1. Context

How do judges decide cases? How should they decide them? What does how they decide them say about the nature of law? These vexed questions have been revisited through the various feminist judgments projects in recent years, building upon other movements and theoretical schools over the past century in particular.Footnote ¹ Of the predecessor approaches, feminist legal scholarship is most commonly associated with legal realism and critical legal studies. For the realists and the Crits, assessment of judgments in cases reaching appellate courts betrays law's character as the contingent outcome of innumerable ideological battles at the heart of the community, where those battles have been won by rich or otherwise advantaged people.Footnote ² Judges are thought to reach their conclusions based on instinctive preferences formed at the outset; they pitch their analyses at levels that enable them to draw on rules, and on interpretations of those rules, that facilitate those conclusions. Law is politics, but with a deceitfully respectable cloak.

The realist and critical theses are troubling insofar as they might be thought to expose the ways in which the ‘liberal legal order’ (so-called) rests on fanciful notions of neutrality and non-arbitrariness.Footnote ³ They conflict with formalist conceptions of law with their ‘reverence for the role of logic and mathematics and a priori reasoning’.Footnote ⁴ Crits and feminists argue that such conceptions – despite their having fallen out of favour among specialists – continue to seep into the minds of law students and often prevail in societal understandings of law and legal process.Footnote ⁵ Textbooks in contemporary legal modules introduce big liberal ideas that are widely and sometimes thoughtlessly proclaimed by lawyers: reasonableness; equality before the law; the rule of law, and so on. Law students are quickly made familiar with statute law; with different sources of statutes and the hierarchies between them; with a half dozen canons of statutory interpretation.Footnote ⁶ They are introduced to the distinction between the ratio decidendi and obiter dictum: judges, they are told, are bound by precedent; they apply the laws that others make.Footnote ⁷ In this climate, lawyers could be forgiven for adopting a more reassuring conception of law as a vast but organised body of rules, mechanically applied in case after case.

This rules-oriented conception resonates with what the feminist scholar Carol Gilligan described in her 1982 opus In a Different Voice as the ‘ethics of justice’ mode of moral reasoning.Footnote ⁸ Prompted by a concern to scrutinise the methodology employed by psychologist Lawrence Kohlberg in work that had claimed that men were capable of reaching higher levels of moral development than women, Gilligan's research led her to identify this as one of two modes of moral reasoning. Whereas the other – the ‘ethics of care’ mode – is more context-oriented and holistic, the ‘ethics of justice’ mode is mechanical, with the agent reasoning in deductive ways from abstract principles or rules, and seeking definitive conclusions. Finding it more prevalent among boys, Gilligan associates it with the 11-year-old ‘Jake’. The 11-year-old ‘Amy’ meanwhile – in keeping with her emphasis on context – is inclined towards elaborating the relevance of relationships, and is less concerned with conclusiveness.

Even prior to the wave of feminist judgments projects – and despite concerns around its supposed essentialising of men and women – feminist legal scholars had begun to draw on Gilligan's work in their critiques of liberal legalism, arguing that legal structures and discourse are informed by Jake-type modes of reasoning.Footnote ⁹ Cases take a rigidly adversarial form: evidence is either admissible or inadmissible; litigants either win or lose. The argument seemed particularly applicable at the site of adjudication where, feminists argued, Amy-like reasoning is dismissed as intuitive and unreliable. As we shall see in the section that follows, Erika Rackley made a particularly effective version of the argument with reference to myth and fairy tale. Like Hans Christian Andersen's little mermaid, the woman judge must trade her voice for acceptance in the liberal legal world: she must ‘shed her difference and fit the fairy tale’ – where the fairy holds that judges in the liberal order set aside their own politics, experiences, and even selves, on appointment to the bench.Footnote ¹⁰

