4.1. Typical narrative images

Bernard Jackson's target in a relatively recent paper, entitled ‘Literal Meaning: Semantics and Narrative in Biblical Law and Modern Jurisprudence’,Footnote ⁵⁷ is what he characterizes as the liberal modern law model of language-based rules. In this model, he says, ‘people can rely ultimately on their disputes being resolved by court adjudication’, which ‘involves the application of linguistic rules’, where the meaning of those linguistic rules, in turn, ‘is normally available in advance, the assumption being that the “literal” meaning is that both intended by the legislator and to be applied by the court’.Footnote ⁵⁸ Jackson argues that a very different model is discernible in the Bible. The features of this model include a view of ‘dispute settlement as essentially private’ which does not ‘necessarily involve the application of linguistic rules’, and where, significantly for my purposes, if ‘linguistic rules are used, their application is not to be identified with the notion of “literal meaning”, but rather with their narrative, contextual sense’.Footnote ⁵⁹ Jackson's dissatisfaction is with a view of language that suggests that words – whether in rules or otherwise – have a ‘literal’ or a ‘core’ meaning; that is, there are ‘situations that the words “cover”’.Footnote ⁶⁰ It would be too crude to suggest that the Hartian, MacCormickian, and Dworkinian model of legal reasoning to which the ILC Report acknowledged its debt is fairly characterized in this fashion, but it cannot be doubted that the complex family of ideas invoked by those theorists broadly falls under the canopy of the liberal modern law model. Jackson urges and offers a different approach to the liberal model, namely one that does not ask ‘what situations do the words of this rule cover?’, but inquires in the following way: ‘what typical situations do the words of this rule evoke?’Footnote ⁶¹ Formulating the question in this way, he says, signifies looking ‘to the narrative images – of situations within known social contexts – evoked by the words’.Footnote ⁶²

Jackson provides some examples of passages from biblical law where an approach bent on discovering the literal meaning necessarily fails: it captures neither the complexity of the language of the rule nor the social context in which such rules were used (i.e. in largely private settlements of disputes). The role of courts in such a context, and within such a model of understanding the language of rules, ‘was restricted to cases perceived as too far distant from the typical narrative images evoked’ such that judges were ‘expected to deploy their intuitions of justice’.Footnote ⁶³ In such a context, Jackson continues, ‘communication is characterised by the availability of common deployment of unspoken social knowledge, knowledge of what are the typical narrative images deployed within the group’.Footnote ⁶⁴ Where that ‘social knowledge is sufficiently internalised within each of the members of a particular social group’, the rules may be said to operate as a ‘semiotic code, based on the internalization of images and feelings’.Footnote ⁶⁵

Jackson goes on to consider the applicability of this ‘semiotic’ model to modern law – an important task, given that that model developed in a context very different, at least at first blush, from the role of modern courts and tribunals. He argues that, contrary to positivistic models, the semiotic model can ‘inform the manner in which law actually works’Footnote ⁶⁶ in the contemporary world. Further, he argues that both H. L. A. Hart and Lon Fuller (the latter especially) ‘unwittingly acknowledge’ that ‘narrative images still underlie much of our case law and jurisprudential theorising about it’.Footnote ⁶⁷ He invokes Fuller's famous example of the ‘upright sleeper’,Footnote ⁶⁸ and argues, convincingly, that the question of whether the upright sleeper can be found to be in violation of the rule against ‘sleeping in any railway station’ can be best understood not as an obvious instance of the rule but as a case of how close the image of the upright sleeper may be said to be to the typical narrative image invoked by the rule. Similarly with Ronald Dworkin's equally famous discussion of Riggs v. Palmer:Footnote ⁶⁹ is not the initial classification of this case as a hard case, asks Jackson, best explained by ‘the feeling that the case of the grandson was so far distant from the “normal” narrative of testate succession that this could not be a straightforward instance of the rule, notwithstanding the words in which the rule was expressed’?Footnote ⁷⁰ Even if the judgment records a justification using the rhetoric of a literal interpretation, what is really happening, according to Jackson, is that the judge is ‘appealing to the values internalised within the “common sense” of a particular community. The literal meaning of the statute is filtered through the aesthetics or values which accompany narrative images at the subconscious level’.Footnote ⁷¹

In his earlier book, Law, Fact and Narrative Coherence,Footnote ⁷² Jackson urges us to take seriously ‘the implications of linguistic scepticism in general, and doubts about the place of “reference” in particular’.Footnote ⁷³ His criticism of the use of the concept of ‘reference’ also leads him to a criticism of a correspondence theory of truth, and in this he is wide-ranging, focusing at first on Galen Strawson, but thereafter widening his critique to analytical philosophy of language in general. He shows, very convincingly, how the problem of reference and the associated correspondence theory of truth underlie an approach to legal reasoning dominated by the device of the syllogism – a ‘formalization of the process of (deductively) applying law to facts’.Footnote ⁷⁴ Chapter 4 of Law, Fact and Narrative Coherence offers an alternative; it argues that ‘rules are themselves meaningful as socially-constructed narratives, accompanied by particular (and increasingly institutionalised) forms of approval or disapproval’.Footnote ⁷⁵ In doing so, that chapter also argues that ‘the fact that legal rules have tended to become, in Western legal systems, increasingly abstract and conceptualised tells us more about the pragmatics of rule-telling (its increasing bureaucratization and specialization) than about the nature of rules themselves’.Footnote ⁷⁶

