Category Archives: Chilling effect

The (Hoary) Roots of Vagueness

When, in 2015, the Supreme Court handed down its judgment in Shreya Singhal vs Union of India, it was celebrated for many reasons. One of them was that the Court’s invocation of vagueness – along with the chilling effect – as a ground for striking down Section 66A of the IT Act represented an important conceptual breakthrough. While in Baldeo Prasad (1960) the Supreme Court had struck down a law criminalising “goondas” on the basis that it did not define who a “goonda” was, in K.A. Abbas (1970) the Court had admitted that in certain circumstances vagueness might make a statute void, and in Kartar Singh (1994) the Court had more or less imported the concept of void-for-vagueness from American jurisprudence, Shreya Singhal was believed to be the first time that the Court actually struck down a speech-restricting statute for being unconstitutionally vague.

Apparently not. It turns out that the origin of the vagueness doctrine in the context of speech-restricting statutes is far older than 2015; in fact, it goes back to 1951, to the beginnings of our constitutional jurisprudence, and a good two decades before the classic American judgment on the point, Grayned vs Rockford.

The case is State of Bombay vs F.N. Balsara, which is part of the constitutional canon, although for very different reasons. We know of Balsara because of its upholding of prohibition, its exposition of the doctrine of pith and substance in determining legislative competence, and its restatement of the classification test under Article 14. However, there was also an Article 19(1)(a) issue in Balsara, which the Court dealt with in a terse paragraph towards the end of its judgment. Not only did the Bombay prohibition law ban liquor, it also penalised people – and advertisements – which would “commend, solicit the use of, or offer any intoxicant or hemp” or “incite or encourage any member of the public or any class of individuals or the public generally to commit any act which frustrates or defeats the provisions of this Act, or any rule, regulation or order made thereunder,” [Sections 23 and 24 of the Act]

With respect to these provisions, the Court had this to say:

“Sections 23(a) and 24(1)(a) in so far as they refer to “commending” any intoxicant are said to conflict with the fundamental right guaranteed by article 19 (1) (a) namely, the right to freedom of speech and expression and there can be no doubt that the prohibition against “commending” any intoxicant is a curtailment of the right guaranteed. and it can be supported only if it is saved by clause (2) of article 19 which, as it stands at present, provides that “nothing in sub-clause (a) of clause (1) shall affect the operation of any existing law in so far as it relates to, or prevent the State from making any law relating to, libel, slander, defamation, contempt of court or any matter which offends against decency or morality or which undermines the security of, or tends to overthrow, the State.” It seems to me that none of the conditions mentioned in clause applies to the present case, and therefore the provisions in question must be held to be void. Section 23 (b) must also be held to be void. because the words “incite” and “encourage” are wide enough to include incitement or encouragement by words and speeches and also by acts. The words “which frustrates or defeats the provisions of the Act or any rule, regulation or order made thereunder” are so wide and vague that it is difficult to define or limit their scope. I am therefore in agreement with the view of the High Court that this provision is invalid in its entirety. So far as article 24(1)(b) is concerned the judgment of the High Court in regard to it cannot be upheld. The learned counsel for the petitioner also conceded before us that he was not going to assail this provision.”*

While the major premise of the Court’s argument was that there was no sub-clause under Article 19(2) that could be used to justify prohibiting the “commending” or “encouraging” or “inciting” the sale and consumption of alcohol, it also made it clear that in any event, the gravamen of the offence – “frustrating or defeating the provisions of the act” – consisted of such vague terms, that the provision could not be sustained under Article 19(2). Interestingly, the Court also gave a nod to a concept often associated with vagueness in free speech jurisprudence – that of overbreadth – when it noted that the impugned words “are so wide and vague that it is difficult to define or limit their scope.” The issue of definition speaks to vagueness, while the issue of limit speaks to over-breadth (i.e., when a statute is worded so broadly that it ends up prohibiting speech that may constitutionally be restricted, as well as speech that may not).

Consequently, six decades before Shreya Singhal, a Constitution Bench of the Supreme Court had already struck down a speech-restricting statutory provision on grounds of over-breadth and vagueness. This conceptual advance, however, was largely forgotten in the intervening years (Abbas makes no reference to Balsara), and Balsara is never taught as a free speech case.

It does make you wonder, though, about how many hidden gems are lying scattered about in the Constitution Bench decisions from the 1950s. Chintaman Rao’s exposition of over-breadth was forgotten until Shreya Singhal resurrected it in 2015; In Re Kerala Education Bill’s doctrine of unconstitutional conditions was cited once in a concurring opinion in Ahmedabad St Xavier’s Education Society, but has been submerged ever since; Basheshar Nath’s invitation to develop a theory of fundamental rights as constituting an objective order of values, following German jurisprudence (via the doctrine of waiver), has never been seriously taken up; and of course, Balsara’s account of vagueness has been lost to the canon. All these judgments were delivered by benches of five judges or more, and consequently, remain good law.

How might the jurisprudence of 2017 be affected if the jurisprudence of the 1950s was to be taken seriously once more?

*Notice also that for the Court, the question of finding limitations upon free speech imported from outside 19(2) don’t even arise (much as they didn’t arise in its 1960 judgment in Sakal Papers). This is a good indication of why a two-judge bench’s recent framing of issues pertaining to whether Article 21 can “limit” 19(1)(a) is entirely misguided; the two-judge bench is bound by the Constitution Benches in Balsara as well as in Sakal: restrictions upon free speech are not to be imported form beyond Article 19(1)).


Filed under Chilling effect, Free Speech

Supreme Court to ‘consider’ restricting free speech on grounds of ‘constitutional compassion’

Bar & Bench reports that in light of UP Minister Azam Khan’s statement that a gangrape was a ‘political controversy’, the Supreme Court has framed the following four “questions” for consideration:

“(i)When a victim files an F.I.R. alleging rape, gang rape or murder or such other heinous offences against another person or group of persons, whether any individual holding a public office or a person in authority or in-charge of governance, should be allowed to comment on the crime stating that “it is an outcome of political controversy”, more so, when as an individual, he has nothing to do with the offences in question?

