‘The different sets of ideas at the back of our heads’: Dissent and authority at the International Court of Justice

Abstract Additional opinions (AOs) – dissenting opinions, separate opinions and declarations, whether authored individually or jointly – are a distinctive characteristic of the ICJ’s jurisprudence. Few decisions of the International Court of Justice (the Court, ICJ) are delivered without any additional opinion attached to it. Yet, despite their ubiquity, there is still significant disagreement as to their relationship to the authority of the Court and its decisions. Although this disagreement is commonly attributed to the different approaches and attitudes traditionally associated with the ‘civil law’ and ‘common law’ traditions, few ask specifically why those traditions take the approach they do, and even fewer consider the appropriateness of the extension of those attitudes to the ICJ, which of course is neither ‘civil law’ nor ‘common law’. In this article, using the work of Mirjan Damaška, I offer a contextually coherent and contextually contingent understand of the theory and practice of additional opinions at the ICJ upon which engagement with this practice – by judges, scholars and practitioners – can be premised. This effort to understand the relationship between additional opinions and institutional authority will, by its very nature, lead to a broader enquiry into the very nature of institutional authority at the ICJ. Having explained the importance of AOs to the structural integrity of the Court’s authority, I will close this article by highlighting the role of various stakeholders when engaging with that practice to ensure that their institutional function is discharged.


Introduction
Commenting upon the ICJ's judgment in Oil Platforms, 2 Jorg Kammerhofer expressed regret at the number of individual opinions attached to the Court's judgment. 3 Opining that '[t]he *Assistant Professor, School of Law, University of Nottingham. I thank Olympia Bekou, Edward Goodwin, Dino Kritsiotis and Marko Milanovic for comments on previous iterations of this work, and the two anonymous reviewers for their thoughtful and constructive comments. appending of individual opinions simply is not healthy', 4 he lamented the negative effect that the publication of such opinions has on the authority of the Court and its judgments. Kammerhofer is not alone in holding this view, nor was he the first to express it. 5 At the same time, supporters of the right of judges to issue individual opinions assert their complementaryeven constitutiverelationship with institutional authority. 6 Despite the longevity of the theoretical debate as to the desirability of permitting additional opinions (individual opinions and joint opinions), 7 existing attempts to address this disagreement fall short of offering a resolution. 8 The differences in the attitude towards AOs are commonly attributed to the different approaches traditionally adopted within the civil law and common law traditions. 9 Assuming that it is possible to identify uniform 'civil law' and 'common law' approaches, 10 few ask specifically why those traditions have taken the approach that they do. 11 Even fewer engage in a sustained critique of the extension of the assumptions associated with those traditions to the international sphere(s), which is(are) after all neither. One writer avoids the apparent impasse between the civil law and common law traditions by diverting attention to particular characteristics of international law that may work in favour of either permitting or prohibiting AOs. 12 However, there has been no attempt to break that impasse and, despite many critics observing that dissent undermines the authority of a given court or tribunal and its decisions, only one writer has attempted to offer an explanation of how dissent offends judicial authority. 13 My objective in this article is to fill this gap in the existing literature and to offer a way forward in the debate as to the relationship between judicial dissent and judicial authority; one that is not hamstrung by the 'different sets of ideas at the back of our heads'. In this article I offer a contextually coherent and contextually contingent understanding of the theory and practice of AOs at the ICJ upon which engagement with this ubiquitous practiceby judges, scholars and practitionerscan be premised.
Thus, while the focus of this article is upon AOs at the ICJa practice that, in the wider scheme of international law may be considered a narrow subject of enquirythe significance of this article is much broader. Firstly, any actor who engages with the jurisprudence of the Court is confronted with the multiplicity of opinions that accompany the judgment, order or advisory opinion of the 4 Ibid., at 716. Court. The interpretation of the significance of those opinions willto varying degreesbe influenced by those different sets of ideas at the back of their heads. While scholars of the ICJ and those closely involved in the Court's activities may have a greater understanding of how such opinions should be understood, the Court's audience is broader and encompasses a range of substantive areas, owing to its systemic role and its general jurisdiction. Secondly, AOs (and their relationship with judicial authority) constitute a frame through which the nature of ICJ authority more generally can be scrutinized. While here, the specific purpose of our analysis of institutional authority is to situate AOs within that authority, the understanding of the nature of that authority that is developed can inform how we think about other aspects of judicial procedure and practice. Thirdly, the theoretical framework presented in this article is of relevance to other (international) courts and tribunals. When applied to those institutional contexts it may assist our understanding of how aspects of their judicial practice and procedurewhether that be the practice of delivering AOs or other practicesfit within the particular structure of authority therein.
In the first half of this article I will focus upon the authority-based critiques of AOs and demonstrate how they are premised upon an understanding of the nature and structure of judicial authority that is not reflected in the ICJ. Using Mirjan Damaška's analytical framework for the comparative study of justice systems, 14 I will demonstrate that rather than being a 'hierarchically'-organized system, wherein AOs are inconsistent with the structure of authority, the ICJ is more accurately characterized as a 'co-ordinately' structured system. Within co-ordinately structured systems, not only are AOs consistent with institutional authority but can also be constitutive of it. I will use the remainder of this article to articulate a theory of the institutional function of AOs in the ICJ, explaining along the way why authority-based critiques are misconceived when made in the ICJ context. At the core of this theory is the notion that, absent the institutional checks and balances upon the exercise of judicial power that exist within hierarchically-structured systems, 15 AOs within co-ordinately structured systems such as the ICJ constitute an essential mechanism by which the parameters of the Court's authority for any given function can be appraised and established. Thus, to the extent that authority-based criticisms proclaim the existence of a tension between the authority of the individual judge (manifested in his or her individual opinion) and the authority of the institution (manifested in the Court's judgment) they are accurate. However, to characterize the existence of this tension as negative per se is misguided. More accurately, the relationship between AOs and the Court's judgments outlined here mirrors the dynamic between the individual judge and the collective in the context of the Court's internal deliberations and judgment-drafting.
Having explained the importance of AOs to the structural integrity of the Court's authority, I will close this article by highlighting the role of various stakeholders when engaging with that practice to ensure that their institutional function is discharged. I call upon both judges and participants in the international law-interpreting communities to exercise greater mindfulness of both the institutional functions of AOs and their responsibilities when engaging with that practice to ensure that those functions are properly discharged.

Additional opinions and judicial authority
Before addressing the authority-based criticisms of AOs and explaining why their application to the ICJ is misguided, to avoid the same ambiguities that plague existing discourse around AOs I will outline what I mean when referring to the concepts of 'additional opinions' and 'judicial authority'.