Jackson is keen to establish a divide between his approach and that of those who account for the application of law by way of the normative syllogism.Footnote ⁷⁷ Adopting his model, he says, entails seeing ‘the relationship between the general rule and the particular case’ as ‘one of inter-discursivity, not the application of a consequence to one particular referent, which the general rule states ought to be applied to all such referents’.Footnote ⁷⁸ The underlying forms of both rule and fact are, according to Jackson, narrativized, where the relationship between the two is ‘one of greater or lesser proximity, in terms of human experience’.Footnote ⁷⁹ Indeed, he argues that the ‘further the form of the “rule” moves from the narrative model to a purely abstract, conceptual formulation, the more we are likely to encounter difficulties in both the application of law to fact and the interpretation of general rules, notwithstanding the clarity in which the rule is expressed’.Footnote ⁸⁰

We must not, he says, mistake the process and expression of justification for the process of ‘how legal decisions are arrived at’.Footnote ⁸¹ If we do so, we make unwarranted distinctions between ‘a) determination of fact, b) justification of determination of fact, c) determination of law, d) justification of determination of law, e) application of fact to law, and f) justification of application of fact to law’.Footnote ⁸² Instead, laws should be ‘regarded as a particular form of narrative representation of human behaviour’ that is then institutionalized as a ‘sanction’ signifying the community's approval or disapproval of that narrative.Footnote ⁸³ Further, there are two forms of narrativization: the first relates to the semantics of the telling of stories in court (their content), and the second to the pragmatics of that storytelling – that is, the process of persuading the adjudicator of the truth of those stories.Footnote ⁸⁴ Jackson illustrates the role of the narrative model, tracking in close proximity to human experience, and thus also exemplifying the ‘inseparability of law, fact and application’ in the now often cited passage of Lord Denning's judgment in Miller v. Jackson.Footnote ⁸⁵ It is instructive to set this passage out in full and to follow Jackson's analysis of it. Here is the passage first:

The image of the intruding outsider is contrasted, negatively, with the overwhelmingly positive cricket-playing community, which is said to exemplify the very best of traditional practices. The newcomer seeks the judicial disapproval of someone in authority who him- or herself, should he make a ruling in favour of the newcomer, would be interfering in a similarly disapproving manner to that of the newcomer. Further, such interference, causing the disappearance of traditional practices, is said to be likely to lead to young men turning to implicitly destructive forms of pleasure-seeking.Footnote ⁸⁷ As Jackson concludes, ‘Lord Denning has clothed his opposition to the application of the legal rule in a vivid, narrative presentation of the facts,’Footnote ⁸⁸ and one, moreover, that already includes within it modes of persuasion. Of course, as he acknowledges, the example is a somewhat extreme one, but, significantly, not because of its use of the narrative mode, but because of its invocation of notions (such as the delights of village cricket) that are not recognized as being capable of being ‘used publicly in the process of justification’.Footnote ⁸⁹

Indeed, the above example sets the scene nicely for Jackson's argument that the ancient forms of expression of rules – for example ‘If a thief is found breaking in, and is struck so that he dies, there shall be no blood guilt for him’Footnote ⁹⁰ – and thus also more recognizably narrativized and concrete than the ‘modern, abstract legal rule’, have not only not completely died out (e.g. there are examples of the use of narrative illustrations in modern criminal codes), but in fact underlie and explain the understanding of rules expressed in that modern, abstract form. Our understanding of rules expressed in such a manner is explicable on the basis of a stock of typical narrative images – narrativized models of behaviour – that contain within them ‘some tacit social evaluation: that such behaviour is good, bad, pleasing, unpleasing etc.’.Footnote ⁹¹ Citing George Fletcher's exposition of the ‘subjective criminality’ (i.e. that of attributing evil minds to agents when they act in certain ways) that infused the law of theft,Footnote ⁹² Jackson argues that however we analyse it, our understanding of legal language is explicable on the basis of the ever developing stock of ‘collective images’ which ‘appear to be socially-constructed narrative models of human experience’.Footnote ⁹³

The conclusion, in relation to decision-making, is simple, but very powerful: ‘decision-making in adjudication consists in comparing a narrative constructed from the facts of the case with the underlying narrative pattern either explicit in or underlying the conceptualised legal rule’.Footnote ⁹⁴ Where, as in modern Western legal systems, the formulation of the rules becomes ‘increasingly abstract and conceptualised’,Footnote ⁹⁵ the more difficult is the task of adjudication.Footnote ⁹⁶ There is much to be said, he concludes, for the formulation of ‘particular rules’ that ‘build upon social experience of typical behaviour-patterns, accompanied by what are considered appropriate institutional sanctions’.Footnote ⁹⁷

There is a great deal more that can be gleaned from Jackson's work, but I do not have the space here to elaborate any further. The important point to take is that of the inseparability of normative consequences (expressed in justification), from the process of fact construction in a narrativistic guise, and, equally, the inseparability of our understanding of rules – or, more generally, the language of law – from our experience of social life as members of communities in specific times and places, where that experience is at least partly expressed in the institutionalized approval or disapproval of certain recurring behaviour patterns.