(b) Should the “State”, the protector of citizens and responsible for law and order situation, allow these comments as they have the effect potentiality to create a distrust in the mind of the victim as regards the fair investigation and, in a way, the entire system?

(c) Whether the statements do come within the ambit and sweep of freedom of speech and expression or exceed the boundary that is not permissible?

(d) Whether such comments (which are not meant for self protection) defeat the concept of constitutional compassion and also conception of constitutional sensitivity?”

It has now become almost tedious to repeat that under Article 19(2) of the Constitution, the freedom of speech can only be restricted by a “law”; so where exactly the Supreme Court receives the power to censor speech of its own accord is unclear.

That said, the four “questions” are paradoxically broad and specific at the same time. Is the Supreme Court going to pass an order prohibiting the usage of the words “it is an outcome of political controversy” every time a “heinous offence” occurs? Or is it going to pass an order prohibiting comments that create a distrust in the mind of the victim regarding “the entire system“?

Secondly, what on earth is ‘constitutional compassion’, and where is it found under the eight sub-clauses of Article 19(2)? And yet, perhaps, the question is a little silly. There was no stricter obscenity threshold for ‘historically respectable personalities’ – but the Court invented it in Devidas Tuljapurkar. ‘Constitutional fraternity’ was never a ground for restricting speech under Article 19(2) – but the Court invented it in Subramanian Swamy to uphold criminal defamation. No doubt, in its wisdom, the Court can invent a ‘constitutional compassion’ standard as well.

What that does (and keeps doing) to the sanctity of the constitutional text and structure is another matter.


Filed under Chilling effect, Free Speech

Why the Supreme Court’s Criminal Defamation Judgment is Per Incuriam

In common law, a judgment that is per incuriam has no legal force or validity, and does not count as precedent. Classically, per incuriam is a very narrow concept, applicable only in two circumstances: to a judgment that is passed in ignorance of a relevant statutory provision, or without considering binding precedent of a coordinate or larger bench. In 2015, however, in Sundeep Kumar Bafna v State of Maharashtra, the Supreme Court expanded the definition of per incuriam, noting that:

“It cannot be over-emphasised that the discipline demanded by a precedent or the disqualification or diminution of a decision on the application of the per incuriam rule is of great importance, since without it, certainty of law, consistency of rulings and comity of Courts would become a costly casualty. A decision or judgment can be per incuriam any provision in a statute, rule or regulation, which was not brought to the notice of the Court. A decision or judgment can also be per incuriam if it is not possible to reconcile its ratio with that of a previously pronounced judgment of a Co-equal or Larger Bench; or if the decision of a High Court is not in consonance with the views of this Court. It must immediately be clarified that the per incuriam rule is strictly and correctly applicable to the ratio decidendi and not to obiter dicta. It is often encountered in High Courts that two or more mutually irreconcilable decisions of the Supreme Court are cited at the Bar. We think that the inviolable recourse is to apply the earliest view as the succeeding ones would fall in the category of per incuriam.”

The expansion of the doctrine of per incuriam in the Indian context makes sense, for at least two reasons. The first is the proliferation of two-judge benches of the Supreme Court pronouncing on important constitutional issues, and invariably at odds with each other. Recent examples include Koushal and NALSADevidas Tuljapurkar and Aveek Sarkar, and (as I shall explain) Subramaniam Swamy (the present judgment) and R. Rajagopal. As the Supreme Court correctly notes in Bafna, “certainty of law” and “consistency of rulings” is a very important principle (one of the basic constituents of the rule of law), and given the sheer volume of cases being decided by two-judge benches, it must be applied with greater rigour.

The second reason is that at present, judgments tend to reproduce submissions of counsel at great length, including cases cited by counsel. The reasoning of the judgment, however, fails to engage with those judgments in any meaningful way, apart from – once again – reciting them while describing the law on the point. A good example is Rajbala vs State of Haryana, where the Justice Chelameswar held that the arbitrariness doctrine under Article 14 could not be used to strike down legislation. This was contrary to the binding decision in Mardia Chemicals. That case was not considered while arriving at the above conclusion, but is nonetheless found in the judgment, in a footnote while reproducing counsel’s submissions. A strict application of per incuriam in such a situation would defeat the purpose of the doctrine itself, since what it stipulates is that a Court is bound to consider precedent (although, having once considered it, it can choose to read it in a way that is obviously incorrect – in such a situation, the latter decision will then become binding law.

It is my submission that Subramaniam Swamy vs Union of India is per incuriam on both the narrow and the broad conceptions of the doctrine. It is per incuriam because of its failure to consider the ratio of R. Rajagopal vs State of Tamil Nadu, that is, the finding that a regime of no-fault liability in defamation causes a chilling effect upon free speech.

Subramaniam Swamy vs Union of India mentions Rajagopal at exactly one point in the 268-page long judgment: at page 5 of the judgment (paragraph 2), where it cites the passage in Rajagopal stating that:

“In all this discussion, we may clarify, we have not gone into the impact of Article 19(1)(a) read with clause (2) thereof on Sections 499 and 500 of the Indian Penal Code. That may have to await a proper case.”

Let us now come to Subramaniam Swamy’s consideration of the chilling effect. The Court holds:

“Once we have held that reputation of an individual is a basic element of Article 21 of the Constitution and balancing of fundamental rights is a constitutional necessity and further the legislature in its wisdom has kept the penal provision alive, it is extremely difficult to subscribe to the view that criminal defamation has a chilling effect on the freedom of speech and expression.”