Additional opinions
Additional opinions are an established and familiar aspect of the international jurisprudential landscape. They are written texts attached to a court's decision under a plethora of familiar labels: 'dissenting opinions', 'separate opinions', 'minority opinions', 'individual opinions', and 'declarations'. They can be authored by individual judges or jointly, by two or more judges. 16 ICJ judges have confined the labels they use to those found in the Court's Statute and Rules of Procedure (English-language and French-language as appropriate). Thus, ICJ AOs exist under the Englishlanguage labels of 'dissenting opinions', 'separate opinions ', and 'declarations'. 17 Irrespective of the label assigned to them, all AOs share the purpose of being a platform upon which judges can express their personal views on aspects of a case, as distinct from the view of 'the Court', as manifested by the Court's decision.
The right to issue AOs is enshrined in the Court's Statute. 18 The culture of practice that has crystallized around that right interprets it as being broad and discretionary. The consequent polyphonic nature 19 of the Court's jurisprudence has become one of its defining characteristics. AOs arein practice and not without controversyused to address any matter of law, fact or policy that the authoring judge(s) deem(s) to have been raised by the case or decision at hand and pertinent to addressing and resolving the issues raised by the dispute as understood by the authoring judge. While the majority of AOs issued are used to offer focused responses to particular aspects of the Court's judgment, there is a tradition among some judges to use their individual opinions to issue lengthy and comprehensive alternative judgments or wide-ranging and general discussions of areas of law raised by the dispute at hand. 20 There has been judicial and non-judicial support for confining the use of AOs to explaining the authoring judge's disagreement or departure from the Court's judgment, 21 or limiting the content of the opinion to the 'framework of the Court's opinion'. 22 The most extensive elaboration of this latter 'restrictive theory' of AOs was elaborated by then President Sir Percy Spender in response to the nine AOs attached to the Court's highly contentious second phase judgment in South West Africa. 23 While this theory gained some traction at the time, 24 the view was met with strong opposition by those judges whose opinions President Spender's theory would preclude, on the basis that to limit AOs in this way would prevent judges from expressing their disagreement with the fundamental logic of the Court's 16 While many use 'individual opinion' to refer to both individually-authored and jointly-authored opinions on the basis that 'individual' can refer to the individuality of the authorship as distinct from the collective of the Court, 'individual' also describes the number of authors of the opinion (a single (individual) judge as opposed to multiple judges (joint opinion). Thus, for the avoidance of doubt and to the risk of confusion with the French-language label of separate opinionsopinion individuelle -I use 'additional opinion' unless otherwise specified. 17 The French version of the ICJ Statute refers to opinion individuelle, opinion dissidente, and déclaration.

18
Art. 57, ICJ Statute; Arts. 95(2) and 107(3), Rules of Court. 19 Or, more accurately in some cases, 'cacaphonic' nature. 20 On the current bench, the opinions of Judge Cançado Trindade would be a prime example of such opinions. Previously, Judge Alvarez held the view that it was necessary for at least one of the judges to provide an extensive review of all the legal issues raised by the case. judgment. 25 Consequently, while there may be a cultural preference for limited opinions, in the spirit of accepting minority perspectives, judges have remained free to use their opinions however they wish.
AOs have been issued in connection with all forms of Court decision: merits and preliminary judgments, advisory opinions, and orders. The labels attached to AOs are typically associated with a spectrum of disagreement, 26 with dissenting opinions on the one end and representing the strongest degree of disagreement, declarations on the other, 27 and separate opinions somewhere in the middle. 28 However, the accuracy of such a linear conceptualization of the typology of AOs should be questioned. The determination of where on that spectrum a given AO lies rests with the individual author, and factors that may influence their labelling choice typically include the nature, scope, and intensity of the disagreement expressed in their AO. 29 Some opinions may consist of aspects that are conventionally understood as 'dissenting' in character and aspects that are conventionally understood as 'declaratory' or 'separate' in nature. 30  Although Arts. 95(2) and 107(3) of the ICJ Rules of Court suggest that the purpose of declarations is for bare statements of assent or dissent, it is common practice for judges to issue substantive, and sometimes lengthy, opinions under the label of 'declaration'.

28
Even conceived so broadly, one could argue with this conception; with some of the most profound disagreements with the Court's judgment found in AOs labelled as 'declarations'.  non-substantive factors may also influence the choice that judges make. 31 For these reasons, the line between different categories of opinion has been described by one former judge of the Court as 'indeterminate'. 32 While the nature of the disagreement expressed will almost certainly have implications for institutional authority, and while the labels ascribed to AOs may themselves have implications for institutional authority through what they connote, 33 labels alone are of limited reliability as objective descriptors of the substantive nature of the disagreement contained therein.
With judges from all backgrounds and legal cultures having embraced the practice, and with few judgments and advisory opinions having been issued without a single AO attached to them, the practice of issuing AOs is an ingrained characteristic of the ICJ's culture. Yet, there is not a clear and shared understanding of the significance of that practice. And, although the notion of authority is frequently invoked when discussing the effect of AOs by both critics and defenders of the practice, what aspect of the Court's authority they risk offending or enhancing, and how, are rarely explained. Consequently, existing attempts to evaluate those views on their own terms prove fruitless. With this in mind, the following section lays out the understanding of 'judicial authority' upon which my argument is premised.

