Skip to content
Accessible Published by De Gruyter Mouton June 21, 2016

Balancing neutrality and partiality in arbitration: discursive tensions in separate opinions

Ruth Breeze
From the journal Text & Talk


Although arbitration is increasingly being used to settle important disputes, particularly on an international level, little attention has focused on the language used by arbitrators. This article contains a qualitative analysis of the discursive moves and resources used in separate (dissenting and concurring) opinions published on the website of the International Center for the Settlement of Investment Disputes from 1987 to 2013. Arbitrators’ discursive practices in this forum are analyzed, with a particular focus on the tensions that arise between the need to sustain the arbitral system and maintain professional relations, on the one hand, and the equally pressing need to display commitment to the losing party’s cause, on the other. These tensions have parallels in other areas of legal practice and professional life, and can be understood as part of the way power systems operate and replicate through discourse.

1 Introduction

Specialists in legal language have recently devoted increasing attention to the area of alternative dispute resolution (ADR), as various forms of arbitration, mediation and conciliation have become more widespread as a means for resolving issues without going through national court systems. One of the most important ADR institutions is the International Center for the Settlement of Investment Disputes (ICSID), established under the auspices of the World Bank, which handles increasing numbers of high-level disputes between companies and sovereign states. Under ICSID rules, cases are usually handled by a panel of three arbitrators, one chosen by each party, and the president, appointed by mutual consensus (ICSID 2006). Despite the importance of this type of ADR, it has received little critical attention from linguists, perhaps because samples of language are rarely available: cases are heard in camera, and only the final award is published. However, one promising source of evidence is available in the form of the separate opinions that sometimes accompany the award. These opinions are of two kinds: arbitrators may sign the award, but publish a concurring opinion explaining their discrepancies on some points; or arbitrators may decline to sign the award and publish an individual dissenting opinion. The practice of issuing separate opinions, thought to have originated in Common Law systems, is currently spreading across different institutions (Lévy 1989; Sanders 1999), and the last 30 years have seen a significant increase in the number of separate opinions published by ICSID: only 14 awards were published with separate opinions between 1987 and 2007, while the ICSID online archive currently reports 27 separate opinions in cases handled from 2008 to 2014.

Bibliography on arbitral practices provides only very limited discussion of the nature and purpose of the separate opinion at ICSID (De Boissésson 1998; Dolzer and Schreuer 2008). Article 48 (4) of the current Convention states: “Any member of the Tribunal may attach his individual opinion to the Award, whether he dissents from the majority or not, or a statement of his dissent” (ICSID 2006). However, since the aim of arbitration is to reach a best-fit solution that is acceptable to the parties, dissent is no trivial matter (Breeze 2012a; Breeze 2012b). The most obvious motivation is to put the record straight when arbitrators feel that the award has misrepresented some aspect, or has applied faulty argumentation. However, another reason is the need to demonstrate to the losing party (usually the party that nominated this arbitrator) that he/she argued cogently for its position during the proceedings. Reasons advanced in the opinions themselves include case-internal issues such as flawed legal reasoning, inadequate explanations or evidence, misunderstandings of fact, and case-external factors such as the purpose of the treaty or the nature of the precedent being set (Breeze 2012a).

Importantly, the arbitrator who issues such a public statement of disagreement runs certain risks, since he/she is drawing attention to flaws in the system, or to open discord with his/her peers, and implicitly criticizing their reasoning or handling of the case. Dissenting arbitrators are therefore faced with a complex discursive task: their commitment to the client must be balanced carefully against the need to uphold the institution of arbitration as such. It is clear that the speaker/writer in a given professional context operates within a framework determined by institutional factors and relationships, and can only make choices within the range of what is available in that context. Dissenting opinions may be highly critical, but they still need to be seen in a dialogic perspective, as “situated and framed by an exchange between speaker and audience” (Amossy 2009a: 265); and the scope within this dialogue is constrained by the values and conventions of the professional situation.

In broad terms, what is happening in such opinions may be usefully compared with the “accounts” identified by discourse analysts and social psychologists, through which speakers/writers defend, justify or play down their actions or words in order to maintain the social order (Billig 1985; Antaki 1988; Arribas-Ayllón et al. 2011). In Sarangi’s words, “accounts reflect culturally embedded normative explanations” (Sarangi 2010: 403), which can be understood in the present context rather narrowly as meaning justifications embedded in the professional practices and ideological underpinnings of the discursive community of arbitral practitioners and their clients. Such “accounts” are essentially dialogic, other-oriented, and shaped in such a way as to negotiate the uneven terrain of what is socially acceptable, often in difficult circumstances. Although they have been extensively researched in other areas, such as everyday conversation, healthcare and counseling (Antaki 1988; Arribas-Ayllón et al. 2011), their importance in legal genres has so far received little attention from discourse analysts, and it is hoped that this study will provide further insights into this discursive phenomenon.

In methodological terms, my study will address the arbitrators’ separate opinions using an integrated, qualitative approach to analysis which situates professional discursive practices within a dialogic perspective, centering on the pragma-linguistic resources involved in negotiating the dissenting arbitrator’s position. Essential to my present analysis is the concept of “move,” a macrostructural feature associated with a particular genre, which is important in that it provides a conceptual bridge between an understanding of textual rhetoric as dialogic argumentation (Perelman and Olbrects-Tyteca 1969), and a recognition of the constraints that shape professional genres and discourses (Bhatia 2004). My exploration of the separate opinions will start by mapping the range of moves encountered in arbitrators’ texts. I will then center on a qualitative exploration of how the choice and phrasing of these moves are shaped by the writer’s strategic purpose within the ongoing negotiation of meaning in this professional situation. Particular attention will focus on the textual realization of the moves, on the grounds that patterns of reasoning must be examined in their exact phrasing, since the wording is “not an exterior garment, but the very body of argumentation” (Amossy 2009b: 317). I shall conclude with an interpretation of some of the aspects of arbitral discourse that have come to light in terms of professional discourse practices.