Previously, I have argued that this paragraph makes some fundamental conceptual errors in its understanding of what the chilling effect is. Be that as it may, let us now consider R. Rajagopal’s Case

In R. Rajagopal’s Case, the Supreme Court modified the common law of civil defamation. It replaced the no-fault liability standard with either a New York Times v Sullivan standard of actual malice, or the English standard of reasonable publication (I say “either” because the judgment seemingly adopts both tests, which are actually very different). Crucially, it cited New York Times v Sullivan, and cited that exact paragraph where Sullivan held that no-fault liability causes a chilling effect upon free speech. It then made the following finding:

“… in the case of public officials, it is obvious, right to privacy, or for that matter, the remedy of action for damages is simply not available with respect to their acts and conduct relevant to the discharge of their official duties. This is so even where the publication is based upon facts and statements which are not true, unless the official establishes that the publication was made (by the defendant) with reckless disregard for truth. In such a case, it would be enough for the defendant (member of the press or media) to prove that he acted after a reasonable verification of the facts; it is not necessary for him to prove that what he has written is true.”

In other words, Rajagopal holds that in case of statements defaming public officials, the standard is the (Sullivan) standard of reckless disregard for the truth. It’s reason for doing so is the chilling effect caused by a no-fault liability standard (and in fact, that can be the only reason for doing so, since that was at the heart of Sullivan). Consequently, the proposition that a no-fault liability standard causes a chilling effect in case of defamation is part of the ratio decidendi of Rajagopal. As recently as 2015, in M/s Fibre Boards v CIT, the Supreme Court clarified that the ratio of a case includes the reasons that form part of its conclusion (paragraph 27)

It may be argued at this stage that Rajagopal was a civil defamation case, and the question of what to do about criminal defamation was left open by that bench. That is true; however, that a no-fault liability provision for defamation causes a chilling effect upon speech is agnostic about whether the liability might take civil or criminal form. The reasoning given in Sullivan was that an individual who will be required to take the stand and prove the truth of his statement in order to avoid liability, will tend to self-censor, and “steer clear of the unlawful zone”, thus impoverishing public discourse. In other words, the chilling effect is caused by the legal standard (no-fault liability), and not by the nature of the liability (chilling effect).

For this reason, Subramaniam Swamy’s finding that criminal defamation’s no fault liability does not chill speech is reached in ignorance of Rajagopal, and is therefore per incuriam.

Even if this argument fails – i.e., even if one believes that since Rajagopal was about civil defamation, it’s ratio cannot be extended to cover cases of criminal defamation – Subramaniam Swamy remains per incuriam under the broader definition of Sundeep Kumar Bafna. This is a much simpler case to prove. After Swamy, the legal standard for imposing criminal liability (no-fault) is lower than civil liability (Sullivan) in case of public officials. This overturns the legal system on its head, and leads to an illogical position as long as both decisions continue to hold the field. In terms of Bafna, it is impossible to “reconcile the ratio” of Subramaniam Swamy with R. Rajagopal.

I submit, therefore, that Subramaniam Swamy vs Union of India is per incuriam, and therefore a judgment that has no legal validity and does not serve as precedent for future cases. It is to be hoped that on an appropriate occasion, a future bench of the Supreme Court will consider these arguments while deciding whether or not to refer another constitutional challenge to criminal defamation for resolution by a bench of five judges (as Article 145 of the Constitution requires).


Filed under Chilling effect, Defamation, Free Speech, Per Incuriam

“Historically respectable personalities”: The Supreme Court invents a new exception to free speech

When the Supreme Court struck down S. 66A of the IT Act in Shreya Singhal vs Union of India two months ago, the verdict was hailed as a landmark milestone in Indian free speech jurisprudence, for two reasons. First, for once, the Supreme Court followed up its platitudes about the value of free speech in a democracy not with a sentence beginning with “but...”, but rather with actual concrete action – striking down a speech-restricting provision. And secondly, it was a judgement in which the Supreme Court began with the text of Article 19(1)(a) and 19(2), proceeded to rigorously examine that text in light of constitutional history, purpose and precedent, and analysed the impugned law on the touchstone of the Constitution, as understood and interpreted over the years.

Today, in Devidas Ramachandra Tuljapurkar vs State of Maharashtra, the Supreme Court reversed the slight gains made by Shreya Singhal on both counts, creating a new exception to 19(1)(a) when it comes to putatively obscene speech about “historically respectable figures”. As I shall presently explain, today’s judgement is entirely misconceived, for the following reasons: first, it misstates and misconstrues fifty years of Indian obscenity law jurisprudence; secondly, it invents a “historically respectable figures” exception with no basis in constitutional text, history, precedent or reason; thirdly, in failing to define this new standard, it opens a Pandora’s Box of future claims brought on behalf of “historically respectable figures”, which will cast a deep and broad chilling effect upon core political speech; and fourthly, in creating an artificial hierarchy between “historically respectable figures” and the rest, with differential standards applying to both, it destroys the very basis of having a fundamental right to free speech in the first place – that is, to enable citizens to contest the most entrenched and basic norms that society subscribes to.

In Devidas Ramachandra Tuljapurkar, the question before the Court was asked to quash charges of obscenity leveled against a poet for writing a poem called “Gandhi Mala Bhetala” (‘I Met Gandhi’), published in 1994 and meant for private circulation among the employees of the All India Bank Association. A member of the “Patit Pawan Sangathan” lodged a complaint with the Police against the author and publisher, alleging offences under Sections 153A and 153B of the IPC (causing disharmony between classes) and 292 (obscenity). The Magistrate discharged the accused under 153A and 153B, but not under 292. After unsuccessfully moving the Additional Sessions Judge and the High Court to quash the charge, the accused finally reached the Supreme Court.