Judicial authority
For the purposes of this article 'authority' is understood in terms of Max Weber's conception of authority or domination, that is, the 'legitimate exercise of power'. 34 Power denotes 'the probability that one actor within a social relationship will be in a position to carry out his or her own will, despite resistance'. 35 'Legitimacy' is understood in the sense of political legitimacy, namely 'the process through which both political power and obedience are justified'. 36 On this basis, legitimacy is defined as 'the governed recognizing the right of the governors to lead and, to a certain extent, their entitlement to the perks of power'. 37 According to this view, 'what makes a certain practice of power legitimate is the process through which authority justifies its exercise of power and gains social acceptance'. 38 There exist a multitude of conceptions of legitimacy premised upon different processes or bases for justification. When speaking of institutional (judicial or otherwise) legitimacy, these conceptions may be grouped into three clusters: consent legitimacy, input 25 July 197425 July , [1974  The use of the language of 'dissent' can invoke the political connotations associated with that language, in turn influencing how the opinion, its author, and their relationship to the Court's judgment is perceived. See R. Jennings, 'The Internal Judicial Practice of the International Court of Justice', (1988) 59 BYBIL 32, at 46, explaining how some judges avoid invoking the language of 'dissent' to describe their opinions; Guillaume, supra note 31, at 433, explaining how the author preferred to style their opinions as 'declarations' in order to avoid having to invoke the connotations associated with 'dissenting' and 'separate'. legitimacy and output legitimacy. 39 Familiar conceptions of consent legitimacy may include state consent and democratic legitimacy. 40 By contrast, input legitimacy encompasses different dimensions of process or procedural legitimacy, including deliberative legitimacy, participatory legitimacy, representative legitimacy, independence, accountability, compliance with 'due process' norms, legality, compliance with human rights, or the attributes of officials (e.g., expertise, experience, personal characteristics, impartiality, independence, individual values). 41 Finally, output legitimacy concerns results-based legitimacy and the extent to which the outcomes of the institutional process achieve the functions attributed to the institution. 42 While it is impossible to be certain what 'ideas at the back of their heads' animated existing contributions to discourse around AOs and judicial authority, the application of the conceptual framework offered in this section to those contributions demonstrates how discourse around AOs and authority can be seen as implicating and interacting with a number of these conceptions of normative legitimacy in different ways.
This conception of authority applies equally to judgments of the Court and to AOs. As platforms for the public expression of individual judicial views AOs are a vehicle through which individual judges can persuade or assert influence in the public sphere. Consequently, they are expressions of individual judicial power, andto the extent that they are accepted as legitimateexpressions of individual judicial authority (individual opinions) or the pooling of the individual authorities of multiple judges (joint opinions). As such, discourse on the relationship between AOs and judicial authority is one aspect of the broader issue of the interaction between individual judicial authority and institutional judicial authority, and more broadly, the role of the individual judge within international law. While the focus here is externally orientedi.e., the authority in the public sphereas Section 5 illustrates, AOs (for consumption in that public sphere) also strike at the heart of the dynamic between the individual judge and the collective in the course of internal deliberations. 43 Moreover, conceiving AOs as expressions of individual judicial authority helps us focus upon those attributes of AOs that distinguish them from other expressions of disagreement with a Court's decision, such as criticisms issued by individuals not holding the office of judge of the Court (e.g., state representatives, scholars, civil society actors). The capacity in which AOs are issued grants them what Weber referred to as 'charismatic authority' 44 such that the contentiousness of AOs lies not simply in the fact that they are the expression of difference or of disagreement, but rather in that they are authored by a judge, and specifically a judge who participated in the case at hand. 45 Authority, and specifically its quality of being an accepted expression of power by a social actor, is particularly important in the context of international justice. Given the geopolitical realities of the international spherethe primacy of the state as the principal unit of action and the dependence of international courts and tribunals upon the consent and co-operation of states to functioninternational courts and tribunals have weak 'power', that being the ability to enforce their will despite resistance and without consent of the affected parties. Thus, international courts and tribunals rely especially upon the acceptance of their legitimacy as institutions and the legitimacy of their decisions to secure the compliance and co-operation necessary for their operation and the effective discharge of their functions. 46 Against this backdrop, the lack of sustained study of the institutional and systemic implications of AOs upon institutional legitimacy (and in turn, authority) is surprising.
3. The perceived negative effect of additional opinions All parties to the Court's Statute have accepted the formal criteria for the legitimacy of decisionmaking, including the quorum for decision-making, 47 the publication of the names of the judges who participated in the case, 48 and the right of judges to issue AOs. 49 Nevertheless, the sentiment that AOs are harmful to the Court's authority is one that has persisted since the drafting of the Statute of the Permanent Court of International Justice (the ICJ's predecessor), 50 and prior to that the negotiations around the creation of the Permanent Court of Arbitration. 51 In this section I elaborate upon this view, before using the remainder of the article to explain why these sentiments, at least premised upon the assumptions that they appear to be, are misguided. This view is epitomized by Kammerhofer who states that in the absence of 'a secure political organization, a separation of powers, and effective enforcement of its judgements by the executive branch' the ICJ depends upon 'the persuasiveness of its pronouncements' and consequently it 'cannot afford to have "in-house", "official" critics'. 52 Although domestic courts may face similar difficulties when faced with inducing compliance by the Executive, international courts and tribunals lack the tradition of acceptance and habitual respect for their decisional authority typically associated with domestic courts. National judiciaries are able to establish their credentials as competent decision-makers in the context of 'mundane' or non-controversial cases. 53 However, international courts and tribunalsparticularly the ICJoften deal with highly politicized and much contested 'big cases', placing judicial authority under further strain. 54 Yet, it is often these very casesperhaps, in part because of their magnitude or sensitivitythat are accompanied by the greatest number of AOs. 55 Thus, concern as to the fragility of international judicial authority is understandable. Using the conceptualization of authority set out in the previous section, the 46 Bodansky, supra note 39. remainder of this section will examine how 'in-house' criticisms in the form of AOs can be perceived to undermine institutional legitimacy.

Output legitimacy
Whether speaking of the Court's decisional authority or of its interpretive authority, authoritybased arguments against AOs may be underpinned by the belief that the revelation of the existence of disagreement between judges casts doubt upon the substantive correctness of the Court's decision. From the perspective of the Court's dispute resolution function, the publication of disagreement between the judges on the applicable law and how it applies to the facts of the dispute weakens the Court's claim to have resolved the dispute. That then diminishes the perceived output legitimacy of the Court's decision and likelihood of compliance with it by the parties. Within those judicial systems that provide for appellate review of decisions, AOseven those that disagree fundamentally with the Court's judgmentmight encourage disappointed parties to continue engagement with the judicial process by seeking appellate review, equipped with the AOs. 56 In the absence of any system for appellate review at the ICJ, AOs may have the opposite effect: they may validate a party's sense of grievance and reinforce any decision not to comply with the Court's decision. 57 However, this argument is not limited to the (non)compliance decision. Owing to the essentially consent-based nature of the Court's jurisdiction as defined by Article 36 of its Statute, states have the opportunity at an earlier stage of proceedings to determine whether they will accept and comply with the Court's judgments and orders in any case that arises. 58 Thus, when considering the effect of AOs upon the Court's decisional authority, not only is it necessary to consider their impact upon the decisional authority of the particular judgment to which they are attached (particular decisional authority), but also their impact upon the Court's general decisional authority and the manner in which states engage with any proceedings brought before the Court.
From the perspective of the Court's interpretive authority, although the ICJ's judgments do not constitute a formal source of international law, its decisions nevertheless constitute a subsidiary means of determining the law. 59 In light of the diversity of the Court's composition, intended to be representative of the international community of states, an articulation of a rule of customary international law by the Court may hold a high degree of authority for this purpose. 60 Added to this institutional and systemic position of the Court as the 'principal judicial organ of the United Nations', belief in the Court's authoritative potential in terms of the clarification and development of international law has long been held. 61 Turning to the interpretive authority of particular judgments of the Court, therefore, in light of the infrequency of disputes and questions submitted to the Court, such that the Court rarely obtains the opportunity to revisit the same legal question, every statement on the law by the Court in its decisions holds considerable authoritative potential. 62 AOs that highlight weaknesses or shortcomings in the Court's interpretation of the law in a judgmentwhether treaty-based or customarycan weaken the degree of consensus 56 W. Brennan, 'In Defense of Dissents', (1985Dissents', ( -1986 62 The Court has been able to revisit its case law in some areas (e.g., maritime and territorial delimitation), resulting in the opportunity for routinization in those areas. In turn, the authoritative potential of articulations of law in such areas of law is greater. that crystallizes around that interpretation within the wider international law-making and interpreting community.