The texts used for this study constitute all the 34 separate opinions issued from 1987 to 2013 which were publicly available on the ICSID website at the time of this study (7 opinions were unavailable). The sample of texts used is therefore small, but comprehensive, in that it includes all the available texts from the given period, which can be taken as solid evidence of arbitrators’ discursive practices in this forum. The 19 opinions cited in the text are listed in the appendix.

The following sections of the article are organized in terms of what might be termed discursive macro-function, beginning with the resources which arbitrators use to mitigate criticism and ostensibly maintain discursive neutrality (Section 2), followed by those used to express vehemence and strong commitment (Section 3). The most characteristic styles of opinion are then briefly summarized (Section 4), followed by some conclusions.

2 Resources for maintaining discursive neutrality

Arbitrators who dissent are evidently motivated by grave reasons and have a strong desire to let their views be known. Nonetheless, they often appear to preserve a professional air of discursive neutrality in the text. The projection of impartiality may strengthen the arbitrators’ arguments, since they are suggesting that something is impartially, objectively true, rather than emphasizing their own subjective conviction. This texture of neutrality is achieved in a variety of ways, which I group here as: emphasizing agreement and downplaying disagreement, externalizing authority and depersonalizing the attack. Since the examples used to illustrate discursive moves always include elements of their linguistic realization, I shall first explain the strategic moves, and then point out the more interesting features of their presentation.

2.1 Emphasizing agreement and downplaying disagreement

2.1.1 Emphasizing agreement

It is striking that even in dissenting opinions most arbitrators make a point of emphasizing their agreement with the majority view.



The Committee also has my complete support when it says [...]

By stressing points of agreement, they not only put the record straight by clearly delimiting the extent of their disagreement, but they also position themselves as being open to reason and willing to accept others’ views. It may be noted that in the following examples, the writer foregrounds agreement, placing his disagreement in second position. In example (2), the arbitrator also prefaces his agreement with an expression of respect for other members of the panel, while in example (3), the arbitrator clearly spells out which parts of the award he supports.



I emphasize both my high regard for my colleagues and the narrow scope of my concurring and dissenting opinion. I agree in most respects with the Tribunal’s factual and legal analysis, which is careful and thorough. I differ only on limited grounds, and do so with reluctance, mindful of the desirability of unanimity in arbitral decision-making.



I agree with the Award on the description of the facts and allegations, on procedural and jurisdictional issues […]. But I do not agree with the conclusions of the Award concerning national treatment and discrimination.

Even where the dissenting arbitrator does not agree, he/she often pays lip service to the validity of the arguments expressed, or even praises some aspect of the award.



I consider it important to append this opinion of my dissenting views, not to denigrate or undermine the reasoning and logic of the Award, but only to point out the key differences between my views and those of the majority.



my dissent does not relate to any particular aspect of this brilliantly drafted Decision.

2.1.2 Regretting disagreement

In some cases, disagreement is expressed clearly, but is fronted with an expression of regret, which again often seems to be accompanied by expressions of respect for the other professionals on the panel.



Regardless of my recognition of and respect for my renowned colleagues [...] I could not find a common position on how my colleagues have construed the facts and the law to dismiss Claimant’s complaint.

Indeed, the individual arbitrator may even construe him/herself as regretting his/her disagreement, whether through a conventional formula (example [7]), or more elaborately (example [8]).



Lamento no coincidir con el razonamiento del voto mayoritario.

‘I regret that I do not agree with the reasoning of the majority vote.’



I nevertheless have the misfortune to be of a different opinion on the outcome.

The use of concession clauses has a particular role here, in that it enables the arbitrator to acknowledge the validity of the panel’s viewpoint before presenting his/her disagreement, as in (9).



No doubt the Committee is right to say [...] but that does not, to my mind [...]

2.1.3 Sharing responsibility for disagreement

Although arbitrators disagree, they often choose to emphasize that they are not alone in this. They thus mitigate their lonely position as dissenters, and work up rhetorical backing for their point of view. In some cases, they put the criticism of the majority view into the mouths of others, such as witnesses at the hearing:



This misappropriation of City Water’s property for use by DAWASCO cannot be characterized as anything other than both expropriatory and a denial of fair and equitable treatment. Indeed, it was recognized as such, in very frank testimony, by one of the Republic’s responsible officers in the course of these proceedings.

In others, they even maintain that the parties themselves have given reason to suppose that they agree with the (dissenting) opinion:



To that extent, therefore, Mexico itself has expressed a position that does not support, and that neatly contradicts, the opinion of the majority constituting the decision contained in the present Award.

Similarly, they may seek to lessen their own responsibility for disagreeing by showing that their view actually builds on that of the majority:



The majority admits that the extent of the above mentioned evidence of discrimination is limited. I would suggest that it is not only limited but null and void.

2.1.4 Mitigating disagreement

When arbitrators do express their disagreement, they often use specific pragma-linguistic means to soften their statements. Hedging is classically used to create a degree of distance between the writer/speaker and what is stated (Hyland 1996). In these texts, the arbitrator uses hedging to reinforce his/her discursive neutrality. This corpus abounds in moves containing hedged expressions of disagreement, some of which are quite elaborate and couched in metaphor:



For these reasons, I would set the annulment bar rather lower than my colleagues, and find that this case crosses it.

In other cases, standard hedging devices are used to mitigate the writer’s criticism of the award, such as epistemic modals of possibility (“may”, “might”), epistemic lexical verbs that lower the writer’s degree of commitment (“seem”, “appear”), or adverbs such as “allegedly” (Hyland 1996):



The Award appears to rely on [...]

