Associative Harm in Ilareva: Clarity, Not Contraction [Guest Post]

(This is a guest post by Tanmay Durani)

Introduction

The commentary by Natalie Alkiviadou on Ilareva and Others v Bulgaria (“Ilareva”) presents the judgment as a moment of doctrinal slippage, an instance where the European Court of Human Rights (“Strasbourg Court”) has blurred the line between death threats and hate speech, and in doing so, stretched Article 14 (prohibition of discrimination) of the European Convention on Human Rights (“Convention”) beyond principled limits. More specifically, the concern is that the Strasbourg Court collapsed the distinction between individualised threats and equality-based harm, thereby expanding the reach of discrimination analysis into a terrain that is traditionally governed by criminal law and Article 8 (right to respect for private life) alone. 

The anxiety is intelligible. When courts expand positive obligations in the domain of speech regulation, the spectre of conceptual overreach is never far away.

How the Case Reached Strasbourg

To appreciate the stakes, however, the factual and procedural context of Ilareva requires closer attention. The applicants — Valeria Ilareva, Lidia Staykova, and Krasimir Kanev — were prominent human rights defenders in Bulgaria working with refugees and migrants at a time of heightened anti-immigrant rhetoric. In January 2015, they became the targets of a barrage of Facebook posts containing explicit death threats, including calls to “hang them at Parliament,” to “shoot them,” and to poison them. A photomontage portraying them as “Freaks of the Year” circulated widely. Their professional identities were publicly known, office addresses easily accessible, and one of the applicants was later physically attacked after appearing on television.

Complaints were lodged with the Bulgarian authorities under two provisions of the Criminal Code: Article 144 (criminal threats) and Article 162 (hate speech). Prosecutors, however, chose to pursue only Article 162 and declined to investigate the more direct charge of threats. Crucially, they then concluded that the applicants could not qualify as “victims” under the hate speech provision because they were not themselves members of a protected minority. The investigation stalled, with authorities characterising the posts as “negative assessments” unlikely to generate justified fear because they were made online. 

In effect, the applicants were excluded from protection both as victims of threats and as victims of hate speech. This failure created a ‘protection vacuum’ where the domestic authorities’ formalistic interpretation of victimhood effectively immunized prejudice-motivated threats from legal scrutiny. It was against this backdrop that the case reached Strasbourg.

Layered Harm in Digital Space

A closer reading of Ilareva, however, suggests not the erosion of doctrinal clarity, but an attempt — albeit imperfect, and perhaps under-theorised — to grapple with the layered nature of harm in contemporary digital environments. Online threats differ from their offline analogues not merely in medium but in scale, permanence, and networked amplification. They circulate rapidly, accumulate visibility, and can generate diffuse but enduring intimidation. And so, to that extent, the Strasbourg Court appears to have recognised that such harm may operate simultaneously at multiple normative levels -

It locates the threats within a matrix of prejudice. The applicants were targeted not randomly, but because of their professional association with refugees and migrants. The hostility was thus directed simultaneously at individuals and at the equal civic standing of the groups they defended. This is the hinge of the disagreement.

The charge of “hate speech by association” assumes that equality-based protection is triggered only when the immediate target belongs to a protected class. Yet anti-discrimination law has long recognised associative harm. One need only recall the reasoning in Beizaras and Levickas v Lithuania, where the Court held that homophobic abuse required investigation under Articles 8 and 14, not solely because of the personal distress caused, but because of its discriminatory character. Or consider the trajectory from Aksu v Turkey to Minasyan and Others v Armenia, where the Court progressively acknowledged that speech reinforcing structural prejudice inflicts harms that exceed individual reputational injury.

Seen in this light, Ilareva does not invent a new category; it extends an existing logic to a context where prejudice operates through proxy. This reflects the doctrine of 'transitive animus,' where the hostility directed at the advocate is not a separate injury but a 'mirrored' manifestation of the structural prejudice aimed at the primary group.

The deeper concern, however, is about hierarchy.

Does the invocation of Article 14 risk suggesting that a death threat becomes more legally significant when prejudice-motivated? That anxiety misconceives the function of equality analysis. Criminal law across many European jurisdictions treats bias motivation as an aggravating factor — not because threats absent bias are trivial, but because prejudice-based violence compounds harm. It intimidates communities, chills participation, and signals exclusion from the civic order. To acknowledge this additional dimension is not to dilute the gravity of threats per se; it is to recognise that harm can be cumulative and relational. Indeed, to exclude activists from equality protection on the ground that they are not themselves members of the vulnerable group would create a perverse asymmetry: solidarity would receive less protection than identity, narrowing Article 14 in a manner inconsistent with its purpose.