Input legitimacy
From the perspective of input (or process) legitimacy, that primarily concerns process, a traditional critique of AOs is that they undermine what critics refer to as 'secrecy'. 63 Secrecy is an attribute of the judicial process designed to preserve and promote the actual and/or perceived independence and impartiality of individual judges. 64 The independence of the judiciary goes into the heart of the rule of law and is central to the notion of the administration of justice, distinguishing the politics of law and legalism from other forms of politics, as well as other forms of thirdparty dispute settlement. 65 The independence of judgeswhether speaking of judicial institutions or the individual judges within the institutionis a principal criterion of input legitimacy and, in turn, a key source of authority. Thus, even when not expressed explicitly in terms of authority, 'secrecy'-based critiques of AOs strike to the heart of institutional authority.
The arguments advanced in favour of prohibiting, particularly, national judges and ad hoc judges from issuing AOs during the negotiation of the PCIJ Statute and deliberations upon amendments to its Rules of Procedure illustrate how and why AOs can be perceived to undermine secrecy, judges' independence and their capacity to decide cases impartially and, in turn, authority. Initially, the fear was underpinned by the belief that judges would always vote against decisions unfavourable to their appointing state and use AOs to record the fact and nature of their disapproval. 66 In turn, this would create the impression of partiality, thereby undermining the independence-based legitimacy of the Court and its constituent members. Later the argument shifted. Rather than potentially revealing latent partiality, AOs were viewed as a mechanism through which states could exert pressure on national judges and thereby undermine their independence: the prohibition of AOs would 'shield the judge from the reproaches of national public opinion'. 67 More recently, in a broader defence of the secrecy of deliberations, Bruno Simma and Thore Neumann explained that the justification for such secrecy is to prevent governments from 'monitoring the discursive behaviour of individual judges, [and seeking] to influence them by indirectly "disciplining" their discursive "inputs" and by using information from the deliberations to thwart a judge's re-election'. 68 Duncan French, for example, has suggested that Judge Weeramantry's failure to secure re-election for a second term of office may have been influenced by the stance he maintained on a number of legal and political issues in his ICJ AOs. 69 63 In the context of the PCIJ/ICJ, see Advisory Committee of Jurists, Procès-Verbaux of the Proceedings of the Committee (June-July 1920), at 531, 570, 591-2. 64 It can be manifested in different degrees and in different waysfrom the complete anonymity of the identity of judges and the suppression of the publication of any information that may reveal their identity, to the suppression of the publication of information that may reveal the views of any individual judges, to on the other end of the spectrum, the suppression of any information that may reveal the content of the Court's internal deliberations. Moving beyond independence, an opinion that implicitly or explicitly suggests that the Court's decision was reached without full due consideration of all issues raised by the case as perceived by the authoring judge may call into question the deliberative legitimacy of the Court's decision. 70 It can be seen, therefore, that AOs may be perceived to impact upon input and output institutional legitimacy in several detrimental ways, all of which may lie behind claims that AOs undermine institutional authority. As explained in the following section, this is just one account of the relationship between AOs, legitimacy, and institutional authority. The validity of these claimsand indeed, any accounts of this relationshipdepends upon the institutional context within which they are made and the structure of institutional authority therein.