After setting out the factual matrix, the Court spent the next thirty-seven pages extracting excerpts from various judgements handed down by American, English and European Courts (paragraphs 12 to 33); thefifty-four pages extracting excerpts from various Indian judgements on obscenity (paragraphs 34 to 60); and then sixteen pages extracting excerpts from the Constituent Assembly Debates and various Indian judgements in praise of free speech (paragraphs 61 to 72). Since the Court made no attempt at a synthesis, however, it is essential to set out, in brief, the development of Indian obscenity law over the last fifty years.

In its 1965 judgement in Ranjit Udeshi vs State of Maharashtra, the Supreme Court had adopted the Victorian-era English test in R vs Hicklin: material was obscene if it tended to deprave or corrupt the people into whose hands it was likely to fall. There were three problems with the Hicklin test as it stood: the broadness of the terms “deprave or corrupt”; its acontextuality (prosecutions under the Hicklin test would focus on pointing out specific words or phrases that were supposedly obscene); and its focus on looking at the text from the point of view of the people who, in the opinion of the Court, were most likely to be corrupted or depraved – something that Madhavi Divan has called “the most vulnerable constituency test”. The second of these issues was resolved by the Court in Udeshi itself, when it clarified that obscenity must be tested by looking at the work of art as a whole, and in context. Subsequently, over the next fifty years, the other prongs of Hicklin were gradually liberalised. In cases over the 80s and 90s, and ending with D-G Doordarshan vs Anand Patwardhan, and Ajay Goswami, the Supreme Court shifted from the “most vulnerable constituency test” to that of the average, reasonable and strong-minded reader; and in Aveek Sarkar, decided last year, it finally expressly abandoned Hicklin after a half-century of chipping away, and replacing it with the 1957 American test in Roth vs US. The Roth test holds that:

“The standard for judging obscenity, adequate to withstand the charge of constitutional infirmity, is whether, to the average person, applying contemporary community standards, the dominant theme of the material, taken as a whole, appeals to prurient interest.”

The Roth test, therefore, is sharper and narrower than Hicklin. It does not cover all material that might corrupt or degrade, but limits itself to whether the dominant theme of the material, taken as a whole, appeals to the prurient interest (and whether it is utterly without redeeming social value). Admittedly, Roth is a problematic test, and as I’ve pointed out before, Aveek Sarkar ought to have considered the developments within American obscenity jurisprudence as well, in future cases, before relying upon a test that has been significantly displaced in its own country of origin. Nonetheless, as a test for obscenity, it is at least more speech-protective than Hicklin – because, for instance, merely (allegedly) vulgar or distasteful work, that might also allegedly deprave or corrupt individuals, would not be caught by the obscenity net.

This survey is essential to understand why the Court proceeds to go entirely off the rails in Tuljapurkar. The test for obscenity, as developed in Roth and adopted in Aveek Sarkar, does not contemplate differential standards for different people. It proceeds on a simple basis: does the work, taken as a whole, to the average reader, appeal to the prurient interest? Presumably, under the old Hicklin test, it would be somewhat possible to make this distinction by (somehow) arguing that disrespecting historically respectable figures would corrupt all those people who follow them as models for the ideal life. I say presumably, because this argument is a stretch, but it is at least conceivable. Under Roth and Sarkar, however, it is the nature of the work that matters, not not who it is about.

The Court gets around this rather basic problem by engaging in what can only be called a sleight of hand. In paragraph 73, it holds that “the test evolved by this Court, which holds the field today is the ‘contemporary community standards test’.”

This, however, entirely misstates the Roth test by omitting what follows after community standards. A complete statement of the test is:

“…applying contemporary community standards, [whether] the dominant theme of the material, taken as a whole, appeals to prurient interest.”

It is not open to judges to pick one part of a judicially evolved test, established through years of gradual and incremental precedent, and omit another part that is inconvenient to the conclusion that they wish to reach. If a judge wishes to depart from a test, than it is incumbent upon him to state the old test, provide his new version, and give reasons for doing so. This is how the Court went about replacing Hicklin with Roth in Aveek Sarkar, and how a Court is normally supposed to act.

By simply ignoring the second part of the Roth test, the Court is able to invent a new exception to free speech: obscenity standards are different when it comes to “historically respectable figures”, because of community standards. To prove the established “community standards” in case of Gandhi, the judgement spends the next twelve pages extracting various Supreme Court judgements that had referred to Gandhian thought and principles, and cites the names of various books about Gandhi. What role these citations are meant to play in the explication of an obscenity test is unclear, but at the end of it, Justice Dipak Misra observes that:

“The issue in the instant case, whether in the name of artistic freedom or critical thinking or generating the idea of creativity, a poet or a writer can put into the said voice or image such language, which may be obscene.”

This, however, is deeply confusing, because the Court begs the question: first, it says that historically respectable figures command a different threshold of obscenity; then, without explaining what that threshold is, it says that the question is whether “in the name of artistic freedom… a poet… may be obscene.” But there is absolutely no quarrel with that proposition! Everyone accepts that obscenity is a constitutional restriction upon the freedom of speech in India. The question is what constitutes obscenity.

Confusion is worse confounded in its concluding paragraph, where the Court says:

“When the name of Mahatma Gandhi is alluded or used as a symbol, speaking or using obscene words, the concept of “degree” comes in. To elaborate, the “contemporary community standards test” becomes applicable with more vigour, in a greater degree and in an accentuated manner. What can otherwise pass of the contemporary community standards test for use of the same language, it would not be so, if the name of Mahatma Gandhi is used as a symbol or allusion or surrealistic voice to put words or to show him doing such acts which are obscene.”