4.2. Constructing facts

The leap from Jackson to the work of Geoffrey Samuel, particularly Epistemology and Method in Law,Footnote ⁹⁸ is not a great one. Like Jackson, although in different terms and by reference to a different set of problems and contexts, Samuel criticizes the ignorance of the role of images (he calls them ‘facts’) in legal reasoning. Samuel's specific problematic is legal epistemology. He argues that contemporary theories of legal knowledge suffer from an overdose of the rule-model in which legal knowledge is said to be knowledge of rules, ‘that is to say, normative propositions capable of being expressed in symbolic language’.Footnote ⁹⁹ Instead, Samuel says, we should pay more attention to non-symbolic forms of knowledge: that is, the life of facts and fact-construction. Samuel's book is a philosophically thorough inquiry into the relationship between rules and facts, and I cannot, in this brief section, do it proper justice. My focus will be exclusively on chapter 5, in which Samuel criticizes a model of legal reasoning based on the more or less straightforward application (by recourse, for example, to the legal syllogism or a kind of legal hermeneutics) of rules to factual situations.

‘In order to relate to factual situations’, says Samuel, ‘rules must contain within them the means by which one can move from pure norm . . . to the world of social fact’.Footnote ¹⁰⁰ Rules, he says, ‘delimit facts. They describe areas and boundaries’.Footnote ¹⁰¹ But acknowledging this requires us to consider what this world ‘beyond rules’ is like. Consider, he says, the development of the law of negligence from Donoghue v. Stevenson Footnote ¹⁰² to Grant v. Australian Knitting Mills.Footnote ¹⁰³ In the first case, as is well known, Mrs Donoghue was said to have suffered nervous shock on account of discovering a snail in her ginger beer (manufactured by the defendant). In the second, Mr Grant suffered acute dermatitis as a result of wearing underpants (also manufactured by the defendant), which contained an excess of a chemical harmful to human skin. There is nothing inherent – even if trained lawyers by habit may think so – in the neighbour principle – that is, nothing within the proposition itself which may indicate that the two cases are related. Instead, there are relationships between images (once again, Samuel calls them ‘facts’): between the harm (nervous shock and acute dermatitis), the means (ginger beer and underwear) and the subjects (consumers and manufacturers). Where the relationships between them are sufficiently close, such that the latter situation can be imagined as an instance of the former, then the cases are sufficiently similar to warrant the use of the same justification for the decision (i.e. the applicability of the neighbour principle). The point is, says Samuel, that it matters how facts themselves and the relationships between them are imagined.Footnote ¹⁰⁴ He offers the following illustration:

But it is no mere, though of course also no small, matter of how one envisages facts and relationships between them – a matter, as he calls it, of analogy or analogical thinking. What matters also is how one categorizes facts. In the case of Mrs Donoghue, for example, one needs to classify the harm, the persons, and the things.Footnote ¹⁰⁶ What kind of injury or damage is nervous shock? Does it matter that Mrs Donoghue is an elderly woman? What kind of product is ginger beer? The very asking of these questions, says Samuel, is evidence enough of the dynamic relationship between facts (as in the Donoghue case) and certain categories taken to be fundamental for the purpose of classification (e.g. persons, things, actions, and damages). The explanation of such a cognitive process is, as Samuel acknowledges, by no means an easy feat. But the difficulties, at the very least, should point us to the insight that ‘the linguistic proposition cannot in itself ever contain information about the imagery which surrounds the actual application process to the facts’.Footnote ¹⁰⁷ Rules, for Samuel, are by no means useless; they help us, in a limited way, to orient ourselves. But what explains the orientation – what drives it – is the work of images, of analogies between sets of images, and that of the framing of the image in the first place. Framing, for example, involves the scale of time within which some event is delineated as a stand-alone event. For instance, in the above case of the oil tanker, ‘one can look at the act of the captain vis-à-vis the discharged oil and the distressed ship or at the wider picture of a proprietor sending out his ships and cargoes’.Footnote ¹⁰⁸ Adopting the latter, for example, is more likely to lead to the imposition of a normative structure that focuses on ‘the activity of the control of things’.Footnote ¹⁰⁹ Another feature of the work of images is the relatively common device of an imaginary bystander, or the reasonable person: this is not a device explicable solely or even persuasively by the alleged priority given to ‘ordinary meaning’. It is better explained as another way of constructing the facts: of making the images appear in a certain way, of using them to delimit the availability of normative structures.