Concluding Thoughts

None of this is to deny that Strasbourg’s reasoning would benefit from sharper articulation. The Court’s positive obligations jurisprudence has expanded steadily, and its doctrinal architecture sometimes lags behind its normative intuitions. The justification for regulating hate speech lies not merely in preventing offence, but in protecting public assurance — the guarantee that all members of society can stand as equals without being publicly denigrated or threatened (Waldron, 2012). If that is the animating principle, then threats aimed at silencing those who defend minorities directly undermine that assurance. What remains for the Court, however, is to specify the analytical thresholds that distinguish (i) unprotected threats of violence, (ii) discriminatory hostility that engages Article 14, and (iii) protected but unsettling expression. Greater clarity as to intent, severity, nexus to protected characteristics, and demonstrable impact would go some way toward stabilising the framework.

Ultimately, Ilareva should not be read as collapsing categories, but as recognising their intersection. Death threats are punishable because they endanger personal security; prejudice-motivated threats implicate equality as well. The convergence of Articles 8 and 14 in such circumstances reflects the dual nature of the harm rather than an inflation of doctrine. The challenge for Strasbourg is not to retreat from this recognition, but to render explicit the theory that underpins it. Without such theoretical precision, critics will continue to perceive expansion where the Court may intend doctrinal continuity. With it, the jurisprudence can evolve without surrendering the commitment to freedom of expression that lies at the heart of the Convention system.

Segregation, Misunderstood: In Defence of Clause 7(d) of the 2026 UGC Regulations

Introduction

The Supreme Court (“Court”), in its interim order staying the University Grants Commission (Promotion of Equity in Higher Education Institutions) Regulations, 2026 (“2026 Regulations”), framed five “substantial questions of law”, including - among others - whether the inclusion of the expression “segregation” in Clause 7(d) could amount to a “separate yet equal” classification infringing constitutional guarantees of equality and fraternity. 

Notably, Clause 7(d) was not, at least on the record, a focus of the petitioner’s challenge. The concern appears to have been raised by the Court suo moto.

The Anxiety with “Segregation”

“Segregation”, in its most basic sense, denotes the enforced separation of groups within shared civic, social, or institutional spaces on the basis of race, sex, or religion. With time, it has come to describe a structured system of hierarchy where separation is backed by law, justified by claims of difference, and sustained through material inequality. 

In the United States, the doctrine of “separate but equal” was crystallised in Plessy v. Ferguson, wherein state-mandated racial segregation was upheld on the fiction that formal parity of facilities satisfied the Equal Protection Clause. The result was the Jim Crow regime, which produced racially divided schools, transport, housing, and public accommodations - all nominally equal yet systematically unequal in resources, access, and dignity. Nearly six decades later, when the U.S. Supreme Court reversed course in Brown v. Board of Education, it did so on the insight that enforced separation in public education was inherently unequal because it generated stigma and stamped Black children with a badge of inferiority. 

And so, the Court’s anxiety is, to be fair, understandable. The semantic charge the term carries has accrued through repeated association with systems that converted separation into hierarchy. The question, however, is whether that historical memory should control interpretation when the word appears within a materially different regulatory design.

In Defence of Clause 7(d)

Clause 7(d) of the 2026 Regulations provides:

“Higher Education Institutions shall ensure that any selection, segregation, or allocation for the purpose of hostels, classrooms, mentorship groups, or any other academic purposes is transparent, fair, and non-discriminatory.”

Properly read, Clause 7(d) appears to be neither an authorisation of segregation nor a flirtation with “separate yet equal.” It is a safeguard simpliciter. The provision states that any “selection, segregation, or allocation” for hostels, classrooms, mentorship groups, or other academic purposes must be “transparent, fair, and non-discriminatory.” The operative word is not separation, but discipline in the face of being transparent and without prejudice or bias. The clause recognises that higher education institutions inevitably organise students across space (eg. rooms are allotted, mentorship cohorts assembled, etc) and insists that such organisation must not reproduce prohibited hierarchies.

The inclusion of the word “segregation” does not create a new power. It acknowledges an existing institutional reality wherein students are sorted, clustered, and allocated for multiple reasons, which can sometimes, very well just be logistical. In a caste-stratified society, these processes might not always be neutral. Spatial arrangements can reflect and entrench social divisions even in the absence of explicit intent. Hostel allotments, informal grouping practices, or mentorship allocations may inadvertently track social background. Clause 7(d) brings these practices within the ambit of an equality review. It renders them visible and therefore regulable.

And thus, to conflate this with the logic of “separate but equal” is to overlook the difference between enforced separation as an instrument of domination and regulated allocation as an administrative necessity. In the former, the state mandates classification on prohibited grounds. In the latter, institutions perform ordinary organisational functions, now subject to equality constraints. 

The clause reflects an anti-subordination impulse. It recognises that equality in educational institutions cannot be confined to admissions and examinations; it must extend to how institutional space is structured. If spatial organisation becomes a vehicle for isolation or exclusion, the promise of equal participation is hollowed out. By placing allocation practices under an explicit equality obligation, Clause 7(d) seeks to prevent precisely that outcome.