The following example is noteworthy in its use of the (classically legal) double negative (Trosborg 1997), often decried by advocates of plain English (Cutts 1995), which creates a strong distancing effect and offsets the implicit criticism:



None of this is of course to say that the Tribunal’s reading of what Article 2 as a whole properly means is not a tenable one.

The use of counterfactual conditionals, also very frequent in legal discourse (Nivelle 2008), might be interpreted as having a similar effect, but is worthy of greater attention. In some cases, such as example (16), it is used persuasively to point out logical flaws, and thereby reflects Rescher’s (1961) category of counterfactuals with a clear interpretation. Following Nivelle (2008), examples of this kind can be seen as “echoic dissociating CTFs” (Nivelle 2008: 481) in that they pick up some argument used by adversaries in order to show its inconsistencies. However, this strategy arguably also lessens the direct impact of the criticism by moving the disagreement onto a theoretical plane.



However, if the existence of the Mexican litigations in this case was not incompatible with the terms of the Article 1121 waiver, the conduct of Claimant in maintaining or appealing those litigations could not have been incompatible with it.

In example (17) below, the arbitrator again uses the counterfactual as a form of argumentative persuasion, suggesting that something might or could have been, if circumstances had been different. He thereby shifts the focus from the specific case to hypothetical others, and diverts emphasis away from the concrete facts to possible other scenarios. This process of abstraction depersonalizes the argument, presenting what is essentially an accusation in a roundabout way. Moreover, it also projects some degree of impartiality, implying that “if circumstances had been different,” the same situation, with the same key players, could have been resolved differently.



As such, it should have been supported by extremely clear and convincing evidence, which was not the case in Marvin Feldman v. México.

Use of elaborate hypotheticals seems also to be a form of courtesy mitigating what is, at bottom, a strong disagreement that the arbitrator has a “duty” to vent (Asher and McCready 2007):



I would fail in my duty if I were to conceal my doubts out of friendship for my colleagues.

2.1.5 Wider lack of consensus

Another argumentative strategy that can be drawn on to mitigate dissent is that of situating the current disagreement within a broader landscape of uncertainty: that of the lack of general consensus in the arbitral community about the issues at stake, and the complexity of the problem at issue:



Whether and to what extent the ILC Articles, and the Commentaries thereto, constitute accurate restatements of customary international law, and to what extent they represent progressive development of international law, is hardly a matter on which there is general agreement.



The present case is a very complex case to which the three members of the Tribunal have been extremely devoted during many years.

Arbitrators also stress the contingent nature of judgments that have to be made, admitting room for discussion, and thereby minimizing the discursive impact of the differences that have arisen. This approach may be interpreted as a face-saving device (Brown and Levinson 1987), in that it seems to maximize respect for the others’ views:



These are always matters of judgement, sometimes quite delicate judgement, and, if I find myself coming down on the other side of the line from them, I doubt whether the distance between us is all that great.

2.2 Externalizing authority

2.2.1 Delegating responsibility to an external authority

Legal discourse tends to cite authorities, because all legal systems have a top-down aspect which involves applying laws or previous decisions to particular cases. Although in theory arbitrators are not bound by precedent, since arbitration is a form of ADR outside the common law system, an increasing tendency for arbitrators to cite the growing body of arbitral precedent from the same or similar fora has recently been noted (Commission 2007; Weidemeier and Mark 2010; Breeze 2014). It is particularly interesting to observe how this may function discursively in the case of separate opinions. Here, recourse to authority in the form of laws, practices, principles or precedents outside the immediate case could be interpreted as a means of directing attention to the theoretical issues, and thus away from the rights and wrongs of the particular case and the roles played by the individuals involved. It could thus be interpreted as a strategy used to preserve an air of neutrality and save face, rather than emphasizing personal commitment. Previous research (Breeze 2014) has shown that authorities invoked in separate opinions at ISCID include national laws; general principles or adages; academic papers or reference works; previous ICSID awards; and previous non-ICSID awards. Arbitrators add weight to their arguments by explaining what the relevant authorities have said, but they also transfer responsibility for what is being said to the authorities that they cite:



The International Law Commission’s Articles on Responsibility of States for Internationally Wrongful Acts (ILC Articles) make clear that a state which commits an internationally wrongful act, such as an expropriation, is under a number of obligations. These include [...]

However, this is not as simple as it might appear, because in ICSID arbitration, apart from ICSID rules and the Bilateral Investment Treaty involved, there are few sources of authority whose use is completely uncontroversial. For this reason, when arbitrators refer to other sources, they often emphasize the importance of those sources discursively, boosting them by using adjectives that index the locus of quality (example [23]), or that of quantity (example [24]). We may note, though, that even when they are underlining the importance of their sources, the arbitrators are delegating responsibility for the propositions expressed to those entities rather than expressing personal commitment.



Other highly significant international agreements […]



Numerous publicists and decisions of international tribunals overwhelmingly support the position that, as a general rule, a host State is not liable […]

Similarly, criticism is also sometimes directed at an absence of sources when the arbitrator perceives that the majority have failed to ground their arguments properly on some authority:



However, Respondent has cited no authority, and I have seen none, supporting the proposition that [...]

2.2.2 Invoking the purpose of ICSID or specific treaties

As well as directing attention towards relevant authoritative texts, dissenting arbitrators also often point to the purpose or “spirit” of ICSID (or the relevant BIT) as being the driving force behind their own dissent. Recourse to the purpose of ICSID, again, tends to redirect the particular arguments at hand, recontextualizing them in a more theoretical way. Very often, the recourse to the purpose or “spirit” of ICSID is also linked to a concern with setting a precedent for future cases.