As I have noted above, Samuel's picture of the work of images, or fact construction, in legal reasoning is more complex than I can afford to explicate here. He spends, for example, a lengthy chapter on the role of taxonomies – allowing for the categorization of facts under certain identifiable clusters of rules (e.g. family law or consumer law) – where those taxonomies are themselves subjected to thorough historical analysis (dating back to the Institutions of Gaius). Perhaps even more importantly, he presents six schemes through which the social sciences have traditionally constructed facts: the causal scheme, the functional scheme, the structural scheme, the hermeneutical scheme, the actional scheme, and the dialectical scheme – all of which interrelate in complex ways, and all of which also may be seen as tools for an understanding of how different methods for fact construction may perhaps better serve certain purposes rather than others.Footnote ¹¹⁰ What is important for my purposes is the model of legal reasoning endorsed by Samuel on the back of his analysis. Ultimately, Samuel thinks of legal reasoning as ‘a question of comparing isomorphies: that is to say, of comparing formal patterns’.Footnote ¹¹¹ Legal reasoning, then, for Samuel, is a complex instance of constructing, categorizing, and comparing factual patterns in the context of certain traditionally accepted terms (e.g. injury, damage, consumer, manufacturer, etc.) and connectors (e.g. cause, intent, motive, etc.), the operation (and thus also our understanding) of which is fundamentally inseparable from that complex process of fact construction. Foreshadowing, for a moment, the importance of specific institutional contexts (discussed in part 5 of this paper), we can supplement Samuel's account by suggesting that the function of the above-mentioned traditionally accepted terms and connectors changes with respect to their institutionalized history within particular communities – that is, the factual patterns accompanying terms will differ as between specialized communities of practitioners (e.g. different factual patterns accompany the concept of intention in the practice of criminal law and tort law).

4.3. Pooling the threads

Where Samuel offers a more sustained epistemological argument for the inseparability of our understanding of rules from their factual adaptability, Jackson provides us with an account of the social dimension of our understanding of legal language. In the latter case, it is our ever-changing and ever-developing stock of narrative models of typical behaviour patterns that we share as members of specialized communities, and the approval or disapproval of which is institutionalized in the form of sanctions. In the former, there is a more extensive focus on the relationships between images, and inextricability of the work that goes into the framing of an image, and constructing analogies between sets of images, from the imposition of normative structures on the set of facts in question. In both theorists, however, there is a common thread: a thorough and deep critique of a model of legal reasoning that relies on a chronological account of determining the applicable rules to some facts, applying those rules to those facts, and then justifying that application. Although their work, I believe, is to some extent representative of rule-sceptical trends in the recent debates over the methodology of comparative law,Footnote ¹¹² it is also work that deserves to be singled out for its sustained philosophical engagement with this alternative model of legal reasoning. That model, moreover, is not one that relies, exclusively or even dominantly, on appeal to the alleged purposiveness of reasoning about rules. Rather it is a model that brings to light the social epistemological dimension of our understanding of legal language. Ultimately, it is a model that comes closer to a vision of legal theory that concerns itself not with the analysis of the content of rules and their logical (or illogical) interrelationships, but rather with a refocusing of the legal theorist's concern away from ‘the law itself’ and towards the use of the language of law as a resource in the exercise of judgement. It is now time to consider how this alternative model of understanding legal language may lead us to the invocation and also the pursuit of a different set of system values.

5.1. Factual mapping

Much of what I have already said in section 4 of the paper is directly translatable to this section. The factual mapping of rules involves explaining the role of images, analogies between images, and the framing of images, as well as the stocks of typical narrative images, that underlie our understanding of rules and principles, and thus of their factual adaptability. Examples of how effective such work can be may be gleaned from the above cited work of Bernard Jackson, George Fletcher, and Geoffrey Samuel. Given space restrictions, I can only point to those works, and stress the necessity of a review of the normative conflicts identified by the ILC Report from the perspective of factual mapping. One example of how this might be conducted, itself partially acknowledged by the ILC Report,Footnote ¹¹³ is in the context of the fragmentation that is said to occur through conflicting interpretations of general law. Here, the Report cites an alleged normative conflict between the two different tests for the responsibility of states for domestic unrest, in Nicaragua v. United States of America Footnote ¹¹⁴ and Prosecutor v. Duško Tadić.Footnote ¹¹⁵ In the former case the test was expressed as one involving examination of ‘effective control’ and in the latter as one involving ‘overall control’. As the Report notes, the ICTY considered that ‘effective control’ was too high a standard, preferring a test which held a power accountable where that power has a ‘role in organising, coordinating, or planning the military actions of the military group’.Footnote ¹¹⁶ By contrast, the ICJ held that the ‘United States had not been held responsible for the acts of the Nicaraguan Contras merely on account of organising, financing, training and equipping them’Footnote ¹¹⁷ – more was required, namely that the United States exercised ‘effective control’ over those contras.