It subsequently informed the approaches taken, and the goals identified, in much of the feminist judgments literature. In an introductory chapter to the UK Feminist Judgments volume, Rosemary Hunter identifies ‘including women… in terms of writing women's experiences into legal discourse…’ as among her ‘seven criteria for judging like a feminist’.Footnote ¹¹ She refers to feminist judges’ disposition to ‘reason from context and the reality of women's lived experience’ and to ‘make individualised rather than categorical or abstract decisions…’Footnote ¹² In an equivalent chapter in the Northern/Irish Feminist Judgments, the co-editors suggest that feminist judges might ‘ask “the woman question” and focus on the gendered effects of apparently “neutral” liberal legal constructs and practices, and/or challenge gender bias in legal doctrine and judicial reasoning’.Footnote ¹³

In the wake of this wave of scholarship, it is timely to reflect on its insights and implications. Should it be understood as a foundational challenge to the liberal legal order: as concerned not only with exposing its myths, but – following Carol Smart's call to arms in Feminism and the Power of Law – with prompting those persuaded by its merits to ‘avoid the siren call of law’ altogether, disengaging from and even actively railing against what is deemed an innately patriarchal discourse?Footnote ¹⁴ Or has it more modest aims such as generally integrating feminist insights into conventional legal discourse? This paper takes the approach of revisiting of Rackley's depiction of Dworkin's Hercules with the aim of shedding light on these questions, and on feminist judging more generally. Do feminist judges, contra Hercules, impose ideology from beyond law's black letter rules, concsiously dissenting from liberal legal process (and, by extension, from the basic foundations of the liberal legal order)? Or might it be that feminist judges, like Hercules, retrieve better, richer meaning from within those black letter rules, thereby promoting the non-dominating potential or even essence of those rules (and, by extension, helping realise the essential justness of liberal legal ideals)? We begin the inquiry with a look at Dworkin's theory of adjudication, and at Rackley's critque.

2. Hercules

Dworkin's account of adjudication might be understood as having emerged with an eye on demonstrating the essential incoherence of two schools of thought, and as having then developed by way of response to the incoherence of a third.Footnote ¹⁵ First, law-as-rules formalists were wrong to ignore law's inescapably interpretive nature. Judges could not be thought of as robotic rule-appliers. Secondly, legal positivists were wrong to overlook the ways in which moral values were an inescapable part of the process of adjudication; judges couldn't but draw on moral ideals to make sense of the rules that they were tasked with applying. Then, in response to the third category of thinkers: the legal realists and critical legal scholars were wrong to think that judges simply worked from their own preferences; that they were not bound by the rules at least in certain and important ways. In Taking Rights Seriously, Dworkin illustrates this with reference to Riggs v Palmer.Footnote ¹⁶ Here, the judges were faced with the beneficiary of a will who had murdered the testator, his misfortunate grandfather, apparently in order to expedite his inheritence. The difficulty was that the black letter rules on testamentary succession did not include any provision that blocked such a beneficiary from taking the inheritence. The criminal rules of murder applied, as the grandson found out to his cost, but, on the face of the ‘hard data’ of legal doctrine – the legislation, the case law and the black letter rules emanating from other such recognised sources – the will was not obviously invalid.Footnote ¹⁷ Yet the majority on the court found against the murderer on the succession question.

Was it just that the judges went beyond the law to find and apply what was the right moral answer, filling in what a legal positivist might see as a ‘gap’ in the law? No, Dworkin insisted.Footnote ¹⁸ They pursued what was their inescapably interpretive task, discovering principles implicit in the legal system that explained and made sense of the black letter rules that applied to the case. Various (and often competing) principles were at play, including one around the idea that ‘no person shall profit from his own wrong’, and another around the idea that ‘the enactments of the legislature should be enforced according to their clear wording’. The judges could not consult particular legal sources in order to identify the deeper principles, but neither could they make them up from thin air. Rather, judges in this case – as in all ‘hard cases’ – identify principles that best fit with and justify the black letter legal rules. As he presents it in Law's Empire, judges decide such cases by ‘trying to find, in some coherent set of principles about people's rights and duties, the best constructive interpretation of the legal doctrine of their community’.Footnote ¹⁹ They engage in an analysis and justification of legal doctrine in a particular domain and reconcile it with a broader analysis and justification of the entire legal system. In Riggs, this interpretive enterprise led them to clarify that the statutory rules on succession were subject to an implicit exception precluding a murderer from inheriting under his victim's will.

