hegemony of courts in the larger Member States: big personalities from small Member States can become leaders instead of followers. 80 Consider for instance the proposal of the Association of the Councils of State and Supreme Administrative Jurisdictions of the European Union for revision of the preliminary reference procedure: the whole process was instigated by the Dutch Council of State which galvanized the other members -not a court from a country one would normally expect to take the lead. One element to consider is of course what happens once these personalities leave the network and we limit ourselves here to observing that there is a clear need for empirical work in this respect.

Europe as network? Some normative implications
What does the rise of horizontal judicial networks mean for the European judicial order, in particular the working relationship between the CJEU and the national courts?
A first point relates to the network concept from the perspective of (constitutional) lawyers. As explained in Section 2, the concept of 'network' has been used to describe and understand a changing legal reality, and is popular in political science. This in part may explain its transition to the legal field. However, the popularity of networking often makes lawyers feel uneasy, and this is all the more the case when it concerns judicial networks. Judges, in our constitutional traditions, speak through the 78 In this respect, consider coherence as a legitimating factor as well: if national judicial decisions are seen to fit into a broader transnational setting and case law, this too could result in an increase in legitimacy. 79 The language factor may gain in relevance if there is also a shared legal tradition. Think for instance of the interactions between courts in common law jurisdictions (notably those belonging to the Commonwealth) and the central role played by the Privy Council in forging a judicial community among them. 80 In this sense, networks may also have an effect on the practice of citing judgments from other courts (the dialogue through case law) by making smaller jurisdictions that one might not normally refer to more interesting because they expound the more stimulating judicial visions or ideas.
judgments they render. They hide their personality behind their robes, and are not considered to be led by personal beliefs and convictions. They must be independent and derive their authority and legitimacy only from the law they pronounce. The network notion is perceived to offend against traditional notions of judicial independence, legal order, hierarchy, ultimate authority and uniformity. Moreover, the networking phenomenon raises fundamental practical questions regarding enforcement and compliance of best practices agreed within a network: are mutual trust and social processes really sufficient in this respect? And perhaps even more important, how do we feel about judges using networks to develop best or common practices? In sum, networks require a change in legal thinking. In this respect, Ost and Kerchove, amongst others, speak of a 'paradigm shift' . 81 Within the EU, and in relation to judicial dialogues more particularly, the constitutional pluralism discourse helps to make this transition, given its insistence on moving away from final authority (and the combative relationship this entails) to a more non-hierarchical (and cooperative) approach. Second, we should note that networks thus require a 'paradigm shift' , yet at the same time they are also the catalyst for this paradigm shift. Judicial networks offer alternative avenues for dialogue, which may support the formal legal channels such as the preliminary reference procedure and the traditional dialogue through case law. Misunderstandings and blockages in those dialogues may be mitigated in the networks. They offer judges the opportunity to discuss face to face, without having to take account of the other participants in the wider legal debate (such as governments, legislatures and other stakeholders), using the law and legal principles as their common language. Networks, we submit, contribute to building mutual respect between courts from different legal orders, and can be used as a forum for deliberation, an advantage that the traditional channels for dialogue do not offer. It is only under conditions of mutual respect and open conversation that a pluralist conception of the European constitution can flourish. There is, however, also something deeply disturbing in this banal finding: the law needs structure, order and coherence, and courts are usually cast as the guardians of that order.
Networks thus have the potential to increase the dialogic qualities of the relationship between the European and the national courts. At the same time, they may affect the balance within that relationship. We have already pointed out how judges from smaller Member States may become leaders in the dialogue -like in the example of the Dutch Council of State and the proposal to revise the workings of the preliminary reference procedure. Interestingly, another network (the NPSJC) joined the discussion. Their initiative was taken over at a late stage by the CJEU, which hosted a symposium with members of both networks to discuss the proposals at the premises of the Court in Luxembourg in March 2009. The position of the CJEU in these judicial networks and its engaging with them will be of great importance in this new environment where national courts are adopting a proactive approach in debating their relationship with the CJEU. The CJEU is a member of the Association of the Councils of State and Supreme Administrative Jurisdictions of the European Union and participates in many activities of this and other networks. Yet, there seems to be room for improvement when considering the CJEU's activities. It is important for the CJEU to be seen to be actively engaging with the national courts, supporting 'bottom-up initiatives' and to be taking the lead where necessary. This is all the more so when it concerns the European legal order and the European judicial system itself. Take for instance its communication strategy. The March 2009 symposium, mentioned earlier in this paragraph, is announced on the Court's website, 82 but one looks in vain for a press release or more substantial document detailing what was discussed and what was decided on. Leaving aside the dialogue with national judges through the preliminary reference procedure, the CJEU has mainly engaged with its national counterparts through its website (for instance the Reflets that provide overviews of recent national and international case law, legislation and some articles). In this respect, consider the dialogues among courts organized by the European Court of Human Rights, as alluded to earlier, for a somewhat different approach. 83 The Strasbourg court has an annual 'dialogue between judges' on a particular theme, which is launched at the opening of the judicial year in January, with judges from all of the Contracting States gathering in person. 84 The CJEU's communication strategy may well require adaptation to the new dialogic environment and the context of pluralism.