Quite apart from the fact that the last sentence is viciously circular again, the question remains: how do you apply the “community standards” test as a matter of degree? What does it mean to say that the test applies with “greater vigour“? Prima facie, juxtaposing “community standards” with questions of degree seems entirely incoherent. Does it mean that the burden of proof shifts? Does it mean that the evidentiary standards for establishing community standards are lower (not that the Court has bothered with evidence of community standards in its obscenity judgements before)? The Court does not provide any guidance upon these issues. And by refusing to do so, it opens the floodgates for a whole host of future 292 claims in which organisations, claiming to speak on “behalf” of historically respectable figures (another term that the Court refuses to define), will make of S. 292 what has already been made of S. 295A and 153A. The chilling effect of Ss. 295A and 153A will not be compounded because of a conceptually incoherent and incorrect judgement has made vague observations about strict “community standards” for “historically respectable personalities”, but has made no attempt to clarify the meanings of those vague and open-ended terms.

Lastly, the philosophical implications of this judgement are deeply troubling. In walling off a class of individuals – “historically respectable figures”, the Court does two things. First, it arrogates to itself the power to define “respectability”. On what basis must we all accept Gandhi’s “respectability”? By now, we have reams of literature that impugn Gandhi’s personal and political motivations, his actions and his character, and his impact upon the freedom movement. Through this judicial fiction of Gandhi’s “respectability”, the Court creates a judicial narrative which it then imposes upon the rest of the country. This is not the judiciary’s province. And secondly – and relatedly – by creating differential standards about what can and cannot be said based upon the respectability of historical personalities, the Court betrays the whole purpose of a free speech right: the freedom to question, criticise, and undermine the most deeply entrenched narratives that society holds dear, the freedom to use speech to attempt to uproot and transform the most fundamental social and moral norms of a community (as held in Khushboo vs Kanniamal), the freedom to offend, shock or disturb. A right to free speech is nothing without the right to attack and undermine one of the most deeply crystallised narratives of modern India – the narrative of the “father of the nation”.

The vagueness and conceptual slippages rife throughout this judgement leave it unclear how much of an impact it will have doctrinally. For the reasons adduced above, it is submitted that this judgement is incorrect, productive of great public mischief, and ought to be overruled the first chance the Supreme Court gets.

Lastly, the following observation is necessary: it is one thing to decide a case on the basis of one’s moral or political instinct, and then reverse-engineer reasoning to make the conclusion fit with text and precedent. That is an unavoidable facet of the judicial task, and demonstrates at least a degree of respect for constitutional fidelity. It is quite another thing to decide a case, and then bowdlerize text and precedent by ignoring inconvenient parts that don’t suit your conclusion. In the Make-Up Artists judgement, Justice Dipak Misra did this with regard to the Trade Union Act, citing one half of a statute and ignoring the other half that entirely undermined his argument. Today, he has done it again, citing one half of an existing test, and ignoring the other that would preclude his conclusion. The results – to repeat – make for good headlines or bad headlines, depending upon one’s predilections, but in the long run, they cause serious damage to constitutional fidelity and constitutional doctrine.


Filed under Chilling effect, Free Speech, Obscenity

Two Arguments against the Constitutionality of S. 66A

(This post first appeared on the CIS blog, here)

In the immediate aftermath of the elections, free speech issues have come to the fore again. In Goa, a Facebook user was summoned for a post warning a second holocaust if Modi was elected to power. In Karnataka, a MBA student was likewise arrested for circulating an MMS that showed Modi’s face morphed onto a corpse, with the slogan “Abki baar antim sanskaar”. These arrests have reopened the debate about the constitutional validity of Section 66A of the IT Act, which is the legal provision governing online speech in India. Section 66A criminalises, among other things, the sending of information that is “grossly offensive or menacing in character” or causes “annoyance or inconvenience”. The two instances cited above raise – not for the first time – the concern that when it comes to implementation, Section 66A is unworkable to the point of being unconstitutional.

Like all legal provisions, Section 66A must comply with the fundamental rights chapter of the Indian Constitution. Article 19(1)(a) guarantees the freedom of speech and expression, and Article 19(2) permits reasonable restrictions in the interests of – inter alia– “public order, decency or morality”. Presumably, the only way in which Section 66A can be justified is by showing that it falls within the category of “public order” or of “morality”. The precedent of the Supreme Court, however, has interpreted Article 19(2) in far narrower terms than the ones that Section 66A uses. The Court has held that “public order” may only be invoked if there is a direct and immediate relation between the offending speech and a public order disturbance – such as, for instance, a speaker making an incendiary speech to an excited mob, advocating imminent violence (the Court has colloquially stated the requirement to be a “spark in a powder keg”). Similarly, while the Court has never precisely defined what “morality” – for the purposes of Article 19(2) – means, the term has been invoked where (arguably) pornographic materials are concerned – and never simply because speech has “offended” or “menaced” someone. Indeed, the rhetoric of the Court has consistently rejected the proposition that the government can prohibit individuals from offending one another.

This raises two constitutional problems with Section 66A: the problems of overbreadth and vagueness. Both doctrines have been developed to their fullest in American free speech law, but the underlying principles are universal.

A statute is overbroad when it potentially includes within its prohibitions both speech that it is entitled to prohibit, and speech that it is not. In Gooding v. Wilson, a Georgia statute criminalized the use of “opprobrious words or abusive language”. In defending the statute, the State of Georgia argued that its Courts had read it narrowly, limiting its application to “fighting words” – i.e., words that by their very nature tended to incite an imminent breach of the peace, something that was indisputably within the power of the State to prohibit. The Supreme Court rejected the argument and invalidated the statute. It found that the words “opprobrious” and “abusive” had greater reach than “fighting words”. Thus, since the statute left “wide open the standard of responsibility, so that it [was] easily susceptible to improper application”, the Court struck it down.