This could not have been the reasonable intent of the NAFTA Parties […]

2.2.3 Setting a precedent

Although arbitration lacks a strong theory of precedent such as might be found in most common law jurisdictions, arbitrators are very much aware of past and future precedents when they deliberate on cases (Commission 2007; Weidemeier and Mark 2010; Breeze 2014). Arbitrators who publish separate opinions are often eager to point out that their actions are motivated by their duty to influence future proceedings.



I should explain why I do so, in the interests of the ICSID system as a whole, and as a pointer for future Tribunals.



The arbitral system should not encourage that kind of behavior.

It is even frequent for arbitrators to frame their dissent as being necessary to protect the future of the institution – a move which engages the locus of precariousness beloved of institutional rhetoric (Perelman and Olbrechts-Tyteca 1969: 91):



If I have decided to dissent, it is because the approach taken by the Tribunal on the issue of principle raised in this case for the first time in ICSID’s history is in my view at odds with the object and purpose of the ICSID Convention and might jeopardize the future of the institution.



It is regrettable, so it seems to me, to put the extraordinary success met by ICSID at risk by extending its scope and application beyond the limits so carefully assigned to it by the Convention.

2.3 Depersonalizing the attack

In addition to using the argumentative moves set out above, the arbitrators also manage a range of linguistic resources which tend to render their discourse less personal, more objective, and thus more likely to reinforce the sense of neutrality and impartiality. It has often been pointed out that impersonal forms and concealed agency are used to legitimize controversial actions (Van Leeuwen and Wodak 1999). These resources mainly occur when the arbitrator is making a criticism of his/her colleagues. They might therefore be regarded as part of a general strategy to depersonalize attacks.

2.3.1 Avoidance of agency

Objectifying language is used in all three languages (English, French, Spanish) of the original opinions when expressing disagreement. This technique avoids establishing an antagonistic “I–you” relationship between the writer of the opinion and the writers of the award. In fact, the agency of both “sides” tends to be obscured. In the most basic sense, the widely used convention of referring to “the Tribunal” as an abstract singular entity is already an example of depersonalized language:



The Tribunal in this case has failed to acknowledge this important Distinction.

The frequent use of passives or impersonal “it” constructions also tends to deflect attention from the agents involved:


(de Witt)

Aucune analyse n’est formulée à cet égard dans la Sentence.

‘No analysis is expressed in this regard in the Award.’



It is essential to note that [...]

Another way of depersonalizing the text is by attributing mistakes through grammatical metaphor to “analysis” or some impersonal entity, rather than to actual people:



In my view, this analysis confuses issues of causation, on the one hand, and quantification or quantum of damages, on the other.

In the following case, a similar effect is achieved by appealing to abstract “logic” as a discursive agent, again, through grammatical metaphor:



The necessary logic of those reasons reinforces the legal analysis adopted by this dissenting opinion. It appears to support only one conclusion.

3 Taking sides

Arbitrators are generally appointed by one of the parties, to whom they owe a certain duty of loyalty. The need for arbitrators to show “their” party that they have presented his/her viewpoint accounts for some of the lengthy sections of explanation and reasoning that are found in both the awards and in the separate opinions (Vazquez and Giner 2012). It is therefore not surprising that arbitrators sometimes cast aside the cloak of neutrality and express their views with vehemence.

3.1 Emphatic criticism

It is noticeable that although most instances of criticism are mitigated, as we have seen above, some contain a blend of mitigation and force. The following example is representative, in that the writer purports to express respect, then expresses strong disagreement phrased in impersonal terms:



With the greatest respect, it is a fundamental error to construe the MFN treatment clause as denoting the treatment to be accorded to all aliens as a general obligation by virtue of customary international law.

3.2 Vehement disagreement

Dissenting arbitrators often position themselves as “obliged” to disagree. As we have seen above, this may be expressed in such a way that it involves a delegation of responsibility. However, at times it is expressed very directly, using the first person:



I am unable to concur with this reading of Article 25(2)(b).

At some point during their opinions, all the arbitrators who publish separate opinions have to indicate why they are taking this dramatic step. This happens most usually at the beginning and end of the opinion: at the beginning, to offer an explanation for the purpose of the document, and at the end as a performative statement. As we have seen, such statements may be softened by some positive reference to the majority:



For these reasons, I am unable to join the Award and must with all due respect issue this concurring and dissenting opinion.

However, in other cases, the expression of disagreement is more vehement:



the manifold shortcomings of the Tribunal’s Award.



I part company with my colleagues, and find the Award so defective that I would be prepared to annul it.

Arbitrators mobilize a considerable range of discursive resources to stress the validity of their own point of view: “the well-recognized fact”; “crystal clear”; “the proper course”; as well as making generous use of adverbs such as “clearly”, “importantly”. They also deploy the rhetorical device of repetition for emphasis. For example, Born uses the terms “wrongful” and “wrongfully” five times (48), while Rovine habitually repeats “none” and “there is no” at least three times in one paragraph. Such repetitions sometimes exploit the “crescendo” effect analyzed by Perelman and Olbrechts-Tyteca (1969: 175) whereby the last instance carries more discursive weight than the first.



City Water was wrongfully evicted from its leased premises, and wrongfully denied the use of its assets [...]

Sometimes, the arbitrator stresses his own willingness to seek agreement in order to show the intransigence of the majority. In (42), the arbitrator stresses that he was open to negotiation, and implies that the majority was not.



I defended my opinion with clear argument in the ad hoc Committee’s deliberations, and endeavoured to reach a consensus with my colleagues. Nevertheless, I remained firmly convinced of my views, which were not congruent with my colleagues’ judgment. Hence, I am stating herein my dissenting opinion.

3.3 Emotive appeal

Although legal genres are sometimes thought to be dry, and lawyers are said to be professionally resistant to the idea that emotions should ever “cloud their judgment” (Ryan 2005: 234), the opinions studied here point to emotional involvement. Indeed, arbitrators occasionally heighten the tone of their opinion by use of pathos:



This issue […] touches on the sovereign prerogatives of a Government fighting for its very life.