The structure of authority
In his seminal study The Faces of Justice and State Authority, 71 Mirjan Damaška offered a theoretical model that illustrates the relationship between procedural design and practice, official authority, and the purposes for which that authority is exercised. To understand the relationship between a particular process or practice (such as issuing AOs) and the authority of the institution, Damaška's theory suggests that it is necessary to situate that practice within the wider systemic contextconsidering the range of factors embodied within his matrix of ideal types and their implications for each other. In particular, he explained how attitudes towards AOs are the product of the structure and organization of authority in the context of which those attitudes are formed. 72 That structure is determined by three categories of attributes of the justice system under question, namely the attributes of the decision-makers, the distribution of authority within the institution among the decision-makers, and the legitimate criteria for decision-making. However, when those attitudes are taken out of the systemic context in which they were developed and transplanted into another, the validity of that transplant depends on the fit within that new context. 73 Arguments of the kind canvassed in Section 3, that the expression of individual judicial authority in the form of AOs is incompatible with institutional legitimacy, are most consistent with those systems of justice that bear the characteristics of Damaška's 'hierarchical ideal type' of authority. The hierarchical ideal type is typified by the following three attributes: (i) a professionalized body of official decision-makers; (ii) the decision-makers are organized hierarchically and among whom official authority is distributed widely and vertically, from the top down; 74 and (iii) the style of decision-making can be characterized as two variants upon legalisticpragmatic legalism and logical legalism. 75 The professionalized system of official decision-making refers to a system of permanent officials who, over time and experience in the same or similar role, carve out a sphere of practice. 76 In turn, this provides for the routinization and specialization of decision-making, namely the ability to address issues in a general, rather than individualized, manner by uniformly applying a narrow range of decision-making criteria. 77 The vertical distribution of decision-making within a pyramid of authority sees decision-making at lower levels subject to superior review. 78 (Judge Weeramantry, Dissenting Opinion) and Nuclear Weapons Advisory Opinion supra note 54 (Judge Weeramantry, Dissenting Opinion). 70 Questioning the fullness of 'collegiate discussion' in the case at hand see Barcelona Traction case, supra note 25, at 86, para. 37 (Judge Fitzmaurice, Separate Opinion). 71 Damaška, supra note 14. 72 Ibid., at 19, 24. Damaška, supra note 14, at 18-23. 75 Ibid., at 21-3. 76 Ibid., at 21. 77 Ibid. 78 Ibid., at 20. Combined with the routinized and specialized characteristics of a professionalized body of decisionmakers, the strict hierarchization of this model of decision-making affords little room for official discretion in the exercise of authority. 79 Accordingly, the exercise of decision-making power is subject to a comprehensive system of institutional and professionalized accountability.
Damaška explicitly addressed the implications of the professionalization of decision-making upon the approach towards the expression of individual views, in the form of AOs. Because the professional decision-maker is expected to decouple their personal views from their official function: [j]udgments become pronouncements of an impersonal entity (a curia) even where a single individual is entrusted with their rendition. And because the institution must be univocal so as not to be equivocal, the announcement of a judgment made by several officials nullifies prior internal dissent: those who disagree must now repress their feelings. 80 If made in the context of a legal system tending towards the hierarchical ideal type, the authoritybased criticisms of ICJ AOs seem logical and consistent with the attributes of the system wherein the expression of individual judicial authority is inconsistent with institutional judicial authority. 81 However, evaluating the ICJ in accordance with the aforementioned three categories of attributes, it can be seen that the ICJ bears few of the structural characteristics of the hierarchical ideal type of authority. Rather, the ICJ bears closer resemblance to what Damaška characterizes as the 'co-ordinate ideal type'characterized by temporary, non-professional decision-makers, appointed for non-permanent periods of time and without any specific specialist training to equip them for official decision-making in the post they hold. 82 Judicial decision-makers within the coordinate ideal type join the bench after a lengthy career and do so from a diversity of backgrounds. While they are not 'lay' in the general sense of the word, co-ordinate decision-makers are lay in that they have not received a rigorous and uniform education and training in 'being a judge'. Instead, the diversity that lay decision-makers bring to the judiciary due to the diversity of their previous careers, experiences and backgrounds is considered a quality that makes them qualified to be a judge (e.g., as practitioners, politicians, diplomats, or academics). 83 Consequently, the decision-making traits of specialization and routinization are unlikely to take hold. It follows that there is greater opportunity for flexibility and an individualized approach to justice. The co-ordinate ideal is typified by 'a wide distribution of authority among roughly equal lay officials; with no one clearly superior to others, there is essentially a single stratum of authority'. 84 Whereas predictability and objectivity are ensured in the hierarchical model by the application of 'textually fixed rules', authority in the co-ordinate model 'depends upon the clarity of consensus in the community or in the dominant group'. 85 The individual attributes and professional experience of ICJ judges, 86 the diversity of their background and identities and the acknowledgement of the benefits that this diversity can bring, 87 as well 79 Ibid. 80 Ibid., at 19. 81 Indeed, it might be argued that within hierarchically-structured systems, beyond the courtroom individual judges do not possess authority as the expression of individual judicial power within those systems is not considered legitimate. 82 Damaška, supra note 14, at 24. 83 And this established and reputed experience and expertise may furnish the individual judge and their opinions with greater individual authority. 84 Ibid. 85 Ibid., at 28. as the appointment of ad hoc judges 88 are all factors that push the ICJ towards the co-ordinate ideal type of justice. While judges may be appointed to the Court bearing in mind their pre-existing experience and expertise, once on the bench the opportunities for specialization and routinization of decision-making to take hold are slim. The diversity in the factual and legal issues raised by cases and questions submitted to the Court, owing to its general jurisdiction and the infrequency with which they are referred, forces the Court to address directly and explicitly the individual and unique characteristics of each case on an ad hoc basis. The Court's 'flat' structure consists of a single level of decision-making authority composed of and within a single body and no mechanism of appellate review, irrespective of whether the Court sits in plenaryas it does in the majority of its casesor in chamber formation. 89 In light of this single stratum of authority, the limited opportunity for formal career progression within the Court calls into doubt the existence of the hierarchical dynamics wherein 'team playing' and obtaining consensus is rewarded by promotion. 90 The ICJ's characterization under the third limb of the organization of authority ideal typesformal versus substantive justice as the legitimate basis for decision-makingis open to greater contention. On the one hand, Article 38(1) of the ICJ Statute suggests that the only legitimate bases upon which decisions can be reached are those under Article 38(1)(a)-(c). However, in practice, the variance in methodologies for the interpretation of treaties and the identification of customwhich in turn result in potentially great disparities in outcomemean that a formalist conception of decision-making involving an appearance that this involves the simple identification and application of the applicable standard is not sustainable. The diversity of approaches towards decision-making represented by the judges of the Courtsome maintaining a strict approach in favour of formal justice and some eschewing the boundaries of formalistic conceptions of justice in favour of more equitable approachesmeans that it is impossible to ascertain a particular culture of decision-making of the institution as a whole. The diversity of legal and philosophical traditions represented through the diverse membership of the Court as required by Article 9 of the Court's Statute suggests that it was never intended for the Court to adopt a single jurisprudential approach towards decision-making, nor even to create the appearance of one. The diversity requirements under Article 9combined with the authorization provided by Article 38(1)(c) for individual judges to canvass the world's array of legal traditions and cultures, synthesized by way of the inclusive process of internal deliberationstogether create a picture of institutional decision-making more consistent with Damaška's co-ordinate ideal type than the hierarchical one.
Again, Damaška addressed the implications of the co-ordinate characterization for AOs. Whereas judges in the hierarchical system are expected to decouple their personal identity from their professional, institutional role, within the co-ordinate ideal type the personal and professional identities are inextricably interlinked. 91 Thus, within co-ordinately structured justice systems such as the ICJ, the individualand the individuality ofjudges are not only consistent with institutional authority but are a legitimizing attribute of the institution itself. In the absence of the checks and balances upon the exercise of institutional power that exist within hierarchical systems, the exercise of institutional power is held more directly accountable to the sense of personal-professional responsibility and reputational concerns of the individual judges who are It is undoubted that informal hierarchies may emerge and while there may be limited opportunities for career progression within the institution, the development of a reputation for consensus-building and 'team-playing' may be rewarded by appointments to other prestigious institutions or regimes. 91 Damaška, supra note 14, at 27. personally associated with the judgment, 92 and the legitimacy-appraising actors. Such accountability is facilitated through the adoption of modalities of transparency. 93 These have the effects of highlighting the personal-professional responsibility of the individual judges for the institution's decision and of providing the information that external actors require to substantively appraise the legitimacy of the institution and its judgments. One such modality is through AOs. Focusing on their effect upon the appraisal of institutional and decisional legitimacy by external stakeholders, the following section will explore how the individual authority of judges expressed through AOs interacts with institutional and decisional authority within the ICJ as a co-ordinately structured system of authority.