A statute is vague when persons of “ordinary intelligence… have no reasonable opportunity to know what is prohibited.” InGrayned v. Rockford, the American Supreme Court noted that a vague law impermissibly delegates basic policy matters to policemen, judges, and juries for resolution on an ad hoc and subjective basis, with the attendant dangers of arbitrary and discriminatory application.” There are, therefore, a number of problems with vague laws: one of the fundamental purposes of law is to allow citizens to plan their affairs with a degree of certainty. Vagueness in legislation prevents that. And equally importantly, vague laws leave a wide scope of implementing power with non-elected bodies, such as the police – leading to the fear of arbitrary application.

While overbreadth and vagueness are problems that affect legislation across the board, they assume a particular urgency when it comes to free speech. This is because, as the American Supreme Court has recognized on a number of occasions, speech regulating statutes must be scrutinized with specific care because of the chilling effect: when speech is penalized, people will – out of fear and caution – exercise self-censorship, and the political discourse will be impoverished. If we accept – as the Indian Courts have – that a primary reason for guaranteeing free expression rights is their indispensability to democracy, then the danger of self-censorship is one that we should be particularly solicitous of. Hence, when speech-regulating statutes do proscribe expression, they must be clear and narrowly drawn, in order to avoid the chilling effect. As the American Supreme Court euphemistically framed it, “free speech needs breathing space to survive.” Overbroad and vague speech-restricting statutes are particularly pernicious in denying it that breathing space.

There seems to be little doubt that Section 66A is both overbroad and vague. However ill-judged a holocaust comparison or a morphed corpse-image may be, neither of them are like sparks in a powder keg, which will lead to an immediate breach in public order – or “immoral” in the way of explicit pornography. We can therefore see, clearly, that the implementation of the law leaves almost unbounded scope to officials such as the police, provides room for unconstitutional interpretations, and is so vaguely framed that it is almost impossible to know, in advance, what actions fall within the rule, and which ones are not covered by it. If there is such a thing as over-breadth and vagueness par excellence, then Section 66A is surely it!

At various times in its history, the Supreme Court has acknowledged the problems of overbreadth, vagueness and the chilling effect, but never directly incorporated them into Indian law. As we have seen, each of these elements is connected to the other: over-broad and vague speech-regulating statutes are problematic because of the chilling effect. Since Section 66A is presently being challenged before the Supreme Court, there is a great opportunity for the Court both to get rid of this unconstitutional law, as well as strengthen the foundations of our free speech jurisprudence.

Leave a comment

Filed under Chilling effect, Free Speech, Internet Freedom, IT Act

Guest Post: Free Speech, Interim Injunctions and Media Guidelines

(In this guest post, Raag Yadava, a 2013 NLSIU graduate, analyses the High Court’s recent order restricting reporting of the sexual harassment allegations leveled by an intern against a former justice of the Supreme Court)

The Delhi High Court, in an order dated 16th January, 2014, granted Justice Swatanter Kumar (“SK”) an interim injunction in a suit for defamation (permanent injunction plus damages) filed against various media houses, the unnamed intern who made the allegations of sexual harassment and the reporter who sourced the information. (Note: the defendants were not permitted to file replies to the injunction application.)

The facts, briefly, were this: on 30th November, the intern sent an affidavit to the Chief Justice of India complaining of sexual harassment by SK. (It is worth noting that SK’s stance is that this lady ‘was neither an intern nominated by the Supreme Court nor by the plaintiff himself.’ – see, paragraph 5 of the order). Discovering this complaint, from 10th January onwards, various media houses published (on TV and in print) news of this allegation. Crucially, these reports named SK and appeared prominently (as headlines more often than not). These are some examples: “Another intern alleges sexual harassment by another SC judge”, “Justice S. Kumar … put his right arm around me, kissed me on my left shoulder … I was shocked”, “Sex taint on another former SC judge”, “Ex-judge Claims Green Plot in Sex Slur” etc.

The Court granted the injunction. In doing so, the Court considered the issue to be one of balancing the right of freedom of speech under Article 19(1)(a), on one hand, and administration of justice (or the individual’s right to “open justice”), on the other. Narrating the development of the law on prior restraints on press publication in Naresh Shridhar Mirajkar (1967), Reliance Petrochemicals (1988) and Sahara (2012), the Court begins with the observation that prior restraints are per se not unconstitutional, the devil rather lying in the details in which the restraint is crafted.

Thus next, at various instances, the Court considers the legal factors that permit exceptions to the otherwise unregulated norm of freedom of speech and expression of the press. Drawing on Reliance, at paragraph 38, and on Sahara and Mirajkar in paragraph 43, the Court considers this factors to be “an interference with the administration of justice”. Thus, “if … there exists a real and imminent danger that the continuance of the publication would result in interference with the administration of justice,” a prior restraint to the freedom under Article 19(1)(a) is justified. The Court also backs this standard by the recognition of the drafters of the Constitution, given that “contempt of court” – which includes the power to punish for obstructing the administration of justice – was included as an express restriction to the freedom under Article 19(2).

Then comes the question of what precisely ‘interference in the administration of justice’ means – a question previously considered by the Supreme Court. As in Sahara, the Court here notes the “obstruction of the justice … include(s) intrusion in right to have open justice unbiased by any public opinion (from a) publication which would give excessive adverse publicity to the accused … which may likely hamper the fair trial in future.” Thus, even if some amount of fairness can be attached to the publication, or where it appears to be fair, the Court’s reasoning implies that prior restraints may still be imposed.