As for Judge Hercules, he is not to be thought of as separate from this theory of adjudication.Footnote ²⁰ Rather, he is a device intended by Dworkin to shed light on it. Drawing on his ‘superhuman skill, learning, patience and acumen’,Footnote ²¹ Hercules does in a complete and self-conscious way what mortal judges do ‘instinctively’ and ‘in a much less methodical way’.Footnote ²² Whereas real-world judges reason ‘from the inside-out’ or ‘from more specific problems to broader and more abstract ones’ (that is, from the relevant legal rules out towards broader justifications of those rules), Hercules reasons ‘from the outside-in’.Footnote ²³ Before he sits on his first case, he builds ‘a gigantic “overarching” theory for all seasons’. He can establish an understanding of ‘what justice and fairness require; what freedom of speech, best understood, means; whether and why it is a freedom particularly worth protecting…’ and so on. He can ‘weave all that and everything else into a marvelously architectonic system’ such that when a case comes to be decided he is ‘very well prepared’.Footnote ²⁴ He can then readily identify principles that ‘fit with’ and ‘justify’ – and that therefore explain or make sense of – the relevant legal rules. In each case he decides, therefore, he effectively constructs a moral and political theory for the whole legal and political order that casts the black letter rules in their best moral light.

Dworkin's theory, just like those of the realists and Crits, has troubling implications. In keeping with the ‘Hercules’ epithet, it presents the judge as a powerful agent, shifting law in new directions in accordance (or partly in accordance) with the moral and political theory that she finds most persuasive. Yet it seems to chime in certain ways with how judges in higher courts decide cases: the apparently legal disagreements that split courts can often seem to reflect deeper disagreement at the level of political theory. We must not take an unduly narrow view of what represents philosophical reasoning here: as Stephen Guest argues, the fact that real world judges do not consciously engage in argument at the level of political theory, or express themselves in the vernacular of academic philosophers, should not in itself be taken to count against Dworkin's theory of adjudication.Footnote ²⁵

Dworkin insisted, of course, that his theory could hardly be further from those of the realists and Crits.Footnote ²⁶ He presented law as an integrated and coherently evolving body of rules and principles rather than as the contingent outcome of intense ideological battles. Adjudication is not a political enterprise governed by the intuitive preferences of judges. Rather, judges write their judgments as co-authors of a chain novel might write their particular chapter.Footnote ²⁷ Just as a would-be author of the next season of The Crown cannot write the royal family out and present a cartoon about a foolish cat and a winning mouse, judges must bring through interpretations of legal rules that those rules bear. They retrieve meaning from within the body of rules and principles that comprise the law, rather than impose their own preferred meaning from beyond law's empire – as we consider again in Part 4.

Among Dworkin's feminist critics, Erika Rackley homes in on Hercules, in a series of articles published as the feminist judgments volumes were beginning to emerge.Footnote ²⁸ Although conceding that he takes account of the discretion that judges undeniably have, Rackley sees Dworkin's theory as ‘ultimately a staunch attempt to defend the notion of judicial constraint’, and Hercules as a ‘more jazzed-up version’ of the formalist or positivist accounts in which judges are conceived as robotic rule-appliers.Footnote ²⁹ She suggests that ideas around Hercules's ‘detached neutrality’ and ‘detached impartiality’ are pervasive in contemporary liberal legal culture. Yes, everyone knows that real-world judges cannot and do not, in Rackley's phrase, ‘dispense justice [like Hercules] from the coolness of Mount Olympus’.Footnote ³⁰ But that does not mean that the myth does not have a ‘normative pull’, or ‘operative effects’, in the real world of law.Footnote ³¹ It penetrates the imaginations of lawyers and judges. They strive to live up to an ideal that is both impossible and damaging. It is impossible because judges cannot but bring their own humanity and experiences to bear in their adjudication. It is damaging in that, insofar as it encourages judges to rein in their own humanity and experiences and to hide the ways in which they actually do bring them to bear, it diminishes their adjudicative capacity.Footnote ³² They fail to account adequately for the inescapable connections, as Robin West would have it, between care, empathy and justice.Footnote ³³ Relatedly – and recalling Gilligan's ideas around Amy-like reasoning – it means that women judges (perhaps feminist judges?) are encouraged to suppress their own voices and selves on appointment to the bench. In Rackley's words:

Rackley goes beyond this to suggest Dworkin's use of Hercules as ‘strategic’: that he consciously ‘harnesses the power of myth to capture and subsequently constrain the legal imagination’.Footnote ³⁵ As she sees it, Hercules has been programmed by Dworkin to generate outcomes ‘consistent with the values and premises of the particular political tradition he is there to serve and preserve’.Footnote ³⁶ He is thus a lead player in the patriarchal liberal tradition, designed to marginalise and exclude. No doubt the particular super-hero he chose – or the fact that he chose a super-hero at all – invites such a reading of Dworkin and Hercules. However, while illuminating, Rackley's account is ultimately problematic, as we consider in later parts of this paper. First, we must turn from theory to practice.

3. Stokes

Stokes v CBS High School Clonmel involved a claim that the admissions policy of Clonmel CBS (an all-boys post-primary school in County Tipperary, Ireland) breached the Equal Status Acts 2000–2008, legislation that had been introduced by the Irish legislature to transpose various aspects of the body of European equality law emerging in the wake of the Amsterdam Treaty.Footnote ³⁷ In circumstances where there were 174 applicants for 140 places, the claimant, John Stokes – a member of the Irish Traveller community – had missed out following a procedure governed by the school's admissions policy. The claim was basically that a criterion in the school's admission policy that gave preference to applicants whose fathers were past pupils of the school indirectly discriminated against John Stokes, as a member of the Irish Traveller community, in the context of historical discrimination against that group and related patterns of low education and school attendance.

Stokes hinges on two questions. The first concerned the meaning of the legal ‘rule’ that was of most immediate relevance. This was a rule in the Equal Status Acts addressing indirect discrimination. It was concerned with instances where ‘an apparently neutral provision’ placed a ‘protected group’ at a ‘particular disadvantage’.Footnote ³⁸ That the Irish Traveller community was a ‘protected group’ was not in issue – the community was specifically designated as such by the legislation. That the school's admissions policy was an ‘apparently neutral provision’ was also not in issue insofar as it did not directly target members of the Irish Traveller community. Hence this question centred on the appropriate interpretation of the phrase ‘particular disadvantage’.

The second concerned what, specifically, might count as the allegedly discriminating provision in the case. Although the admissions policy in general was in issue, it was the ‘father rule’ that was troubling from the perspective of the protected group. The evidence – undisputed in the case – was that less than 100 Irish Trallevers attended post-primary schools in Ireland in the period where the applicant's father was of school-going age.Footnote ³⁹ Because of their prevalance in schools across the country, such ‘parent rules’ would tend to be problematic for Travellers in general. The complication was that under the particular school's admissions policy the ‘father rule’ ran in tandem with a ‘sibling rule’; that is, the two sub-rules operated on an either/or basis. This latter rule provided that applicants whose siblings attended the school would qualify automatically regardless of whether they satisfied the ‘father rule’. Very often, if not usually, schools would not be over-subscribed, and so all applicants would be successful in that particular year. In any subsequent year where the school might be oversubscribed, younger siblings of Traveller pupils/past pupils would thus automatically qualify. In practical terms, this sibling sub-rule offset the extent to which the overall rule impacted on Travellers. They were much more likely to satisfy the sibling than the father rule, and if they satisfied it, the father rule did not appear to impose any burden on them.

The two questions interlap, of course, and analysis of the latter helps illustrate the slipperiness of the overall concept of indirect discrimination. On one reading – insofar as the relevant part of the admissions policy allows qualification on the basis of either the father or the sibling rule – the question of whether the policy actually puts members of the Irish Traveller community at a particular disadvantage depends on how difficult it is for applicants from that community to qualify through either sub-rule by comparison with non-Traveller applicants. Such an approach might be thought practical and rational: it looks at the text in the relevant legal rule (‘particular disadvantage’); looks then at the combination of sub-rules in the admissions policy through which applicants can qualify (father rule/sibling rule), and compares the chances of Traveller and non-Traveller applicants in a mechanical way. Alternatively, it is specifically the father rule that is problematic in light of the overall idea or context of structural discrimination. That is, if indirect discrimination is concerned with the notion that the access of any individual to a social good like a place in the local public school should not be governed by historical patterns of discrimination, any assessment of indirect discrimination should apply to the father rule on its own, in isolation from the sibling rule.