The majority opinion hardly adverts to the fact that the insurrection had developed into a full-scale civil war with tragic loss of life on both sides.

3.4 Irony and sarcasm

Occasional instances of irony or sarcasm also appear to be used to reinforce the arbitrator’s divergent opinion in a more measured way. The problematic nature of irony as a strategy has often been pointed out (Perelman and Olbrechts-Tyteca 1969: 208): for irony to be effective, a minimum of agreement on both fact and value is essential, which makes it at best a risky strategy. Example (45) contrasts “extraordinary” with “leisurely” to underline the tribunal’s misreading of the situation, while example (46) exploits the notion of the “freudian slip” which betrays a subconscious wish or thought:



These conditions of civil war, in my opinion, constituted an extraordinary situation which did not admit of reliance on the type of leisurely police precautionary measures envisaged by the Tribunal.


(Abi Saab 1)

one notes a freudian slip by omission in the quotation by the Majority.

3.5 Expressions of surprise

One interesting move which dissenting arbitrators sometimes use to convey disagreement is that of expressing surprise at the majority judgment. They may even stage a rhetorical boosting of the panel’s eminence in order to heighten their surprise at its decision – a rather theatrical move which serves to dramatize their disagreement:



the underlying element of mutuality must surely have been obvious to a Tribunal of this eminence even if neither of the parties brought it out four-square in its argument.

3.6 Attribution of doubtful motivation

Although arbitrators are usually careful to preserve discursive neutrality and to avoid judgmental language, this is not always the case. These opinions contain a few instances of arbitrators who criticize the parties’ behavior (example [48]), or impute dishonorable motives to them (example [49]). We may note that these instances of criticism are phrased indirectly and couched in technical language, which might be understood in terms of reluctance to perform a bold face-threatening act:



Rather, it was the Republic which chose, through its expropriatory conduct, to depart from and render nugatory, the contractual mechanisms of the Lease Contract.



I believe the Tribunal should have [...] called for evidence as to whether this exculpatory language was designed to ensure [...] or whether its effect [...] was to exclude any liability of Sri Lanka arising from its conduct [...] including a duty to negotiate in good faith.

3.7 Rhetorical questions

One of the most notable ways in which the arbitrator reinforces his/her own point of view discursively is by use of rhetorical questions, which are classically used to rope in the audience (Perelman and Olbrechts-Tyteca 1969). We may note how the arbitrators in examples (50) and (51) rhetorically (over)load the dice before asking the question:



What is a Claimant (one might say “an ICSID Claimant”) entitled to expect of the Tribunal, and what indeed are we all, as users of the ICSID system, entitled to expect when that sort of claim is put forward?


(Abi Saab 2)

This misguided tendency [...] is undermining the credibility not only of the ICSID system, but of the very idea of objective international adjudication [...]. What is the proper and prudential course for ICSID tribunals to take in this regard?

3.8 Appeal to “simple” or “normal” reasoning

Several of the divergent opinions apply the strategy of saying that the panel failed to adopt “normal” or “simple” measures or interpretations that would have solved the problem quite simply and easily. Appropriate legal reasoning obviously lies at the heart of arbitration, and it is essential that tribunals should not only reason correctly, but provide a full account of their reasoning processes in the award (Lalive 2010). The implication that the majority reasoning is either tortuous or blatantly wrong therefore amounts to harsh criticism of the panel’s decision.



There is nothing special about this list; the items in it are simply the normal tools of treaty interpretation.

In the following case, the “simplicity” of the issue (on which the majority disagreed) is stressed by use of both a cleft sentence and intensification through repetition (italics):



What is decisive in our case is the simple, straightforward, objective fact that the dispute before this ICSID Tribunal is not between the Ukrainian State and a foreign investor but between the Ukrainian State and a Ukrainian investor.

Indeed, arbitrators often refer to “simple”, “normal” or “natural” meanings, as a way of framing a sharp criticism of the panel’s interpretations (Endicott 2000):



The natural and ordinary meaning of the phrase “with respect to” is specific, narrow, and precise.

They also write dismissively of the type of reasoning used by the majority, sometimes (example [55]) combining indications of respect with a strongly emphasized disagreement (italics):



I respectfully submit that the Claimant has not proved his claim and that it is neither rational nor legally valid to make assumptions of unknown facts (the treatment of the Poblano Group) on the basis of a chain of inferences.

3.9 Criticizing the evidence produced

The evidence presented at the arbitral sessions is obviously of crucial importance, so if it is insufficient, this provides strong grounds for dissent. Some dissenting arbitrators base their disagreement on the opinion that the evidence produced by the parties was inadequate, and that the decision made on this basis is therefore invalid:



It is only on the basis of such documents that the majority presumes that CEMSA has been discriminated against.

Another way of undermining the evidence presented is to cast aspersions on the trustworthiness of the people who presented it:



Minister Sícaro’s testimony regarding the termination decree’s political character, in my view, was patently evasive and definitely not credible.

3.10 Reduction to absurdity

The reductio ad absurdum is a rhetorical strategy related to irony (Perelman and Olbrechts-Tyteca 1969: 207; Nivelle 2008: 487) used to defeat an opponent’s argument which is also present in some of these opinions:



If the approach taken in this Award were to prevail, it would suffice for any investor from a NAFTA State to show that another State party to the same Treaty has made only one mistake or miscalculation in the administration of a tax, favoring a single national investor – whose circumstances are apparently similar – to claim and obtain a benefit from that State, to the detriment of its public finance.

Covarrubias’s argument is essentially that since, in this case, a Mexican company received a tax rebate to which it was not entitled, it is ridiculous to pay damages to a US company (which was also not entitled to a rebate) on grounds of discrimination. His analysis is structured as a reductio ad absurdum (relevant text in italics).