Till this point, the Court’s reasoning is consistent with the approach of the Supreme Court previously. In such cases, the Supreme Court proceeds to examine on facts whether such “excessive adverse publicity” exists, and whether it “may likely hamper” fair trial. The Court in this case, however, adds another principle to this legal reasoning – that such prejudice “exists in the cases of persons who are seen with the eyes of public confidence and public faith like judges of the Supreme Court or the other superior Courts of justice.” Given that any aspersions cast on a judge reflect on the judiciary, the individual’s integrity and reputation is linked to that of the institution. Thus, for those holding public office, such allegations cast doubt on the “institution as a whole”. Facially, it does not cohere that an individual’s or for that matter, an institution’s interest in maintaining its reputation also justifies a prior restraint, so the Court links this to loss in public confidence to the right to open justice in these terms: “The person who is accused of such allegations is seen with extreme suspicion and the same also creates a kind of pressure of adverse public opinion which may affect his likelihood of getting fair trial or may lead to interference in the course of the justice.” There are two issues with this reasoning: the Court does not explain why “such” allegations would ipso facto affect a trial court, or what “such” allegations are. If by this, the Court means allegations against high ranking public officials, it would imply that since all allegations against public officials affect their institution’s integrity, a prior restraint is justified in principle without the need for any further examination on facts.

Let us parse this: if I allege that A Raja is corrupt and has allocated spectrum improperly, surely this leads the public to reflect on the Ministry of Telecommunication, and “such” allegations would justify a prior restraint. The Court’s answer – when it does ‘consider’ limited facts – comes in paragraph 53, noting that “the allegations made in the complaint have neither been examined or tested in any Court of law nor have they been proved … not any cogent evidence has been produced along with the complaint.” This, however, is unappealing. Either the Court means that no allegations can be published unless proved in a Court of law – which is a patently absurd conclusion. Or perhaps the Court means that to justify a prior restraint in an ad interim injunction, the Court hearing the matter would itself enter into the facts to see whether there is any element of truth on the basis of evidence. This is problematic not only because sexual harassment complaints (being by their very nature private acts) rarely if ever have “cogent (external) evidence”, but also because this would mean the press’ otherwise untrammelled and unfettered freedom of speech to publish daily and contemporary news would be subject to a half-baked appreciation of evidence by the Court, until the matter is finally heard and judgment pronounced several months later.

Curiously, the Court proceeds to abandon this line of reasoning altogether, concerning itself now with SK’s “impeccable reputation” as a judge who “has dealt with many important cases and has always protected and preserved the interests of justice.” Given this, the Court then records these observations, which are best quoted in full:

56. Assuming for the sake of example that a false complaint is filed against the retired judge of high judiciary after his death by raising similar nature of allegations after the retirement of about 10 or 20 years. One would fail to understand that after his death who would protect his interest and defend the case in Court of law when he had in his career given landmark judgments and had a great name and reputation in bar and bench. These questions are to be examined by the Court when the fresh cases are considered 57. In view of the recent stringent provisions incorporated in the Sexual Harassment of Women at Workplace (Prevention, Prohibition and Redressal) Act, 2013, which provides for a mechanism of dealing with the cases of sexual harassment, this Court is of the opinion that strict view would have to be applied equally to both the sides, i.e. complainant as well as alleged accused specially in cases where the complaint is filed after the lapse of long period. Thus, this Court is also of the view that there should be a limitation of time for the purpose of filing of such complaints, otherwise no one would know when the complaint ought to have been filed and decided. Thus, a balanced approach has to be taken, particularly, in these types of matters. 58. In the present case, assuming the complaint filed by the defendant No.5 is found to be false after inquiry, then who would ultimately compensate and return the repute and sufferings of the plaintiff and mental torture caused to him and his family members.” (emphasis supplied)

Thus, the link between an injury to reputation and the right to open justice, crucial to satisfy the tests in Reliance etc, is forgotten by the Court, speaking solely of how harm to reputation is in itself the relevant factor. Equally, the Court’s remarks on how there “should” be a limitation on sexual harassment complaints seems entirely unwarranted in the facts of this case, and seems to bear no legal relevance to whether the administration of justice may be subverted.

Combing its reasoning till here, the Court finally concludes: (a) given there was only a “stray” allegation (the Court, here, has judged the veracity of the allegation, it seems, on the ground that there was only intern who alleged sexual harassment and not more), (b) there was a ‘delay’ of two and a half years in filing the complaint, (c) the allegations are being excessively published, (d) that such publicity is “destructive of (the individual’s and the institution’s) reputation”, grants the injunction. In line with the general trend, relying on ESN Software Pvt. Ltd., the Court extends the injunction to all non-party media outlets as well.

The injunction itself is found in paragraph 64 – the operative part of the order. It prohibits publication of any material highlighting the “allegation in the form of headlines, without disclosing in the headlines … that they are mere allegations”, and the publication of any photos of SK. The injunction thus curiously seems to fall short of what the reasoning requires. Paragraph 64 only prohibits allegations “in the form of headlines”, thus leaving free publication in less conspicuous parts of the publication, and even then, permits reporting provided that the fact that these are only allegations is made clear. It is unclear how such reporting would also not – if we were to accept the Court’s reasoning – harm SK’s and the Supreme Court’s reputation.

Two crucial questions thus appear: Is it correct for the Court to permit a public official to piggy-back on the institution’s reputation? Is the emphasis on harm to reputation, absent a factual finding of prejudice to a fair trial, consistent with Mirajkar and Reliance? The answer to these questions determines how broad truly the freedom of our press under Article 19(1)(a) is – whether the press’ reporting of facts bona fide against those in high places is constitutionally permissible; whether the “free and healthy press (that) is indispensible to the functioning of a true democracy” is truly indispensible; whether this freedom exists only on paper, to stop the newspapers that matter.