The Report warns that ‘differing views about the content of general international law’ can both ‘diminish legal security’ and entail that ‘legal subjects are no longer able to predict the reaction of official institutions to their behaviour and to plan their activity accordingly’.Footnote ¹¹⁸ Further, it says, the rights enjoyed by legal subjects will then be unequal as between jurisdictions, and in the absence of an appeal court (as in domestic systems) there is no other way to avoid this conflict than to encourage international tribunals to ‘coordinate their jurisprudence’.Footnote ¹¹⁹ What such warnings assume is that the two decisions are factually commensurable – that it is possible, in other words, to compare the two tests and thus, also, the two different justifications – from a vantage point of the content of general international law in which that content can be given legal meaning outside its factual adaptability. Significantly, the Report itself acknowledges this shortcoming. It states in a footnote that the identification of this alleged normative conflict ‘need not be the only – nor indeed the correct – interpretation of the contrast between the two cases’.Footnote ¹²⁰ It acknowledges that ‘some commentators have suggested’ that ‘the cases can also be distinguished from each other on the basis of their facts’ and that ‘in this case, there would be no normative conflict’.Footnote ¹²¹ But it insists that ‘the point of principle remains, namely that it cannot be excluded that two tribunals faced with similar facts may interpret the applicable law differently’.Footnote ¹²² The point here is that the Report assumes that it is possible to separate a norm (if you prefer, a legal test or a justification) from its factual instantiation where the work of legal reasoning is implicitly restricted to the application of the content of a norm to some set of facts taken – and this is crucial – to fall under the same subject matter. In contrast, what is required is an acknowledgement of the process by which facts are constructed and then connected to the normative justification proffered. Thus, in the above case, it is a matter of considering how the constructed facts in the Tadić decision were thought to fall within the typical narrative images evoked by the normative justification used – that is, that of ‘overall control’ – and, simultaneously, how those constructed facts were thought to be sufficiently dissimilar from the typical narrative images evoked by the normative justification used in the Nicaragua decision, namely that of ‘effective control’. There is, in other words, no such thing as an identifiable normative conflict that operates in isolation from the manner in which some sets of facts are constructed, all set against, as the Report elsewhere acknowledges, a different set of purposes and the values of two different areas of law with two different stocks of accumulated typical narrative images with which the understanding of the normative justifications used is intertwined. Put another way, the positing of a normative conflict arises only when we establish the same factual platform, and what the process of factual mapping can reveal is how facts are construed differently in different institutions.

Factual mapping, however, is not enough on its own; what is necessary is that that analysis is conducted by reference to the specific institutional contexts of the relevant bodies. Speaking broadly, what is required is an account of the institutionalization of the factual adaptability of rules and the normative justifications that correlate with that adaptability, within members of a specific legal community (such as those, as in the above cases, that surround the work of the ICJ and ICTY). I turn to a brief exposition of that social theory in the next section.

5.2. Specific institutional contexts

I do not have space in this paper for a full review of the immense literature on institutions and institutional design.Footnote ¹²³ My focus is not, in any event, on our understanding of institutions and institutional design per se, but on its inextricability from the operation of legal reasoning and thus, also, from our understanding of legal language. To that effect I shall consider, very briefly, two recent works: Robert Summers's Form and Function in a Legal System,Footnote ¹²⁴ and John Bell's Judiciaries within Europe.Footnote ¹²⁵

In Form and Function in a Legal System, Summers criticizes theories of law that claim that legal systems are ‘essentially a system of rules’,Footnote ¹²⁶ identifying H. L. A. Hart and Hans Kelsen as the primary exponents of that theory. His theory of form is applied to what he calls the diverse functional units of a system, including institutions, legal precepts (rules and principles), non-preceptual species of law (such as contracts), interpretative and other methodology, sanctions, and remedies.Footnote ¹²⁷ According to Summers, one cannot reduce the explanation of these functional units into sets of rules. Rather, one must employ a form-oriented analysis that allows one to break down the various elements of functional legal units, which include its purposes, overall form, constituent features thereof, and complementary material.Footnote ¹²⁸ The overall form that comprises all these elements is represented as ‘the purposive systematic arrangement of the unit as a whole: its organizational essence’.Footnote ¹²⁹ Identifying and reflecting on the components and their interrelations within a functional unit allows one, among other things, to ‘organise further the mode of operation and the instrumental capacity of the unit’.Footnote ¹³⁰

In the case of the legislature, for instance, ‘internal committee structures and operational procedures within a legislature must be designed and internally coordinated to facilitate the study, debate, and adoption or rejection of proposed statutes’.Footnote ¹³¹ Not only do the various kinds of forms within functional units need to be considered, but also the relationships between functional units themselves; so, he says, there are devices that ‘consist of basic operational techniques that integrate and coordinate institutions, precepts, methodologies, sanctions and other functional units’, where each of those devices is a ‘formal organizational modality of wide-ranging significance’.Footnote ¹³²

Thus, unlike rule-oriented analysis, form-oriented analysis of a legal system does not analyse the ‘contents of those reinforcive rules that are taken to prescribe the facets of functional legal units generally’.Footnote ¹³³ Instead, it recognizes that there ‘can be no legal content without form’,Footnote ¹³⁴ placing emphasis on the need for the rational design of the components of functional units in order to fulfil the purposes of those units. Without, for example, a well-designed floor debate, statutes are less likely to beget good laws.

A proper study of Summers's book would need to engage in the dense detail of analysis of the formal components of an enormous range of functional legal units. For my purposes here, the important point is that Summers provides us with a theoretical framework within which we can readily acknowledge that rules themselves, as well as methodologies for making, interpreting, applying, and generally understanding them, cannot be understood outside and in neglect of their institutional life. Further, the set of components with which Summers analyses institutions can, I think, be very profitably applied to enhance and improve the institutional environment in which legal work is performed – but, and this is vital, we can only do so once we provide an account of what that institution should be striving for (an account I sketch in the next section).

