In the best piece of free speech news since 1960, the Supreme Court on Tuesday struck down Section 66A of the IT Act, that criminalised “grossly offensive”, “menacing” and “annoying or inconvenient” speech over the internet. There has been a substantial degree of commentary about the judgment over the last two days. In this post, I will attempt a detailed excavation of the many fascinating strands of the judgment, from the point of view of free speech doctrine.
Use of American First Amendment jurisprudence
A month ago, I wrote a rather exasperated post about a bizarre Delhi High Court decision allowing the police to pull down political posters from the walls of private property, on the ground that “Delhites have short fuses”, and that political posters could therefore be a threat to public order. In particular, when American free speech jurisprudence was cited before Justice Endlaw, he refused to consider those cases, on the ground that while the American First Amendment is “absolute” (“Congress shall make no law… abridging… the freedom of speech”), Article 19(2) is subject to reasonable restrictions. This cavalier dismissal highlights the highly opportunistic manner in which the Indian judiciary has engaged with American First Amendment law over the years. On occasions when it helps to buttress a point, judges have shown no compunctions in quoting the grand, eloquent pronouncements of the American judiciary on the freedom of speech; but of course, American First Amendment law has historically been far more speech-protective than its Indian counterpart. Consequently, when judges wishing to uphold far-reaching restrictions upon the freedom of speech are faced with contrary American cases, instead of engaging with the reasoning and demonstrating why it is invalid or inapplicable, they invariable invoke the “First Amendment is an absolute!” trope, and save themselves the trouble of having to consider contrary reasoning.
As I’ve tried to argue before, a refusal to engage with judicially trained minds grappling with very similar issues achieves nothing but stifling exposure to a range of rigorously thought-through and developed viewpoints, and harms the overall quality of reasoning. More importantly, though, the argument that the First Amendment is an absolute, and 19(2) contains reasonable restrictions, is a canard. The First Amendment is not absolute. No American judge, apart from Hugo Black, and possibly William Douglas, has held it to be. The First Amendment permits regulation of incitement to imminent lawless action, obscenity, fighting words, true threats, blackmail, copyright infringement, insider trading, consumer fraud and commercial speech. This was clearly understood by the framers. In the Constituent Assembly Debates, Ambedkar himself observed:
“It is wrong to say that fundamental rights in America are absolute. The difference between the position under the American Constitution and theDraft Constitution is one of form and not of substance. That the fundamental rights in America are not absolute rights is beyond dispute. In support of every exception to the fundamental rights set out in the Draft Constitution one can refer to at least one judgment of the United States Supreme Court.”
He then specifically went on to cite an American judgment on restrictions upon free speech – Gitlow vs New York.
In Paragraphs 14 to 19, Justice Nariman clinically demolishes the aforementioned canard. Citing Chaplinsky vs New Hampshire, the classic American judgment affirming that the First Amendment is not absolute, he correctly points out that the American Supreme Court has never given literal effect to the “make no law” phrase. He then correctly notes that the crucial difference between the American and Indian positions is that while in the United States, a compelling public interest must be demonstrated in order to regulate speech, in India, a restriction must be covered by the eight themes specified in Article 19(2). In other words, there could be occasions when the Indian Constitution protects more speech than the American! In any event, subject to this rider, Justice Nariman notes:
“Viewed from the above perspective, American judgments have great persuasive value on the content of freedom of speech and expression and the tests laid down for its infringement.”
And this is exactly as it should be. There is much to disagree with in American free speech law. But what is undeniable is that over a hundred years, American judges have crafted a deep, thoughtful and complex set of principles for understanding the purposes of the freedom of speech in a constitutional democracy. We might reject their principles completely, but we need to engage with them.
Public interest cannot be a ground for restricting speech
In specifying that under the Constitution, speech can be restricted only under one of the eight listed grounds under Article 19(2), Justice Nariman states twice that “public interest” is not one of the grounds, and so cannot be invoked to justify a speech restriction. In paragraph 21, he notes:
“Under our constitutional scheme, as stated earlier, it is not open to the State to curtail freedom of speech to promote the general public interest.”