Leave a comment

Filed under Chilling effect, Free Speech, Media Guidelines

The Chilling Effect in India

In our discussion of surveillance on this blog, we have argued that although bulk metadata surveillance does not track and store the content of communication, it may nonetheless violate the freedom of expression and freedom of association by having a chilling effect. Briefly, the chilling effect – a doctrine that is used regularly in the United States, Canada, South Africa and the ECHR, to name just four jurisdictions – applies to cases where governmental laws and governmental (or private) activities are of a nature that – while not directly censoring free speech – nonetheless have the impact of self-censorship. The classic example is that of excessively vaguely worded libel laws – in order to keep on the right side of the law, citizens will end up refraining from engaging in completely legal and legitimate forms of speech. In the United States, we observed how disclosure of the member lists of organizations advocating (at the time) unpopular causes such as the NAACP was held to violate the First Amendment rights of association and expression because there was a probable effect that if individuals knew that a hostile government had access to their membership of such organizations, they would think twice before joining them. We noticed similar judgments in the context of revealing trade union membership, and so on. And we argued that S. 66A of the IT Act, which criminalized, inter alia, speech that is “annoying” is a classic case of a law that would exercise a deeply chilling effect on free expression.

In this context, it is surprising to note that the doctrine of the chilling effect has virtually no foothold in India. The phrase is mentioned in seventeen Supreme Court opinions, primarily in the context of how rent control legislation has a chilling effect on housing (this mischief seems to have been caused by a 1984 opinion, and Krishna Iyer J.’s perhaps somewhat overzealous penchant for evocative prose)! A few opinions refer to the chilling effect on government functionaries if all their action is subjected to public scrutiny – which, as well, is a different context. The Bombay High Court has two opinions, both of which are on rent control. The Calcutta High Court has four, two of which relate – again – to public officials, and two of which seem to have taken the phrase a little too literally, using it in the context of noise pollution caused by firecrackers! It is in one judgment of the Madras High Court that we find it used in a free speech context – that is, accompanying a citation of New York Times v Sullivan, the classic US case on free press and libel laws (the other Madras High Court judgment is on rent control). The Allahabad, Punjab & Haryana and Andhra Pradesh High Courts have a combined six judgments on rent control. In a solitary Karnataka High Court judgment, the petitioners make a 19(1)(a) argument in the context of over-zealous contempt of court laws, although the Court does not refer to it in its judgment (it does seem to accept the argument – or a variant of it – in substance) (the other two Karnataka High Court cases are about the conduct of public functionaries). As for the rest of the High Courts, there is a deafening silence.

The one exception is the Delhi High Court. In four judgments, all within the last ten years, the Court has directly recognized the impact of the chilling effect on the freedom of expression. Indeed, in Ram Jethmalani v Subramaniam Swamy, the Court not only cited the Sullivan doctrine, but read the Supreme Court case of R. Rajgopal to have incorporated it into Indian law! The Delhi High Court observed: “if a person is under fear of being sued, he may not express himself freely on public issues, and this would chill the public debate.” M.K. Sharma v Sangeeta Guptis a case with a complex factual matrix, but in essence, the Court held that statements made to an investigating committee set up under the aegis of the Vishaka guidelines on sexual harassment cannot be subject to a defamation suit, since this would “inevitably have a chilling effect thus deterring an aggrieved person from availing of a remedy provided by the Supreme Court against sexual harassment.” In 2009, in Petronet v Indian Petro Groupthe Court again cited Sullivan and Rajgopal, observing that the problem that both the US Supreme Court and the UK House of Lords were trying to address what could be “damaging self-censorship by the media to the impoverishment of political discourse.” Similar were the observations in Nirmaljit Singh v Yashwant Singh

Thus, all the Delhi High Court cases refer to libel laws, two of them being about the media, and two of them about individuals. While the Delhi High Court cases represent a start, it must be noted that the chilling effect does not only operate where there is a direct threat of legal prosecution – if that was so, NAACP and the trade union cases would have been decided the other way. The doctrine is predicated upon the understanding that coercion can take place in various extra-legal ways: through reprisals in the workplace for putatively unrelated reasons, for instance. The basic point at issue is that if one is aware that the government (or other institutions that possess authority or power over one, or even society at large) know the details of the relationships one is entering into, one will think twice before associating with entities/persons that are deeply unpopular (and yet, perfectly legal).

Is there any discernible reason why this doctrine has not taken root in Indian free speech and free association law? Even our Southern neighbor, Sri Lanka, has recognized this: in Channa Pieris v Attorney Generalresponding to a case of governmental arrest and detention, of members of a movement called Ratawesi Perumana, the Court observed – in words that are directly on point, and which deserve to be stated in full:

There was in the matters before us no direct call to desist from expressive activities… The fact that the respondents took no direct action to restrict the right of the petitioners and members of the Ratawesi Peramuna to associate freely in orderly group activity however, does not end the matter… Freedoms such as these are protected not only against obvious and heavy handed frontal attack, but also from being smothered or stifled or chilled by more subtle interference. We need to consider the probable deterrent effect of the arrest and detention even though such effect may have been unintended. 

The Court then quotes Justice Harlan’s opinion in NAACP that we have discussed before, and goes on to hold:

“In the matters before us, the arrest and detention of the petitioners must be regarded as entailing the likelihood of a restraint upon the exercise by them of their right to freedom of association in much the same way as manifestations of hostility to their activities by the armed persons who stole their posters at Matara. The arrest and detention was in my opinion likely to adversely affect the ability of the petitioners to pursue their collective effort to foster beliefs which they admittedly have the right to advocate, in that it may induce some of them to withdraw from the Peramuna and dissuade others from joining it because of fear of police action. It must also have certainly had a chilling effect on the expressive and associational activities of those who had the temerity to continue to be members of the Ratawesi Peramuna.”

And this, surely, must be correct. A doctrine that finds firm constitutional ground in countries having constitutional cultures of free speech and association as diverse as the US and Canada, the ECHR and South Africa and Sri Lanka, is surely near-univeral in its scope. If it hasn’t yet found footing in India, perhaps that is simply because it hasn’t yet been litigated. The Delhi High Court cases, and the scattered observations from Madras and Karnataka do indicate, however, that the argument is there to be made; and what better context to make it in than that of bulk surveillance?

Leave a comment

Filed under Chilling effect, Free Speech, Surveillance