Additional opinions within the co-ordinately structured ICJ
The previous section explained how, in the absence of the structural guarantors of legitimacy characteristic of hierarchical systems, co-ordinately structured systems of authority place greater emphasis upon the assessment of the substantive legitimacy of the exercise of institutional power by the Court by external stakeholders. Structural and procedural mechanisms are, thus, required to facilitate the appraisal of both the input and the output legitimacy of expressions of institutional powers by those external stakeholders. AOs are one such mechanism, and through the tension that is reached between the authority claimed by (and, as this section will argue, for) the institution and the individual judicial authority embodied by AO, the legitimacy of the expression of institutional power at any given time and for any given purpose can be established.
Consequently, within both co-ordinate and hierarchical contexts, AOs should be understood as having potentially the same effect, that is, as a constraint upon the institutional authority. The difference lies in how that effect is conceived. Whereas within hierarchical systems this is perceived to be a negative constraint, within co-ordinate systems it is considered one of constructive restraint, and the fact that they exist as a potential restraint itself being a legitimizing attribute of the Court's procedure. The relationship between AOs and institutional authority is constructive in other ways. As this section will demonstrate, they hold the potential to reinforce both input and output legitimacy. They fulfil these functions in at least two (often, but not always, related) ways: one formal and expressive, and the other substantive. In some instances, the effect of AOs upon the appraisal of institutional legitimacy lies simply in the fact of their existence and possibility as institutional practice, irrespective of the substantive way in which they are (not) used by individual judges. In other instances, in addition to what they represent, their significance also lies in what they say: their effect upon the appraisal of institutional legitimacy lies also (or predominantly) in the substantive views expressed therein. The remainder of this section will elaborate upon these ideas and will explain how, specifically, AOs act upon how institutional legitimacy is appraised.