In addressing these questions in his judgment for the Supreme Court, Mr Justice Frank Clarke (hereafter, Clarke J; as it happens, Clarke has since been appointed chief justice) begins with some contextualisation, invoking the example of a minimum height rule in a job application and its disparate impact on women compared to men. He does this partly to emphasise the group-oriented rationale of indirect discrimination, and partly in order to distance himself from the earlier approach taken by the High Court judge, Mr Justice Partick McCarthy, to the meaning of ‘particular disadvantage’.Footnote ⁴⁰ McCarthy J had consulted the Oxford English Dictionary on the question, finding that ‘particular meant ‘restricted to… belonging only to a specified person or thing’.Footnote ⁴¹ On that basis he (McCarthy J) held that non-Traveller children of parents who had not attended the school were no better off than Traveller-children in the same boat (ie it did not apply only to them) and that there was therefore no violation. McCarthy J thus drained the indirect discrimination norm of its meaning through insipid subservience to a key word in the provision. He basically conflated it with direct discrimination, despite its emergence as a distinct norm, concerned at countering a distinct ill. (So much for the idea that judges can robotically apply legal rules.)

From there Clarke J appears to shift towards what looks like a methodical, rational approach. Conisder the following extracts from his judgment relating to ‘particular disadvantage’:

• • ‘In order for a protected category to be said to be at a disadvantage in comparison to an alternative category… then it seems to me that it is necessary to attempt to analyse the effect of the measure on both of those categories respectively. Such an exercise necessarily carries with it some degree of statistical analysis…’ (emphasis added);Footnote ⁴²

• • ‘… I am satisfied that the use of the term “particular” brings with it a requirement, as a matter of law, that it must be established that the extent of any disadvantage is significant or appreciable…’ (emphasis added);Footnote ⁴³

• • ‘… But it is important to emphasise that statistical analysis is notoriously capable of giving rise to inaccurate indications when very small numbers are involved… The important point to emphasise is that a proper analysis of disadvantage will normally require sufficient numbers to make that analysis meaningful… Looking at the position of a single individual does not really add very much to the analysis of disadvantage…’ (emphasis added).Footnote ⁴⁴

The emphasis is on statistics: comparable groups; measurable difference; sufficient quantities of data to be representative etc. In the same vein, he insists upon the provision of sufficiently localised data too: ‘it does not necessarily follow that a measure applicable to an educational institution in one area of the country necessarily has the same effect as an identical rule in another part of the country’.Footnote ⁴⁵ Indeed, he brings the statistical methodology to its logical conclusion by insisting that a proper analysis would have added into the formula the fact that Traveller applicants who did not automatically qualify, just like non-Traveller applicants, went into a lottery system for any remaining available places (as did John Stokes, although he was unsuccessful in the lottery). Their chances there would have to be combined with their chances under the cumulative father/sibling rule to demonstrate their overall chances compared with a non-Traveller applicant.

Clarke J notes that the decision-makers/interpreters lower down the chain whose approaches he was reviewing (the director of the Equality Tribunal, the Circuit Court judge) had found in favour of the Traveller family on the substantive claim based on ‘figures produced [that] clearly demonstrated that the likelihood of a typical potential secondary school pupil from the Travelling Community nationwide having a parent who attended any secondary school was very substantially less than that applicable to a non-Traveller’.Footnote ⁴⁶ In other words, they had done so because they were satisfied that the evidence that less than 100 Irish Travellers had attended post-primary schools anywhere in Ireland during the period that John Stokes’ parents would have been of school-going age was sufficient to demonstrate that a Traveller child of John Stokes’ age would be indirectly disadvantaged by a parental-preference criterion in school admissions. He finds the approach problematic: ‘extrapolating from those figures without further analysis, that such a situation applied in the context of Clonmel High School, was an inappropriate inference’.Footnote ⁴⁷