4 Styles and choices

The vast repertoire of moves and discursive realizations documented above might generate a sense of almost infinite potential. When we consider the opinions as separate texts, it is true that many play on a wide range of resources to create the desired effect. However, it is also noticeable that individual arbitrators tend to be characterized by particular styles, which are presumably the product of the arbitrator’s own approach to his/her professional activity and the strength of disagreement which he/she aims to express in the case at hand. The arbitral writing styles encountered in this sample range from the dry and neutral to the hot and impassioned, although it should be noted that the evaluative range tends to vary along the text of the opinion, with the strongest assertions reserved for the final sections. These examples serve to illustrate the main types of arbitral voice found in this corpus, focusing mainly on the closing sections of the opinions.

4.1 Technical, impersonal

Some arbitrators adopt what might be termed a “technical” or “technicist” approach to summing up their disagreement, which centers on specific issues of law or accounting. This tends to conserve neutrality, focusing on actions or omissions and avoiding a confrontational stance. In (59), we should note how “failure to discount” is not attributed to any agent, while the phrasing “the Tribunal shares in the error” tends to mitigate the tribunal’s responsibility for the mistake.



I view the failure to discount the set-off amount for actual rents received post-Lesivo as an arithmetical error that could be addressed at this stage of the proceedings. Insofar as Claimant’s expert erred in not discounting these rents, it is my view that the Tribunal shares in the error. Therefore, the Tribunal should correct it and I would do so. For that reason, I respectfully dissent from this part of the decision.

4.2 Rhetorical elaboration

On occasions, the opinions in this sample show evidence of considerable – almost baroque – rhetorical development. The following example, with its exaggeration, emphasis and metaphorical overload serves to illustrate the rhetorical arbitral voice. Interestingly, however, by accusing the specific tribunal of violating the principles of ICSID arbitration, Abi Saab implicitly presents himself as a defender of the institution, and of arbitration in general.


(Abi Saab 1)

In these circumstances, I don’t think that any self-respecting Tribunal that takes seriously its overriding legal and moral task of seeking the truth and dispensing justice according to law on that basis, can pass over such evidence, close its blinkers and proceed to build on its now severely contestable findings, ignoring the existence and the relevance of such glaring evidence. It would be shutting itself off by an epistemic closure into a subjective make-believe world of its creation; a virtual reality in order to fend off probable objective reality; a legal comedy of errors on the theatre of the absurd, not to say travesty of justice, that makes mockery not only of ICSID arbitration but of the very idea of adjudication.

4.3 Rational and responsible

Most of the opinions studied are characterized by their measured, responsible tone. In (61), the arbitrator provides a clear rationale for his opinion and voices his need to dissent politely, fronting it with an expression of regret.



NAFTA was designed in large measure to create a law-based system pursuant to which an investor would be encouraged to invest in Canada, Mexico and the United States. For that, an investor needs predictability and stability, encouragement from the host state, legal standards that include basic minimum protections of investments, a right of legal redress in case of breach of any of those protections by the host state, and an atmosphere in which an investor need not fear that his investment will be diminished [...]. With regret, I cannot accept my learned colleagues’ approach [...] because I believe their approach is not consistent with one of NAFTA’s key reasons for being, or NAFTA’s provisions, or customary international law.

4.4 Careful and respectful

Most opinions include asseverations of respect in their conclusions, carefully delineate the extent of the disaccord and emphasize that the failure to agree on a specific point should not affect future professional relations. In addition to reflecting conventional courtesy, the high degree of facework in these statements also points to a willingness to continue professional relations with those involved, which doubtless springs from a genuine desire to undertake further well-remunerated arbitral work.



For the foregoing reasoning, I did not agree with my colleagues’ Decision. Yet, I should affirm that my different views on certain aspects of this case did not diminish my respect and admiration of my colleagues’ learning and wisdom, which were reflected in my deliberations with them, and in their well-reasoned and distinguished Decision.

5 Conclusion

I have examined the discourses of international investment arbitration, focusing particularly on how arbitrators preserve their discursive neutrality and/or express strong commitment. To this end, I identified the moves and resources used in all the available separate opinions from the period 1987–2013 published by ICSID. This study thus represents a comprehensive analysis of the material available, and sheds further light on the discourse of investment arbitration.

Above all, my analysis points to the complex dilemmas that arbitrators face as they weigh up their need to balance their commitment to the party that appointed them against their desire to maintain positive relations with their fellow arbitrators and with the arbitral institution. A key assumption which underpins arbitration of the kind provided through ICSID is that arbitrators will be capable of resolving issues in a way that is acceptable to both parties. Obviously, the dissenting arbitrator’s refusal to express complete agreement challenges this presupposition. It has been shown how heavily hedged and mitigated much dissent is, and how often face-saving elements (the complexity of the issue, the minor nature of the discord) are stressed to lessen the discursive impact of disagreement. The dissenting arbitrator has been seen to shelter behind some other entity – a law, a treaty, a precedent, or the “spirit” of ICSID – which undermines the majority decision, rather than taking full discursive responsibility for the disagreement him/herself. Disagreement is often rendered impersonally, accompanied by fulsome declarations of respect for the persons of the tribunal. On the other hand, many familiar courtroom strategies are used to vent criticism, including rhetorical questions, mock surprise or the use of pathos. Although all of these may be regarded as evidence of commitment and overt disagreement on the part of the arbitrator, they are usually blended throughout the text with moves that suggest a desire to maintain neutrality, and it is rare to find a highly antagonistic dissenting opinion in this corpus. On a few occasions, arbitrators do choose to express their disagreement directly, using exaggeration to hammer home their points, but even these are tempered by discursive bolstering of the arbitral institution itself.