Output legitimacy
There are two principal ways in which AOs can influence the appraisal of the output legitimacy of the Court's judgment, which in turn will affect the Court's decisional and interpretive authority. In the first instance, AOs are conceived as a supportive force: one that strengthens the claim to output legitimacy. In the second instance, the force exerted upon institutional and decisional legitimacy can work either way: they can be supportive, but they can also be forces of institutionaland necessaryrestraint.
First, not only are AOs an expression of the Court's representative legitimacy (supporting the claim to general decisional authority) but as Andreas Paulus has suggested, they can also enhance the Court's output legitimacy by appealing to, and expressing, the value-based legitimacy criteria held by parties to the dispute and more widely, stakeholders within the international law-interpreting community. 94 In doing so, they support the Court's claim to both particular and general decisional authority. 95 An example of this might include Judge Ammoun's separate opinion in Western Sahara, which concurred with the Court's conclusion that the territory of Western Sahara was not terra nullius at the time of Spanish colonization. 96 However, Judge Ammoun reached that conclusion not on the basis of an analysis of the practice of those states recognized as 'civilised nations' at the time of colonization (as had the Court) but rather on the basis of African philosophical and legal thought, drawing upon a spiritual understanding of the relationship between humanity and the land, recalling notions of ancestral ties, rather than territorial control. 97 More recently, Judge Cançado Trindade, while voting in favour of the Court's dispositive paragraphs, frequently pens lengthy separate opinions that advance his human-centric conception of international law as the basis and rationale for the conclusion with which he concurred. 98 Furthermore, judges may use AOs to supplement the Court's reasoning and in so doing enhance its clarity and aid full understanding of how and why the Court reached its judgment. This lends support to both the general and particular decisional authority of the Court. ICJ judges have described the Court's judgment as a composite of the views of each of the judges aligned in the majority, rather than the articulation of a singular comprehensive judgment of 'the Court'. 99 In turn, legal scholars and ICJ judges alike have observed how AOs restore the 'conceptual richness and colour' of the Court's pronouncements, 100 which in turn aide appreciation of the judgment, how and why it came about, and its implications. 101 This has led some to refer to the 'indissoluble' 102 relationship between AOs and the Court's judgments, one where they 'belong to each other, and ideally, illuminate each other', 103 perhaps by serving as a foil to the Court's reasoning. 104 This idea that AOs can act as a foil to the Court's reasoning takes us to the second way in which AOs influence the appraisal of the Court's institutional and decisional output legitimacy. In the absence of any appellate process, AOs allow external stakeholders to evaluate the substantive See Hernández (supra note 8, at 124) who concludes that 'the Court's practice of publishing individual opinions is part of its claim for the wider authority of its judgments, not vis-à-vis the parties before it but with respect to the wider audience'. 96 Western Sahara, Advisory Opinion of 16 October 197516 October , [1975 ICJ Rep. 12 (Vice-President Ammoun, Separate Opinion). 97 Ibid., at 85-7. Fitzmaurice, supra note 29, at 1-2; expanded upon by Judge Jennings, supra note 33, at 352. 104 Bedjaoui, supra note 99, at 58. merits of the Court's judgment in light of alternative possibilities and in turn affirm the superiority (or, perhaps, inferiority) of the judgment over possible alternatives. 105 Thus, in addition to exposing the Court's judgment to more rigorous scrutiny, AOs expose the authoring judge's reasoning to scrutiny for the purposes of ascertaining whether their disagreement or divergence from the position adopted by the Court was well-founded. To the extent that AOs do directly or indirectly alert external stakeholders to deficiencies in the Court's judgments, AOs should restrain the output legitimacy of the Court's judgments andin turnits claim to authority regarding those deficient matters. This constructive restraint is particularly important in the context of the Court's interpretive authority.
In the context of international law, the degree to which any given proposition of law asserted by the ICJ is accepted as an accurate and authoritative articulation of the law is determined by the wider law-interpreting community on an articulation-by-articulation basis. 106 As such, AOsrather than being viewed as harmful to the Court's authorityshould properly be understood as being a mechanism that assists the law-interpreting community when making that determination. Support for this conception of the institutional function of AOs can be found in the views expressed by participants in both the drafting of the PCIJ Statute 107 and the negotiations on the amendments to the Rules of Court, 108 and more recently in the extra-judicial writings of former ICJ judges. 109 It is implicitevenin the pre-judicial observation of a current ICJ judge that 'a decision [of the Court], especially if unanimous or near unanimous, may play a catalytic role in the development of the law'. 110 Beyond the Court, this view has also gained traction within the wider law-interpreting community. 111 Indeed, one only needs to consider the context in which Kammerhofer's critique of AOs was made: in the conclusion to a critical case comment on the Court's judgment in Oil Platforms. As he observed, '[a]nyone who reads the separate and dissenting opinions is made very much aware of the shortcomings of the present judgement [sic] '. 112 If one were to accept that the shortcomings as perceived by the authors of the AOs are indeed shortcomings of the Court's judgment, then the degree of authority enjoyed by the judgment should reflect those shortcomings.
While, as noted above, there is certainly a legitimizing value to consensus, 113 whether that be evidenced by unanimity or near unanimity in the Court's vote or in the lesser disagreement expressed by judges with the decisionthat value is only rendered possible by the opportunity for the expression of disagreement. Thus, desire for and pursuit of, unanimity should not be confused with the creation of a fiction of unanimity through the adherence to a strong conception of secrecy. To shroud majority-determined decisions with the cloak of unanimity with a view to 105 Fitzmaurice, supra note 29, at 1-2; expanded upon by Judge Jennings, supra note 33, at 352, and Bedjaoui, supra note 99, at 58. The authority-limiting potential of AOs is valuable not only from the perspective of constraining the interpretive authority claimed by the Court but is also important from the perspective of the authority claimed for the judgment by other political actors. As Karen Alter has observed, through their jurisprudence international courts and tribunals influence political outcomes by: empower[ing] those actors who have international law on their side, increasing their out of court leverage. [International courts and tribunals] then alter political outcomes by giving symbolic, legal, and political resources to compliance constituencies, ever-changing groups of actors that for a variety of reasons may prefer policies that cohere with international law. 114 On this basis, Alter recognizes international judges as being not only legal actors but also political actors. Acknowledging how judicial pronouncements can also empower other political actors (for example, by lending support to particular historical narratives or by delegitimizing political opponents), AOs serve as a potential check on the use or abuse of judicial authority by other political actors.
One can look to the Nuclear Weapons advisory proceedings by way of illustration. 115 Devised and promoted by a coalition of (non-nuclear) states and NGOs 'whose entire purpose was to achieve a political objective', 116 the request for an ICJ advisory opinion was motivated by the hope that the resulting opinion would strengthen 'both governmental and civic anti-nuclear pressures'. 117 While the 14 AOs attached to the Court's Nuclear Weapons advisory opinion diluted the authority of the Court's opinion, 118 that dilution was necessary for the preservation of the Court's authority. Martti Koskeniemmi has described the international law-making as a process of consensus-building, that is, 'a hegemonic process in which some agent or institution has succeeded in making its position seem the universal or "neutral position"'. 119 On this view, there is no objective centre or neutral point independent of political arguments, but rather the centre ground is actually an ever-contested 'terrain of irreducible adversity'. 120 All law, he claimed, 'is about lifting idiosyncratic ("subjective") interests and preferences from the realm of the special to that of the general ("objective") in which they lose their particular, political colouring and come to seem natural, necessary or even pragmatic'. 121 This understanding of consensus forming and lawmaking may be applied to the analysis of decision-making and judgment forming, and AOs.
As the Court's dispositive paragraphs and the AOs in Nuclear Weapons demonstrated, the 'process of contestation' or deliberation within the Court failed to reach a consensus, with the principal question determined by the President's casting vote. In the absence of an emergence of a middle ground whereupon the law could be identified, for the Court to have bestowed upon one political position additional leverage by way of having the opinion of the Court 'on its side', would have resulted in a politicization of the Court in a manner inconsistent with the nature of judicial authority. Instead, the number and strength of the AOs counteracted the assumed authority of the Court's opinion, indicating that the process of contestation was still open and the outcome undecided. Rather than privileging one side in that contest the Court, through its advisory opinion and the accompanying AOs, contributed to that ongoing process of contestation, by equipping those actors accepted as legitimate lawmakers on either side of the contest with juridical arguments supporting their positions. 122 In this case, AOs served less as a means by which the authority of the Court is restricted, but more as a restraint upon how the authority of the Court can be used (and/or abused) by other political actors.
Finally, there is one important qualification to the argument advanced in this section, concerning the constructive effect of the authority-limiting potential of AOs. It is possible that how individual judges use AOs may have a destructive effect upon judicial authorityboth that of the individual judge and that of the institution. As I have argued elsewhere, 123 although AOs are not inherently inconsistent with judicial authority (whether individual or institutional), the use of particular language in AOs, or the use of AOs to launch personal attacks against colleagues, can be harmful to judicial authority by undermining collegiality and by undermining the charismatic authority of judges. 124 For example, the image of judicial quarrelling generated by a number of AOs to the Nicaragua judgment that addressed allegations of partiality made by one judge against another, 125 was neither in the interests of the institution nor the individual judges concerned. Judges should be mindful of the grave consequences for their own reputation as well as that of their colleagues and the institution when considering what they use their AOs to address and the way they express their concerns.