John Bell's book, Judiciaries within Europe,Footnote ¹³⁵ though primarily involving a comparative study of the judiciary in France, Germany, Spain, Sweden, and England, is significant for my purposes in its enunciation and endorsement of an institutionally oriented methodology for the understanding and explanation of the operation of judicial reasoning. There are, says Bell, three possible ways of approaching the understanding of judicial reasoning: first, the personal perspective, which ‘looks at the way individuals perceive their role and career’; second, the institutional perspective, which ‘looks at the judiciary as a collective and examines the way in which the structure of the career and the organization of the judges, as well as legal processes, affect the judiciary as a social institution’; and third, the external perspective, which ‘looks at the judiciary from the perspective of its impact on the wider world’.Footnote ¹³⁶ Bell's preference for the institutional perspective is not merely pragmatic; it is not, in other words, merely a matter of facilitating what he considers to be an insightful comparison between judicial institutions. The institutional perspective, he says, is fundamental because ‘it relates to the nature of law . . . because this is how one operates as a legal actor . . . and because it is how the law relates to the wider world’.Footnote ¹³⁷ Like Summers, Bell argues that the law ‘is something more than simply a system of rules or legal standards . . . [it] is as much about practices, what people do, as about what they think’.Footnote ¹³⁸ ‘On the one hand’, says Bell, ‘legal culture is a pattern of behaviour or an activity’ and ‘on the other hand, there is a set of ideas and values, which are communicated through language and signs that express attitudes and values towards the activity.’Footnote ¹³⁹ Bell adopts the concept of law as institutional fact, first introduced by Neil MacCormick and Ota Weinberger,Footnote ¹⁴⁰ and argues that ‘law is an interpretive reality’ that is preceded and pervaded by institutionally based practices within the legal community.Footnote ¹⁴¹

Thus one can only understand judicial reasoning by way of an ‘institutional fact analysis’ which focuses ‘attention on the judge as an actor whose actions are invested with meaning by the legal community through shared understandings’, only ‘some of which are expressed in legal norms’.Footnote ¹⁴² We cannot separate our understanding of instances of judicial reasoning from the expectations placed on those judges in specific institutional contexts in which they act as part of a delimited legal tradition, for to do so would be to disregard their role as institutional legal actors operating within a specific legal community and working within a specific institutional culture. One must first understand the various kinds of institutional pressures on a judge to ‘interpret legal texts and perform legal procedures in ways that are considered appropriate not just by her, but by the legal community’.Footnote ¹⁴³ And one must understand how judges come to acquire and internalize, over long periods of time, that sense of appropriateness, embodied in activities and practices that themselves give rise to ‘norms and standards for why the activity should be conducted in the future’.Footnote ¹⁴⁴ As he emphasizes, ‘this structure of organizational learning does not deny change, but seeks to understand how deeply change operates’.Footnote ¹⁴⁵

All this will not come as a surprise to members of the ILC study group, but it is important to see that its significance does not lie merely in comparative law methodology. For if we are to take the increasing diversification and specialization of international law seriously as a social phenomenon, then we must have a theoretical framework in which we can understand the specific social dimension of the development of distinct legal communities. Further, we must first understand the long-term acquisition of a common sense of appropriateness of certain activities utilized by the actors within those communities, including the peculiar habits of factual construction that result in, and are informed by, stocks of typical narrative images that accompany the factual adaptability of rules used by those actors.

The factors shaping the judiciary that Bell outlines in his last chapter can serve as a useful beginning for a social theory of that common sense of appropriateness shared among members of international legal communities divided into the various identifiably distinct clusters of international law. Summers's work, in turn, can serve as a compendium of tools with which to dissect the forms of institutional life. Both works can serve, more generally, to support the argument for the inextricability of the form of rules and the procedure for their making, applying, and justifying within specific institutional contexts – with the proviso, of course, that we remain conscious of the epistemological artificiality of such categories.Footnote ¹⁴⁶ Combining the more epistemologically minded accounts of legal reasoning provided by Bernard Jackson and Geoffrey Samuel with the attention paid to specific institutional forms by Robert Summers and John Bell can free legal theorists of the international legal order from the prison of ‘the law itself’, and direct international legal theory towards an epistemologically and socially rich account of legal work performed in international legal institutions.

5.3. Rethinking system values: the priority of responsiveness

As I have already mentioned, the ILC Report, particularly in its conclusion, identified and acknowledged the importance of the systemic value of the responsiveness of a legal system. It made that acknowledgement specifically in the context of the values of legal pluralism. There is, however, another – much less cited – source of important work on the notion of the responsiveness of law, namely the work of Philip Selznick and Philippe Nonet.Footnote ¹⁴⁷ It is not a mere happy coincidence that I invoke Selznick's work immediately after stressing the importance of a community-oriented understanding of the operation of judicial reasoning. In his own work, most accessible in The Moral Commonwealth,Footnote ¹⁴⁸ he stresses the importance of a conception of community composed of the following elements: historicity, identity, mutuality, plurality, autonomy, participation, and integration. It is a conception, potentially very usefully supplemented by some of the more recent work of Roger Cotterrell,Footnote ¹⁴⁹ that can assist in developing the social theoretical approach to judicial (but more generally, legal) reasoning that I invoked by reference to Summers and Bell above.