The evidence from this article thus provides a window onto the discursive choices open to arbitrators at such venues, shedding light on the practices of this high-profile yet somewhat shadowy profession. In a wider sense, it may offer interesting material for comparison with discursive practices in other areas of the legal genre colony, such as courtroom language or lawyer–client interviews. As Conley and O’Barr (1998: 9) note, the workings of power through law are not simply a top-down phenomenon: power relations sustain the legal system at many different levels and are enacted through many different relationships. Moreover, in law even more than elsewhere, power relations are sustained through discourse. The elaborate expressions of respect for colleagues and the discursive bolstering of the arbitral system reflect the practitioner’s need to reproduce the system and support its representatives even while criticizing it. As Sarat and Felstiner (1986) observed, lawyers’ work involves helping clients to restructure their perceptions in order to “facilitate a reconciliation between client objectives and the needs of legal institutions” (Sarat and Felstiner 1986: 126), and to achieve this, they must create appropriate degrees of both doubt and confidence in the legal system and its representatives. In separate opinions, arbitrators oscillate between outspoken criticism (rare) and heavily circumscribed, hedged and depersonalized criticism (frequent) of the majority decision: but even the most impassioned attacks on the majority decision (such as example [61]) are framed in a way that is respectful towards – or even protective of – the institution within which it was produced and the “face” of those involved. Rather than interpreting this apparently equivocal attitude as entailing “a heavy dose of cynicism” (Sarat and Felstiner 1986: 127), we should perhaps view this phenomenon in the light of research into discursive tensions in other professions, where similar balances have to be struck (Erickson and Shulz 1982; Meisenbach 2008; Sarangi 2010), and as providing a further angle on the pervasive phenomenon of “accounts” in social interaction (Billig 1985).

Finally, it is hoped that this study will prove useful for those involved in legal training worldwide. Investment arbitration operates across the planet and involves professionals from all legal cultures, and further analysis of its hitherto unexplored discourses should be of relevance to lawyers and legal linguists from all areas of the world.

Appendix: cases cited

Abi Saab 1. Conoco Phillips Petrozuata B.V. (etc.) v. Bolivarian Republic of Venezuela (ICSID Case No. ARB/07/30). Dissenting opinion, Georges Abi Saab.

Abi Saab 2. Abaclat and others v. Argentine Republic (ICSID Case No. ARB/07/5). Dissenting opinion, Georges Abi Saab.

Asante. Asian Agricultural Products Limited v. Democratic Socialist Republic of Sri Lanka (ICSID Case No. ARB/87/3). Dissenting opinion, Samuel K. B. Asante.

Berman. Industria Nacional de Alimentos, S.A. and Indalsa Peru´, S.A. v. Republic of Peru (ICSID Case No. ARB/03/4). Dissenting opinion, Franklin Berman.

Born. Biwater Gauff v. United Republic of Tanzania (ICSID Case No. ARB/05/22). Concurring/dissenting opinion, Gary Born.

Brouwer. Daimler Financial Services AG v. Argentine Republic (ICSID Case No. ARB/05/1). Dissenting opinion, Charles N. Brouwer.

Covarrubias. Marvin Roy Feldman Karpa v. United Mexican States (ICSID Case No. ARB(AF)/99/1). Dissenting opinion, Jorge Covarrubias Bravo.

Czar. Sociedad Anónima Eduardo Vieira v. Republic of Chile (ICSID Case No. ARB/04/7). Dissenting opinion, Susana B. Czar de Zalduendo.

Eizenstat. Railroad Development v. Guatemala. (ICSID Case No. ARB/07/23). Dissenting opinion, Stuart B. Eizenstat.

Highet. Waste Management, Inc. v. United Mexican States (ICSID Case No. ARB(AF)/98/2). Dissenting opinion, Keith Highet.

Lalonde. Sempra Energy International v. Argentine Republic (ICSID Case No. ARB/02/16). Partial dissenting opinion, Marc Lalonde.

Morales. Renée Rose Levy de Levi v. Republic of Peru (ICSID Case No. ARB/10/17). Dissenting opinion, Joaquín Morales Godoy.

Nabulsi. Hussein Nuaman Soufraki v. United Arab Emirates (ICSID Case No. ARB/02/7). Dissenting opinion, Omar Nabulsi.

Rovine. Archer Daniels Midland Company and Tate & Lyle Ingredients Americas, Inc. v. United Mexican States (ICSID Case No. ARB(AF)/04/5). Concurring opinion, Arthur W. Rovine.

Shahabuddeen. Malaysian Historical Salvors, SDN, BHD v. The Government of Malaysia (ICSID Case No. ARB/05/10). Dissenting opinion, Mohamed Shahabuddeen.

Stern. Occidental Petroleum Corporation and Occidental Exploration and Production Company v. The Republic of Ecuador (ICSID Case No. ARB/06/11). Dissenting opinion, Brigitte Stern.

Suratgar. Mihaly International Corporation v. Democratic Socialist Republic of Sri Lanka (ICSID Case No. ARB/00/2). Concurring opinion, David Suratgar.

Weil. Tokios Tokeles v. Ukraine (ICSID Case No. ARB/02/18). Dissenting opinion, Prosper Weil.

de Witt. African Holding Company of America, Inc. and Société Africaine de Construction au Congo S.A.R.L. v. Democratic Republic of the Congo (ICSID Case No. ARB/05/21). Dissenting opinion, Otto de Witt Wijnen.