Input legitimacy
Whereas critics of AOs in the context of international law point to the fragility of international judicial authority owing to the weak institutional power, 126 the response of co-ordinately structured systems of authority is to emphasize those legitimizing attributes that it does possess to maximize their legitimizing effect. For the ICJ, its representative diversity is often posited as one of its greatest legitimizing attributes. 127 Beyond the courtroom, AOs are the principal public manifestation of the ICJ judge: not only is the mere fact of AOs an expression of that diversity, but they offer an opportunity for judges to give substantive manifestation to the diversity of legal and philosophical traditions in the Court's composition and to demonstrate the contribution of that diversity to the process and outcome of adjudication by the ICJ. 128 122 Beyond diversity, AOs can be an expression of the legitimizing attributes of independence and impartiality. Whereas the logic of secrecy is preferred within hierarchically-oriented systems to defend judges against threats to their independence or impartiality, within co-ordinate systems AOs are an expression of individual judicial independence and more substantively, they offer judges the opportunity to demonstrate their own independence. In the absence of hierarchical mechanisms of oversight and accountability, and recalling the scepticism around the willingness and capacity of international judges to act independently and impartially, 129 the policy of secrecy as implemented through the suppression of AOs would do little more than reduce the likelihood that any improper influence is uncovered.
Not only do AOs offer judges the opportunity to express legitimizing attributes, both theirs individually andtaken in sumthat of the institution, they invite external stakeholders to scrutinize the Court's judgment (and the opinions of individual judges) in light of their content to evaluate the degree of independence and impartiality exhibited. Within co-ordinate systems that emphasize the role of external stakeholders in the substantive appraisal of legitimacy, this facilitative effect of AOs is particularly important. Althoughand cruciallyjudges are free to determine whether to issue an AO and, if so, on what matters to write, 130 AOs hold the potential to uncover any undue pressure upon authoring judges by opening up the personal decisionmaking process of the authoring judge to the same kind of scrutiny to which the Court's judgment is subjected and in extreme circumstances, by existing as a platform for judicial whistleblowing. 131 And, because they hold the potential to reveal latent partiality or improper influence upon the Court or upon individual judges, AOs have a prophylactic effect, discouraging the adoption of potentially delegitimizing practices upon threat of their revelation in an AO should they occur. 132 Finally, both within co-ordinately structured systems of justice in general and within the ICJ specifically, there has been an understanding that the production of draft AOs (for consumption in the external deliberative sphere) enhances the quality of deliberation in the internal deliberative sphere, thereby enhancing the judgment's claim to deliberative legitimacy. The production of written opinionsnotesby individual judges as the basis for internal judicial deliberations is an established part of the ICJ's procedure. 133 Requiring individual judges to formulate their own tentative positions and to commit them to writing for circulation among colleagues, not only assists the deliberations of the Court by presenting a range of possible juridical solutions to the dispute at hand that can then be debated and compared, but also enhances personal deliberations of the individual judge. As deliberations progress and a majority position crystallizes, the transformation of notes into draft minority opinions provides the emerging majority with a counterpoint against which to test the coherence and persuasive rigour of their argument. 134 Being presented with a fully articulated alternative to the decision before it is rendered and made subject to public scrutiny provides an opportunity for the majority position to be clarified, modified, or strengthened in the face of weaknesses or matters arising in draft minority-judge authored opinions. The potential power of AOs in the internal deliberative sphere lies not only in their written form but alsoand perhaps more significantlyin the prospect of their publication and, in turn, their capacity to influence how the Court's judgment is received and perceived in the external 129 See Section 3.2, supra.

130
Which itself limits their potential as mechanisms by which to monitor judges' 'discursive inputs'. See above, text accompanying note 68. deliberative sphere. 135 This dynamic is neatly encapsulated in an anecdote recounted by Hugh Thirlway, from his time working in the Court's Registry. After Thirlway had alerted the Court's drafting committee to an argument advanced by Judge Oda in his draft dissenting opinion, the committee duly strengthened the Court's reasoning to address the point in question. Thirlway recalls how Judge Oda, stated 'in mock bitterness, "Mr. Thirlway, you keep moving the targets that I am firing at!"', highlighting the constructive role that the prospect of the publication of credible critique can play in incentivizing the strengthening of the Court's judgment. 136

Conclusion
Writing on dissent in a different context, Roland Bleiker has observed that dissent is a: field of enquiry that has the potential to reveal far more about power and agency than one might think initially. The process of undermining authority says as much, for instance, about the values and function of the existing social and political order as it does about the urge to break out of it. 137 The same can be said for judicial dissent within the context of international adjudication. Our effort to understand the relationship between AOsas expressions of dissentand institutional authority has led to a broader enquiry into the very nature of institutional authority at the ICJ.
AOs are not beyond reproach, both as a matter of principle and in terms of substantive practice. Nevertheless, I have proposed that those criticisms based upon the inherent nature of judicial authority are misconceived. When doing so, I have sought to articulate a more appropriate understanding of the relationship between AOs and the judgments to which they are appended. I argue not only in favour of the 'mere' consistency of AOs with the structure of the ICJ's authority, but also in favour of their significance for the structural integrity of that authority. Considering this, it is important that judges and external stakeholders appreciate their roles and responsibilities of judges and other actors when authoring and engaging with AOs.
Firstly, when exercising their discretion when deciding whether to issue an AO and on what matters, ICJ judges must do so mindful of the institutional context in which that discretion is being exercised. AOs do have consequences for institutional authority but those consequences are contextually contingent. The Court's composition is designed to reflect the principal legal systems and civilizations of the world and it is thus expected that judges will bring to the Court their views on AOs that have been informed by their prior experience and training. However, as I have argued in this article, it would be incorrect to assume that those views may be transplanted to the ICJ context. Rather, judges should be guided in their use (or non-use) of AOs by an accurate understanding of their relationship to ICJ authority. On the use of AOs as vehicles for accounts of the entire legal-philosophical outlook that informs the personal decision-making of a judge in the case, 138 the discretion that judges retain in how they use AOs must be exercised with a similar mindfulness of the institutional function of AOs. Where the production of such opinions delays the Court's proceedings owing to the time taken to produce such opinions and/or preventing them from contributing to the Court's internal deliberations, this is neither in the interests of the Court nor the individual judge in question. 139 Once a judge has established his or her legal-philosophical orientation on the judicial record, it is certainly questionable whether it is necessary for that judge to reproduce it in full in every opinion. Each AO does not exist in a vacuum: they will be read with knowledge of the author's jurisprudential leanings andin some casesit will be their prior record that will have influenced their election to the Court. 140 While it is not within the Court's custom to cite or refer to AOs explicitly, 141 it may be both in the interests of judicial economy and judicial transparency for greater cross referencing by individual judges to prior AOs (authored by themselves or by others) when they have influenced or informed the opinion at hand. 142 Secondly, ICJ judges enjoy a broad discretion over whether to issue an AO and on what matters. How judges exercise that discretion will affect the substantive contribution of AOs to the legitimacy-appraising enterprise of external stakeholders. Given the co-ordinate structure of authority at the ICJ, the fact that judges are afforded an opportunity to issue AOs and that they exist as a mechanism of transparency is alone a legitimizing attribute. Nevertheless, the full realization of the legitimizing attributes of AOs as a mechanism of transparency depends upon how they are used by judges and the extent to which they can be said to offer an accurate, if only ever partial, window into how the decision was reached. It is for this reason that this article has emphasized the potential effects of AOs.
Thirdly, this article has highlighted the responsibility of external stakeholders to hold the exercise of institutional authority to account. Thus, more broadly, I join the calls for greater scholarly engagement with the work of the Courtincluding the work of its individual judges. 143 Audiences may well be able to fully appraise the input and output legitimacy of the ICJ without recourse to AOs. However, they are a resource that can aide that appraisal and when engaging with AOs, external stakeholders must do so on a contextually accurate understanding of their relationship to institutional authority.