But it is not this aspect of community on which I wish to focus here, but the notion of legal responsiveness. In the above cited work of Selznick and Nonet, the authors argue forcefully that the study of ‘the foundations of law’ cannot be divorced from ‘the place we give law in society’, and in that spirit they call for an integration of legal, political, and social theory.Footnote ¹⁵⁰ In their book they offer their own view for ‘assessing the worth of alternative modes of legal ordering’,Footnote ¹⁵¹ that is, ultimately, for assessing the place of law in society. In setting the scene for responsive law, which they offer as the mode of legal ordering against which the current states of affairs should be evaluated, they criticize two other identifiable modes, namely repressive law and autonomous law. In the case of the former, they argue that ‘every legal order has a repressive potential because it is always at some point bound to the status quo and, in offering a mantle of authority, makes power more effective’.Footnote ¹⁵² Under a repressive form of legal ordering, ‘short shrift’ is given to ‘the interests of those governed’, resulting in their position becoming particularly ‘precarious and vulnerable’.Footnote ¹⁵³ They acknowledge that to some extent all modes of legal governing are repressive, and that the emergence of that repression depends on many factors, including ‘the distribution of power, patterns of consciousness and much else that is historically contingent’.Footnote ¹⁵⁴ Nevertheless, there are identifiable forms of avoidable repression, particularly where the use of coercion is unrestrained and results in the suppression of deviance and the putting down of protests.Footnote ¹⁵⁵

The emergence of autonomous law, the second alternative mode of legal ordering, assists in ‘taming repression’.Footnote ¹⁵⁶ More commonly referred to as the rule of law, such taming is made possible when ‘legal institutions acquire enough independent authority to impose standards of restraint on the exercise of governmental power’.Footnote ¹⁵⁷ Such autonomous institutions must themselves have only ‘qualified supremacy’, and be subjected to ‘defined spheres of competence’.Footnote ¹⁵⁸ But there is a price, ultimately too high according to the authors, for the preservation of this kind of institutional integrity. Sharp lines are drawn between politics and law and thus, also, between the legislative and judicial function.Footnote ¹⁵⁹ The legal order is understood as a model of legal rules which does ‘enforce a measure of official accountability’, but also ‘limits . . . the creativity of legal institutions’.Footnote ¹⁶⁰ Regularity and fairness, rather than substantive justice, become ‘the first ends and the main competence of the legal order’.Footnote ¹⁶¹ Finally, ‘fidelity to law’ is ‘understood as strict obedience to the rules of positive law’.Footnote ¹⁶²

Thankfully, however, the autonomous mode of legal ordering contains within it the seed for the development of responsive law. Where law is responsive to social needs, it is ‘competent as well as fair’, it helps to ‘define the public interest’, and it is ‘committed to the achievement of substantive justice’.Footnote ¹⁶³ Such an approach demands, among other things, the enlargement of legal knowledge – an enlargement I invoke in this paper via the inextricability of first, normative justification and factual adaptability and, second, both of those and their instantiation in specific institutional contexts. The vision of a responsive legal order is one which takes ‘affirmative responsibility for the problems of society’.Footnote ¹⁶⁴ The ideal of responsive law does not entirely replace the warnings of repressing legal ordering and the aims of autonomous law, for it recognizes that these levels of development may at times be historically necessary. It does, however, move the ideal beyond them, calling, ultimately, for ‘larger institutional competencies to the quest for justice’.Footnote ¹⁶⁵

The vision of responsive law, then, works in tandem with the understanding of legal language developed in section 3 of this paper. Legal language cannot be analysed for normative inconsistencies outside the utility of the factual adaptability of rules and their correlative normative justification for the resolution of social problems within specific communities and institutional contexts. Indeed, it is perhaps only in the development of specialized communities, imbued with an institutional culture that takes responsibility for social problems falling within its realm, that that responsiveness can be best realized. But that responsiveness will not be achieved as long as we focus unduly on the rule of law, accompanied by the values of legal security and predictability. And, as I have argued throughout this paper, the first step in extricating ourselves from the stifling nature of principles dear to the autonomous mode of legal ordering is to recognize that the language of law cannot be understood to contain a content independent from its factual adaptability – that, on the contrary, its successful use depends on the accumulation, within specialized communities, of stocks of typical narrative images that necessarily accompany the life of the community itself. Understanding legal language in this way helps us to see it as a set of resources used in the exercise of judgement by legal officials, and thus to consider, also, how that judgement – how that institutionally located process of legal work – can be more responsive to the demands placed on the specific institution within which it operates. Understanding the forms of those institutions – as we can, should we learn from the work of John Bell and Robert Summers, as well as the institutional design literature in general – can help us to see what kinds of procedural changes and what sorts of resources we might need to introduce in order to assist those institutions to maximize their responsiveness.

No doubt much work needs to be done on the ideal of responsive law. My primary purpose in this paper has been to show how that ideal goes hand in hand with the understanding of legal language that I provide in section 4 of the paper. My own feeling is that the ideal of responsive law needs to be buttressed by a rich theory of vulnerability –that is, a theory of vulnerability that does not restrict itself to the kinds of harm that individuals may suffer, but also takes seriously the unique forms of harm suffered by communities. Moreover, we need to make more of an effort to pay attention to poor and isolated individuals, as well as poor and isolated communities, both of whom, no less in international law than in domestic law, continue to suffer from a lack (often an absence) of representation and integration. It is only such a rich theory of vulnerability that can inform the substance of an account of international social justice – resulting, hopefully, in investigations as to how we can design our institutions so that they are responsive to those forms of vulnerability.