Immediately after, he cites Sakal Papers vs Union of India in support of this proposition.
This might sound like an innocuous statement, but it is critically important. This is because, over the years, judges have repeatedly ignored the fact that Article 19(2) exhaustively lists eight grounds of restriction, and that public interest is not one of them (unlike in Article 19(6)). Judges have upheld restrictions upon the nebulous grounds of “public interest” and “social interest”. In Ranjit Udeshi vs State of Maharashtra, for instance, while upholding obscenity law, Justice Hidayatullah traveled beyond the terms of the Constitution to observe that the freedom of speech “is subject to reasonable restrictions which may be thought necessary in the interest of the general public.” He then used that to elide the “decency and morality” clause with “public morality.”
In K.A. Abbas vs Union of India, while upholding a regime of pre-censorship upon cinema, that same Justice Hidayatullah observed:
“… social interest of the people override individual freedom. Whether we regard the state as the paren patriae or as guardian and promoter of general welfare, we have to concede, that these restraints on liberty may be justified by their absolute necessity and clear purpose. Social interests take in not only the interests of the community but also individual interests which cannot be ignored. A balance has therefore to be struck between the rival claims by reconciling them. The larger interests of the community require the formulation of policies and regulations to combat dishonesty, corruption, gambling, vice and other things of immoral tendency and things which affect the security of the State and the preservation of public order and tranquillity.”
In In Re Arundhati Roy, the contempt of court case, the Court held:
“… whether there can be a balancing between the two public interests, the freedom of expression and the dignity of the court.”
Examples may be multiplied, but there is a clear pattern here. The worst free speech judgments of the Supreme Court have come when the Court has traveled beyond its brief, collapsed the specific terms of 19(2) into a boundless and boundlessly manipulable vision of “public” or “social” interest, and then upheld the far-reaching restrictions that the government has sought to impose. There come times in the history of any constitutional court, when jurisprudence becomes so utterly untethered from the constitutional text and principles, that what is most urgently needed is simply a reaffirmation of the basic meaning of text and structure. In doing so, Justice Nariman has performed the incalculably important service of re-orienting free speech jurisprudence back towards its fundamental goals and purposes: interpreting the Constitution of a liberal-democratic polity.
“In the interests of”, “tendency”, and the requirement of proximity
Article 19(2) permits “reasonable restrictions” “in the interests” of the eight prescribed themes. The major contestation in Indian free speech law has turned upon the interpretation of “reasonable” and “in the interests of”. In Ramji Lal Modi and Virendra, two cases decided in 1957, the Supreme Court stated that the phrase “in the interests of” was wider in ambit than the phrase “for the maintenance of”, and consequently, authorised the government to regulate any speech that had a “tendency” towards, for instance, public disorder. In Modi, the Court upheld S. 295A on the ground that intentional outrage to religious feelings had the “calculated tendency” (sic!) to disrupt public order. In Virendra, the Court held prior restraint upon the press under the colonial era Press (Emergency Powers) Act to be constitutional, for the same reason. A plea that there must be proximity between speech and disorder was expressly rejected in Modi.
The word “tendency” is pernicious and damaging. It speaks back to American free speech jurisprudence in the 1920s, when the Supreme Court used a “bad tendency” test to persecute anti-war dissenters, trade-union leaders, leftists and pacifists of various hues. Understanding the sheer unworkability of this test, the Court abandoned it in the 1940s, adopting Justice Holmes and Brandeis’ test of “clear and present” danger. The reason is obvious: “tendency” can mean just about anything, from imminence to the faintest causality. If I start smoking now, I will have a tendency to cancer, even though cancer may set in forty years later. But “tendency” has attained an ubiquitous place in Indian speech restricting laws. The Contempt of Courts Act, for instance, criminalises speech that can have the “tendency” of lowering the repute of the Court; S. 292 of the IPC criminalises material that “tends” to deprave or corrupt. And so on.