A Brief Endnote

The validity of Clause 7(d) ultimately turns on the method of judicial approach. If adjudication proceeds by isolating historically charged text from the design and objective of the Regulations, it risks converting memory into reflex. But if it proceeds by examining structure, purpose, and social context, the analysis shall look markedly different. Clause 7(d) neither mandates separation nor legitimises hierarchy – it subjects allocation to equality discipline. That is not a regression into “separate but equal,” but an attempt (and a rather modest one) to ensure that the organisation of space within universities does not replicate entrenched social divisions.

On the (Un)Constitutionality of the CBI

The Central Bureau of Investigation (CBI) occupies a rather peculiar space in our constitutional design. It is neither textually entrenched nor born of a comprehensive post-Constitution statute designed for a federal republic. Its lineage runs through the Delhi Special Police Establishment Act (DSPE Act), and a 1963 executive resolution that formally constituted the CBI. From this somewhat improvised foundation has emerged the country’s premier investigative body. 

The question, then, is not whether the CBI performs important functions - one can be sure, that only a few would deny that - but whether its constitutional basis is commensurate with the breadth of its authority.

The starting point is legislative competence. Article 246 read with the Seventh Schedule distributes power across the Union and the States with careful specificity. “Police” is placed in Entry 2 of the State List. Now, investigation of crime is not peripheral to policing, but it is its core. On a straightforward reading, therefore, the power to investigate offences within a State belongs to that State. The DSPE Act attempts to reconcile this allocation by limiting the CBI’s jurisdiction and requiring State consent under Section 6 before exercising powers within a State’s territory. Consent, thus, functions as the hinge: for it allows a Union agency to enter a State domain. But this solution raises a prior question. If Parliament had plenary authority over criminal investigation, consent would be redundant. If it does not, consent begins to resemble a negotiated exception rather than an organic product of constitutional design.

The federal tension becomes more visible when States withdraw “general consent,” compelling the CBI to seek case-specific permission. This phenomenon illustrates that the agency’s jurisdiction is contingent and revocable. The Supreme Court’s judgement in State of West Bengal v. Committee for Protection of Democratic Rights (para 44 (vii)) adds another layer. There, the Court held that constitutional courts, acting under Articles 32 and 226, may direct a CBI investigation even in the absence of State consent. The reasoning was grounded in the protection of fundamental rights - i.e. where State machinery is compromised, judicial power must step in. The holding is defensible. Yet structurally, it renders the consent requirement porous. 

One might ask whether this move subtly recentralises investigative authority under the aegis of rights adjudication.

Institutional design compounds the anxiety. The CBI was constituted through executive resolution. Its powers are expanded through executive notifications. Its superintendence, except in corruption cases under the oversight of the Central Vigilance Commission, lies with the Union government. In Vineet Narain v. Union of India, the Court attempted to insulate the agency from political interference by mandating fixed tenure and procedural safeguards. The judgement effectively constitutionalised aspects of the CBI’s functioning without conferring upon it constitutional status. And so, what emerges is a body whose independence depends less on textual entrenchment and more on judicially crafted norms and executive self-restraint. But for an agency wielding arrest, search, and prosecutorial authority across jurisdictions, this provenance appears thin. 

Yet the unconstitutionality claim cannot rest solely on structural discomfort. Parliament is competent to legislate for Union territories and for matters concerning Union employees. Inter-State and transnational crimes (corruption, economic offences, organised crime) often exceed provincial capacities. A purely State-centric investigative regime may be normatively faithful to the text but practically inadequate in a complex polity. The existence of a central investigative body is therefore not inherently suspect. The difficulty lies in the elasticity of its growth. The DSPE Act was designed with a narrow mandate; over time, through executive accretion, the CBI has become a general investigative powerhouse. When institutional practice stretches far beyond original statutory contours, constitutional scrutiny becomes unavoidable.

There is also the question of federalism as part of the basic structure. In S.R. Bommai v. Union of India, the Court affirmed federalism as a foundational principle. If investigative authority becomes a tool through which the Union exerts pressure upon States (whether or not such pressure is demonstrably exercised) structural imbalance becomes a legitimate concern. The problem here is not merely misuse but design. A framework that allows concentration of coercive power without robust statutory insulation may be vulnerable to political distortion.

To characterise the CBI as unconstitutional in an absolute sense may be satisfying but bad in law and overbroad. The stronger claim, I argue, is subtler. The basis of CBI sits uneasily with the constitutional scheme. It is neither fully anchored in the distribution of powers nor comprehensively insulated from executive influence. It relies upon consent, judicial innovation, and convention. That combination may function tolerably in ordinary times, but appears fragile in moments of political contestation.

Should Parliament enact a comprehensive statute explicitly reconciling federal competence with central investigative needs? Should investigative autonomy be textually entrenched or institutionally strengthened? These are design questions, not merely adjudicatory ones. To raise them is not to undermine the fight against corruption or complex crime. It is to insist that coercive power (precisely because it is necessary) must rest upon clear foundations. And so, in that sense, debating the constitutionality of the CBI compels us to look beyond functionality and ask whether institutional evolution has kept pace with principle. 

The answer may not lie in declaring the agency void. It may lie in recognising that its present form is a compromise - one that deserves closer, and perhaps more candid, reflection.