5.4. An agenda for the ILC?

Before proceeding, in the light of the above, to suggest, modestly, an agenda for the ILC, it will be in the spirit of this paper to comment on the specific institutional context of the ILC itself. Of course, this paper is not an occasion for a detailed review of the literature on the institutional form of the ILC,Footnote ¹⁶⁶ although such a study, particularly in the light of Summers's recent work, would, I wager, be of great value.Footnote ¹⁶⁷ Nevertheless, to ignore the institutional context of the ILC altogether would be to eschew the bottom-up perspective for which this paper argues.

As Ian Sinclair points out, and as is well-known, although the ILC was established in 1947, the ambition for codification and progressive development of international law began much earlier, with the Hague Peace Conferences of 1899 and 1907, the Council of the League of Nations of the Committee of Experts for the Progressive Codification of International Law (established in 1924), the Hague Codification Conference in 1930, and the Committee of Seventeen, to mention but a few.Footnote ¹⁶⁸ The historical perspective is important, as a detailed engagement with it would, I believe, reveal that many of the current problems faced by the ILC have roots in the original vision (and early attempts at the institutionalization of that vision) for the codification and progressive development of international law – aims which may have had more purchase in earlier times.

In providing an institutional sketch of the ILC, it is important to consider the intimate relationships between the various elements, such as the composition, the selection of topics, and the working methods of the ILC. Anyone involved in the ILC will find it difficult to deny that all three are often highly politicized – no doubt a somewhat inevitable result, at least partly because of the way in which the ILC is overseen by the UN General Assembly. Composed, as it is, of academics, practitioners, politicians, civil servants, and diplomats – in total, 34 members from Africa, Asia, America, and Europe, elected for five years from lists submitted by national governments – the ILC has not always found it easy to co-ordinate the full involvement of all members, given their many pressing commitments at home. Achieving consensus on issues such as the management of international watercourse systems reveals the undeniably stark political nature of the process – not only of topic selection, but also of the manner of achieving consensus, for example, as was noted by Ramcharan, ‘to push a decision through by a majority vote is a sure way of killing it in the General Assembly or at a subsequent codification conference’.Footnote ¹⁶⁹ Despite the safeguards in Articles 2 to 8 of the ILC Statute – requiring the election of independent members, and ‘in no sense as representatives of governments’ – the reality has been that the ILC has included and continues to include foreign-ministry personnel and MPs.Footnote ¹⁷⁰ The debate over part-time or full-time status mirrors the membership issue – meeting for three months a year in Geneva is likely to result in a conference aimed at political compromise rather than a sustained scholarly exercise of the requisite epistemological and social complexity. Of course, no one would deny the important work achieved by the ILC. The question, in the context of this paper, is whether the ILC is well placed to deal with the peculiar problems thrown up by the increasing specialization of the international legal order – problems, as I have sought to show, that place in some doubt not only the theoretical underpinnings of the aims and objectives of the ILC, but also its institutional arrangements. Crucially, in the light of the defence of the system value of responsiveness in this paper, serious consideration needs to be given to the representation of non-state interests, and, in particular, as I have noted above, to those poor and isolated individuals and communities who not only have little say in topic selection, but who, it would appear, are not generally consulted when it comes to solutions that could otherwise result in an improvement of their legal empowerment.Footnote ¹⁷¹

Furthermore, if the approach to the understanding of legal language that I describe in this paper is accepted, it places the very viability of the codification of a general international law in jeopardy. For why should we attempt to codify general international law if there is no such thing as a universal international legal language? Less controversially, that approach would indicate that the production of general international legal instruments needs to be supplemented by the institutional design of the international legal order, such that the products of codification are given the opportunity to acquire institutional histories.Footnote ¹⁷² One of the criticisms made of jus cogens norms is that they lack sufficiently robust meaning to provide effective constraints on decision making.Footnote ¹⁷³ Against the background of the above approach to legal language, this is not surprising; to be effective, jus cogens norms need institutionalization – they need to be internalized by actors within specific institutions, whose habits of factual construction result in typical narrative images that accompany those terms and then inform and direct the process of legal work in that institution. I would not wish to argue that no useful role is played by the influence that attempts at codification of a general international law have on the construction of an international community – clearly, appearances and experiences matter, so that even the thinnest appearance and the merest experience of international togetherness is better than none. However, we cannot hope to control the life of that instrument of codification, for it only acquires a life (i.e. it only acquires meaning) within specific institutions.

Given the above remarks on the institutional context of the ILC – and thus, also, the requisite changes, to be worked out in more detail in later work, that may be required by way of institutional arrangement in order to accommodate these recommendations – I modestly recommend the following agenda for the ILC:

1. 1. review the Report on the fragmentation of international law from the perspective of deep coherence, namely one in which the inextricability of normative justifications from factual adaptability within specific institutional contexts is recognized;

2. 2. initiate a research project that focuses on the understanding of legal work undertaken within the various institutionally distinct clusters of international law – that being an understanding that will require the use of the methodology of deep coherence; and

3. 3. adopt the responsiveness of international legal institutions as the system value of the international legal order, and take steps towards investigating, in partnership with those institutions, the requisite changes in institutional design that may need to be made in order to make those institutions more responsive to the social problems faced by those institutions, and thereby also considering the effect of the resolutions of those problems on the vulnerabilities of poor and isolated individuals and communities.