Before turning to the feminist judgment, let us consider Clarke J's approach through generous lens. Even a cursory glance at his public statements indicates that he cares about social justice, and there can be no doubt but that he appreciates Traveller disadvantage and the importance of equal access to education.Footnote ⁴⁸ Were he a legislator, it is safe to assume that he would support policy that promoted Traveller access to education. His task in Stokes, however, is not to develop policy; it is to interpret and apply legislation in the context of a particular set of facts. That means establishing what might count as ‘particular disadvantage’ in the circumstances of the admissions policy of a school in Clonmel. Surely that requires rigorous engagement with statistics, including as they might apply to the particular catchment area of the school whose admissions policy is the subject of the litigation?

What then of Olivia Smith's (hereafter, Smith J) feminist judgment? Consider first of all her assessment of the ‘particular disadvantage’ question. In order to ‘provide necessary context in determining the correct approaches to proof of indirect discrimination’, she sets out developments in the relevant area at the level of EU law since the concept of indirect discrimination first emerged in the case law of the European Court of Justice in the 1980s: she references the various directives, and the related amendments to the Equal Status Acts in Irish law.Footnote ⁴⁹ She points out that these amendments emerged in response to ‘well-noted difficulties associated with proof of a prima face case of indirect discrimination’, specifically those arising in contexts where, for historical and other reasons, states had not collected statistical data over relevant periods.Footnote ⁵⁰ The Race and other EU Equality Directives ‘welcomed a simplification in statutory language away from a consideration of the proportions of members of the respective groups disadvantaged or advantaged by the provision, along with a recognition that indirect discrimination may be demonstrated otherwise than by statistical evidence’.Footnote ⁵¹ The fact that the amended section 3A of the Equal Status Acts states that ‘in any proceedings statistics are admissible for the purpose of determining whether discrimination has occurred’ should, she argues, be read in light of all this background. In other words, it is no longer the case that statistical evidence would always be mandatory in such claims.Footnote ⁵²

Smith J then introduces an ‘accepted social facts’ thesis which, as she elaborates with reference to case law, has been taken by certain expert tribunals as sufficient in some instances to make out a case of indirect discrimination. The approach ‘would not permit as acceptable forms of evidence deriving from knowledge of social circumstances that are predicated on unsubstantiated personal opinion or belief’. Rather, it would refer to facts whose social acceptance is grounded in ‘available and continuingly relevant research findings, both qualitative and quantitative that the tribunal or deciding court finds relevant, credible and continuing’.Footnote ⁵³ By way of illustration Smith J cites the widely accepted fact that the majority of lone parents are women, or that the majority of residents of halting sites are Travellers – as per Clarke J's own example that the average height of women is shorter than that of men.Footnote ⁵⁴ The idea is that accepting ‘accepted social facts’ of this kind can obviate the need for comprehensive statistical analysis in individual cases, where such analysis might be difficult or even impossible to present. In the Stokes case, the fact that less than 100 Traveller children attended Irish post-primary schools in the relevant time was enough, Smith J argued, to make the case that the father-rule indirectly discriminated against Traveller children of John Stokes’ age. There was no need to present such comprehensive data of the kind insisted upon by Clarke J, not least in light of the fact that such data was unavailable.

As for the feminist judge's approach to what represented the ‘apparently neutral provision’, Clarke J had preferred the reading described earlier as practical and rational: that ‘at the level of principle, where a number of alternative means are provided for complying with a qualifying measure, and where it is only necessary to qualify under one heading, the “provision” must mean the totality of the alternative measures available’.Footnote ⁵⁵ Smith J preferred the alternative approach (see above), emphasising the point or rationale of the norm of indirect discrimination – which she saw as concerned with counteracting networks of privilege that operate as barriers to certain public goods – along with broader contextual considerations.Footnote ⁵⁶

