The epilogue concludes the book by summarizing the book’s key findings. In particular, it draws out five categories of implications following the book’s examination of coherence. These implications relate to fundamental aspects of international legal practice, including, crucially, to conceptual understandings of law, doctrine, epistemology, professional ethics in investor-state dispute settlement, and future institutional design in investor-state dispute settlement.

Chapter 5 examines how considerations of coherence manifest in the use of analogical reasoning by investor-state tribunals. In particular, it demonstrates through concrete examples and case studies that the persuasiveness and correctness of an arbitral award based on analogical reasoning depends on the degree of its internal coherence. It is argued that coherence in an analogical inference manifests in two ways. Firstly, in a methodological sense, coherence manifests itself in the way the adjudicator frames the legal question at issue and in the degree to which the analogy, as drawn, satisfies the elements of similarity, structural parallels, and purposiveness. Secondly, in a substantive sense, coherence manifests itself in the normative contextualisation of the legal question and in the moral appeal of the proposed interpretation derived from the analogy.

The International Law Commission has described the operation of the VCLT rule of interpretation as a ‘crucible’, whereby it is the combination of the interpretative principles contained in the VCLT rule that in each case yields the meaning of treaty provisions. Taking that view as its starting point, Chapter 4 points out that different interpretative VCLT principles may be assigned different weights in different cases, depending on the drafters’ constructed intention followed by the adjudicator in each case. The underlying idea, therefore, is that all these interpretative principles must hang together well and must be combined in a way that makes sense in the context of each case. This is a sign that considerations of coherence are apposite in the operation of the VCLT rule, acting as reasons that justify the above combination in each case and ultimately account for the outcome’s persuasiveness. In the process, the chapter identifies two key coherence-related processes: framing and normative contextualisation

Concerns of incoherence surrounding investor-state decisions have been widely discussed by scholars over the past several years and have moreover been overwhelmingly raised by state delegations during the UNCITRAL Working Group III discussions on investor-state dispute settlement reform. In the coda, the discussion is briefly brought back to its beginnings and the topic of investor-state dispute settlement reform. Are there any conclusions in relation to reform that may usefully be drawn from the book’s discussion? The coda argues that states may be required to reframe and recontextualise their understanding of the concept of coherence in light of the book’s overall analysis. In so doing, they may also be required to address coherence from the ‘bottom-up’ rather than the ‘top-down’.

Coherence is a highly valued notion in law and it is especially sought after in investor-state dispute settlement. At the same time, coherence is a largely underexplored concept in international law literature. The introductory chapter serves to set the stage for the book’s investigation. To that end, the chapter outlines: (i) the impetus behind the choice of coherence as a subject for inquiry; (ii) the principal, so-called ‘bottom-up’ perspective from which the subject of coherence is examined in the book; and (iii) the core thesis advanced in relation to the dual, substantive and methodological, nature of coherence and its role in judicial reasoning in investor-state dispute settlement.

The question with which this chapter grapples is the following: What kind of a concept is coherence and what is its content? The chapter begins by a general introduction on concepts. Three different concept types are identified: criterial concepts, natural-kind concepts, and interpretative concepts. As coherence is clearly not a natural-kind concept, the chapter analyses coherence as a potential candidate concept of the criterial kind. It identifies three elements often associated with, and deemed necessary for, the existence of coherence in a legal setting, namely: consistency, correctness, and comprehensiveness. Incidentally, these are also key concerns regarding the existing ISDS regime as expressed by state delegations and scholars. The section ultimately concludes that none of the three elements is necessary for coherence to exist in non-ideal practical situations. Based on this examination, the chapter then shifts perspectives and characterises coherence as a concept of the interpretative kind. In so doing, the chapter makes a preliminary case for the existence of a dual, substantive and methodological, dimension of the interpretative concept of coherence

Chapter 2 concluded that coherence is linked to legal reasoning, but where exactly within the process of legal reasoning does coherence fit? Chapter 3 presents and contrasts two views about how legal reasoning is deployed — the positivist view and the non-positivist (interpretivist) view. Both agree that the law carries expectations with respect to the achievement of certain values, i.e., legal certainty and substantive correctness (justice). But they disagree about the relationship between those two values when it comes to legal reasoning, hence the different models for coherence that each view gives rise to. The positivist view gives rise to a model of ‘double coherence’, whereas the non-positivist view gives rise to a model of ‘single coherence’. The chapter ultimately sides with the latter view. Two grounds are offered for this conclusion. Firstly, the core assumptions of the positivist view regarding the different processes of legal reasoning allegedly at play when one is deciding easier and harder cases do not seem to hold. Secondly, following the positivist view may result in certain methodological pitfalls for adjudicators which the non-positivist view seems to avoid.

Chapter 7 addresses the following question: How can reflexivity be promoted in the collective context of investor-state dispute settlement, so as to help bridge individually held views by arbitrators that often come into competition or conflict with one another? The response that this chapter offers is that collective reflexivity can be promoted by acknowledging the presence of moral responsibility in arbitrators and by arbitrators committing to five distinct judicial virtues, namely: faith, humility, acquiescence, integrity, and candour. Judicial virtues are habits and mental dispositions, not an equation for the courtroom. They are thus meant as a framework offering guidelines and a roadmap to develop better deliberative practices. The chapter analyses the content of each virtue and assesses observable behaviour in investor-state dispute settlement under each of them.