Amossy, Ruth. 2009a. Argumentation in discourse: A socio-discursive approach to arguments. Informal Logic 29(3). 252–267.Search in Google Scholar

Amossy, Ruth. 2009b. The new rhetoric’s inheritance. Argumentation and discourse analysis. Argumentation 23. 313–324.Search in Google Scholar

Antaki, Charles. 1988. Analysing everyday explanation. London: Sage.Search in Google Scholar

Arribas-Ayllón, Michael, Srikant Sarangi, & Angus Clarke. 2011. Genetic testing: Accounts of autonomy, responsibility and blame. London and New York: Routledge.Search in Google Scholar

Asher, Nicholas & Eric McCready. 2007. Were, would, might and a compositional account of counterfactuals. Journal of Semantics 24. 93–129.Search in Google Scholar

Bhatia, Vijay K. 2004. Worlds of written discourse. London: Continuum.Search in Google Scholar

Billig, Michael. 1985. Prejudice, particularization and categorization: From a perceptual to a rhetorical approach. European Journal of Social Psychology 15(1). 79–103.Search in Google Scholar

Breeze, Ruth. 2012a. Dissenting and concurring opinions in international investment arbitration: How the arbitrators frame their need to differ. International Journal for the Semiotics of Law 25(3). 393–413.Search in Google Scholar

Breeze, Ruth. 2012b. With the greatest respect for my colleagues. Politeness in dissenting opinions in international arbitration disputes. In Manuel Padilla Cruz & Lucía Fernández Amaya (eds.), New perspectives on (im)politeness and interpersonal communication, 149–165. Newcastle: Cambridge Scholars.Search in Google Scholar

Breeze, Ruth. 2014. Constructing authority in international investment arbitration: Insights from separate opinions at ICSID. In Vijay K. Bhatia, Giuliana Garzone, Rita Salvi, Girolamo Tessuto & Christopher Williams (eds.), Language and law in professional discourse: Issues and perspectives, 93–108. Newcastle: Cambridge Scholars.Search in Google Scholar

Brown, Penelope & Stephen Levinson. 1987. Politeness: Some universals in language usage. Cambridge: Cambridge University Press.Search in Google Scholar

Commission, Jeffrey P. 2007. Precedent in investment treaty arbitration: A citation analysis of a developing jurisprudence. Journal of International Arbitration 24. 149–150.Search in Google Scholar

Conley, John M. & William M. O’Barr.1998. Just words: Law, language and power. Chicago: University of Chicago Press.Search in Google Scholar

Cutts, Martin. 1995. Oxford guide to plain English. Oxford: Oxford University Press.Search in Google Scholar

De Boissésson, Matthieu. 1998. Le droit français de l’arbitrage interne et international. Paris: GLN Joly.Search in Google Scholar

Dolzer, Rudolf & Christopher Schreuer. 2008. Principles of international investment law. Oxford: Oxford University Press.Search in Google Scholar

Endicott, Thomas. 2000. Vagueness in law. Oxford: Oxford University Press.Search in Google Scholar

Erickson, Frederick & Jeffrey Shultz. 1982. The counselor as gatekeeper: Social interaction in interviews. New York: Academic Press.Search in Google Scholar

Hyland, Ken. 1996. Talking to the academy: Forms of hedging in science research articles. Written Communication 13(2). 251–281.Search in Google Scholar

ICSID (International Center for the Settlement of Investment Disputes). 2006. ICSID convention regulations and rules. (accessed 12 January 2016).Search in Google Scholar

Lalive, Pierre. 2010. On the reasoning of international arbitral awards. Journal of International Dispute Settlement 1(1). 55–65.Search in Google Scholar

Lévy, Laurent. 1989. Dissenting opinions in international arbitration in Switzerland. Arbitration International 1, 35–42.Search in Google Scholar

Meisenbach, Rebecca J. 2008. Working with tensions. Management Communication Quarterly 22, 258–287.Search in Google Scholar

Nivelle, Nele. 2008. Counterfactual conditionals in argumentative legal language in Dutch. Pragmatics 18(3). 469–490.Search in Google Scholar

Perelman, Chaim & Lucie Olbrechts-Tyteca. 1969. The new rhetoric: A treatise on argumentation. Indiana: University of Notre Dame Press.Search in Google Scholar

Rescher, Nicholas. 1961. Belief-contravening suppositions. The Philosophical Review 70(2). 176–196.Search in Google Scholar

Ryan, Erin. 2005. The discourse beneath. Emotional epistemology in legal deliberation and negotiation. Williamsburg: William and Mary Law School Faculty Publications, 235.Search in Google Scholar

Sanders, Pieter. 1999. Quo vadis arbitration?: Sixty years of arbitration practice: A comparative study. The Hague: Kluwer Law International.Search in Google Scholar

Sarangi, Srikant. 2010. Practising discourse analysis in healthcare settings. In Ivy Bourgeault, Robert Dingwall & Raymond G. de Vries (eds.), The SAGE handbook of qualitative methods in health research, 398–416. London: Sage.Search in Google Scholar

Sarat, Austin & William L. F. Felstiner. 1986. Law and strategy in the divorce lawyer’s office. Law & Society Review 20, 93–134.Search in Google Scholar

Trosborg, Anna. 1997. Rhetorical strategies in legal language. Tübingen: Gunter Narr.Search in Google Scholar

Van Leeuwen, Theo & Ruth Wodak. 1999. Legitimising immigration control: a discourse historical analysis. Discourse and communication 1(1). 91–112.Search in Google Scholar

Vázquez Orta, Ignacio & Diana Giner. 2012. Contrastive study of international commercial arbitration awards and judgments: Intertextuality through metadiscourse in action. In Vijay K. Bhatia, Giuliana Garzone & Chiara Degano (eds.), Arbitration awards: Generic features and textual realizations, 171–191. Newcastle: Cambridge Scholars.Search in Google Scholar

Weidemaier, Mark. 2010. Toward a theory of precedent in arbitration. William and Mary Law Review 51. 1895–1958.Search in Google Scholar

Published Online: 2016-6-21
Published in Print: 2016-7-1

©2016 by De Gruyter Mouton