Thus, drawing on a kind of republican ‘domination-without-interference’ idea, she refers to the potentially ‘dissuasive impact that facially neutral yet indirectly discriminatory provisions may have in terms of screening out otherwise qualified candidates or applicants’. Here she may seem to have the broader precedential implications of her judgment in mind. That is, if a Traveller applicant had a sibling attending Clonmel CBS, that applicant would surely not be dissuaded from applying in virtue of the existence of the indirectly discriminatory father rule: he would go ahead and apply regardless, presumably with confidence that he would succeed. But this ‘screening out’ idea recalls the broader rationale of the indirect discrimination norm, and it points Smith J, as an interpreter of the particular rule, towards an interpretation whereby the ‘offensive’ sub-rule is to be distinguished from any alternative sub-rule that may in practical terms offest its discriminating effect. Smith J notably complements this principle-based reasoning with reliance on black letter legal doctrine. She points out, for instance, that the relevant provision of the Race Equality Directive refers to indirect discrimination as arising ‘where an apparently neutral provision, criterion or practice’ would put persons of a racial or ethnic origin at a particular disadvantage compared with other persons. Thus the EU law that the Equal Status Acts were transposing in fact used wider terms than the Irish term of ‘provision’.Footnote ⁵⁷

4. Analysis

If we never had the benefit of the feminist judgment, we might be inclined to ruefully accept Clarke J's judgment. Yes, as a matter of social justice, there is something wrong with the ‘father rule’. But such wrongs should be remedied in a democratically legitimate way, and so in the political rather than in the legal domain. The applicants just did not manage to demonstrate ‘particular disadvantage’, and so Clarke J's hands were tied. The feminist judgment seems to cast things in a different light, particularly when considered alongside broader insights from the feminist legal scholarship. Clarke J's judgment might instead be thought to summon the images of Gillgan's Jake and Rackley's depiction of the super-hero judge who is taken to have somehow managed to shake off his own self on ascending to the bench. His emphasis on statistics, including his insistence on sufficiently large and localised samples, is initially suggestive of conclusiveness: if there is particular disadvantage, it must be demonstrated rather than vaguely felt or experienced. But he is light on what might be described as the black letter data. Leaving aside his analysis of a complex jurisdictional element in the case (as I have done in this paper) he cites only one precedent – and at that, only in passing – in 21 pages of analysis.Footnote ⁵⁸ He makes no reference at all to the background EU legislation or case law.Footnote ⁵⁹ He relies instead on his own considered sense of how particular disadvantage might best be demonstrated. Has he really cast off his own self and experience here, or might his educational background in maths and economics have informed his approach?Footnote ⁶⁰

Before offering an alternative reading of the two judgments, let us consider a further elaboration from Dworkin's work, again relating to adjucation, but this time more specifically on the notion of interpretation.Footnote ⁶¹ In Law's Empire, Dworkin reflects on interpretation of the social practice of courtesy in an imaginary community. We are asked to imagine this community in which at one point in time the many rules of courtesy includes one that roughly says that ‘peasants should take off their hats to nobility’. The participants in this social practice will adopt what Dworkin calls an ‘interpretive attitude’, which attitude has two components. First, everyone will share the assumption that the practice of courtesy does not just exist but that it has some purpose; some value ‘that can be stated independently of just describing the rules that make up the practice’.Footnote ⁶² Secondly, everyone will assume that the requirements of courtesy – the actual behaviour it calls for and rules it comprises – are ‘not necessarily or exclusively what they have always been taken to be but are instead sensitive to its point [ie its value or its purpose] so that its strict rules must be understood or applied or extended or modified or qualified or limited by that point’. Once the two components of this interpretive attitude are in play – even if probably in a subconscious way among participants in the practice – then interpretation decides ‘not only why courtesy exists but also what, properly understood, it now requires’. As Dworkin puts it, ‘value and content have become entangled’.Footnote ⁶³

These ideas are presented as part of Dworkin's broader thesis around dismantling legal positivism. That is, there is no separating the ‘is’ and the ‘ought’, fact and value. Law inescapably involves interpretation, and interpretation is inescapably value-oriented. But it also points to the way in which the social practice necessarily evolves over time and how that evolution comes about not out of any sense of rebellion against, but rather out of commitment to, the practice. It may be that at one point in time, participants in the social practice of courtesy generally tend to understand its purpose to be something around showing respect to social superiors.Footnote ⁶⁴ This generates a certain consensus on the general rules and behaviours that make up ‘courtesy’ while also, inevitably, generating certain disagreements around what specific rules and behaviours might be required in order to honour its overall purpose – or around which versions of those rules and behaviours might honour it best. These rules and practices thus shift, perhaps a little in one period, a lot in another, as participants try to find how best to abide by the point of the social practice. Later – in light of the evolved rules and behaviours – participants in the practice are inclined to query the actual purpose of courtesy in the first place. They might ask if it really is about showing respect to social superiors, or whether it is concerned with some other value or values. They reflect on the question in large part in light of the actual behaviour and the rules that it has come to comprise in that period, ie the evolved behaviour and rules. And so on.

For the present discussion, it is critical to appreciate that the interpreting agent is not at large to make of the practice what she will. As Dworkin puts it:

The image emerging is one in which raw behavioural data – what in law we might think of as the statutory provisions and relevant precedents – necessarily interact with each interpretor-participant's judgment as to what purposes or values best capture or make sense of those data. Each interpretor-participant digs down into the data in order to identify the values that might cast those data in their best moral light. They cannot invent a value that flies in the face of the black letter rules and legal precedents. But neither can they avoid engaging with value in their reading of the data. Just as a producer of the school musical cannot declare to her audience at the beginning of the show that ‘this is not a mere interpretation of West Side Story… this is West Side Story’, a judge cannot simply call ‘Open Sesame’ and find the unimpeachably true meaning of the phrase ‘particular disadvantage’ revealed to her (echoing Lord Reid's point about adjudication).Footnote ⁶⁶

With this in mind, a more generous reading has Clarke J as a kind of mortal Hercules, engaged in an interpretive enterprise in which he seeks to cast the standing rules around indirect discrimination – as well as those around statutory interpretation, the separation of powers and judicial constraint etc – in their best moral light. Likewise in the case of Smith J, just that she does what could be argued to be a more Herculean job.Footnote ⁶⁷ Notice that she steadfastly stays within the domain of law. At no point in her judgment could she be said to imply: ‘this is an oppressive rule, and I am ignoring it’. Neither can she be said to circumvent an inconvenient rule through excessively creative interpretation. Rather, she mines down into the intricate web of law in order to try to make sense of the conundrums presented by the most immediately relevant rule concerning ‘particular disadvantage’. She identifies all kinds of relevant black letter data – case law, directives, legislative amendments – and combines them with principles that correspond with and make sense of them. She gives them context, pointing out – as Claire Bruton does in her corresponding commentary – that some of the most relevant legislative revisions were aimed precisely at counteracting the problems recognised by the authoritiatve law-makers around the possibility that a statistics-only approach might drain the norm of indirect discrimination of its normative and practical force.Footnote ⁶⁸ To paraphrase Dworkin, Smith J sets out a plausible and morally attractive conception of ‘not only why the norm of indirect discrimination exists but also what, properly understood, it requires in the circumstances of the Stokes case’.

If the feminist judgment might be thought ‘Herculean’ in this way, isn't the reasoning also reminiscent of Gilligan's Amy? Smith J embraces the ‘contextualisation and particularity’ which Rosemary Hunter cites as a central tenet of feminist judging?Footnote ⁶⁹ The opening sections of her judgment are heavily contextual, both in terms of the concept of indirect discrimination (group-oriented, designed to counter the tendency of historical patterns of discrimination to persist) and the experience of the protagonists (indeed even those experiences are contextualised, abstracted away from the individual as an atomistic agent to the individual within the group, in line with the overall context of the main legal concept in the case).Footnote ⁷⁰ Her ‘accepted social facts’ might initially look vague, in that supposedly Amy-like sense. But she justifies her reliance on it through both established doctrine and higher principle. This is not ideology imposed from beyond, with Smith J doing her bit for the higher feminist cause. It is rules-and-context-and-principle retrieved from within the law.