Fortunately, we did not have to wait as long as the US to push back against the idea of “tendency”. In Ram Manohar Lohia’s Case (1960), the Supreme Court cleverly “distinguished” precedent, and held that there was a requirement of proximity between speech and the threatened disorder, and that the connection must not be remote, arbitrary or fanciful. In Lohia, a law criminalising instigating people to not pay their taxes was struck down, because it was held not to have a proximate connection to public disorder. Subsequent cases have tightened this test – in S. Rangarajan, the Court held that the connection must be that of a “spark in a powder keg”, and in Arup Bhuyan, that there must be “incitement to imminent lawless action.” But “tendency” has also continued to be invoked by the Courts with alarming regularity. Soon after Lohia, the Court upheld the crime of sedition in Kedar Nath Singh (1962), on the ground that the State could legitimately criminalise speech that had a “tendency” to public disorder.
As in the case of “public interest”, we can immediately see that “tendency” has been responsible – again – for some of the most regressive and speech-restricting decisions of the Indian Supreme Court.
In the 66A judgment, Justice Nariman emphatically adopts the requirement of proximity. He cites Ram Manohar Lohia’s case, highlighting the need for an “intimate connection” between speech and the prohibited 19(2) category.
Incitement vs advocacy: Collapsing “tendency” into imminence
The requirement of an intimate connection is expressed by Justice Nariman in the form of a crucial distinction: between advocacy and incitement. In paragraph 13, he observes:
“Mere discussion or even advocacy of a particular cause howsoever unpopular is at the heart of Article 19(1)(a). It is only when such discussion or advocacy reaches the level of incitement that Article 19(2) kicks in. It is at this stage that a law may be made curtailing the speech or expression that leads inexorably to or tends to cause public disorder…”
The distinction between “advocacy” and “incitement” is grounded in the idea of proximity, or imminence. For instance, my “advocating” a violent revolution against the State by writing articles in magazines, or even delivering public lectures, does not constitute “incitement”. However, my whipping up a mob into a frenzy and directing it to imminent violent action does. The enquiry is contextual, and is clearly limited to emergent, specific situations. In other words, Justice Nariman emphatically rejects the proposition that an idea, or a message, can be criminalised because of its communicative content. It is only when there is a relationship of immediacy between speech and action – “speech brigaded with action”, in the words of Justice Douglas – that the law can kick in. I have argued elsewhere that this view is consistent with our most basic ideas of individual autonomy and responsibility. But what is most interesting here is that Justice Nariman then collapses “tendency” into the incitement requirement. He says that it is at the stage of incitement when the law kicks in, to curtail speech that tends to cause disorder.
In other words, the concept of “tendency” that motivated the Court in Modi, Virendra and Kedar Nath Singh is emphatically discarded here. Justice Nariman holds that even if speech has a “tendency” to disorder (say, for instance, a revolutionary tract), that in itself is not a ground for restriction: there must, further, be incitement. To come back to my smoking example – I have a tendency to cancer when I start smoking, but cancer is “imminent” (or “inexorable”) only at a much, much later stage.
Thus, even while maintaining continuity with precedent, by continuing to use “tendency”, Justice Nariman effectively knocks the bottom out of the entire rationale for upholding 295A and sedition.
In paragraph 36 onwards, he buttresses this by collapsing tendency into the American “clear and present danger” test, even citing Modi and Kedar Nath Singh! In paragraph 41, he concludes the public order enquiry by holding:
“Viewed at either by the standpoint of the clear and present danger test or the tendency to create public disorder, Section 66A would not pass muster as it has no element of any tendency to create public disorder which ought to be an essential ingredient of the offence which it creates.”
There is a small quibble here: in paragraph 37, Justice Nariman observes:
“The test of “clear and present danger” has been used by the U.S. Supreme Court in many varying situations and has been adjusted according to varying fact situations. It appears to have been repeatedly applied, see- Terminiello v. City of Chicago 93 L. Ed. 1131 (1949) at page 1134-1135, Brandenburg v. Ohio 23 L. Ed. 2d 430 (1969)”
Strictly speaking, this is incorrect. The “clear and present danger” test was found to be prone to gross abuse during the McCarthy years, when the Court invoked it to convict communist party members and other dissidents. Far from using “clear and present danger”, Brandenburg vs Ohio was the case that rejected it, in favour of a more stringent “incitement to imminent lawless action” test. The Brandenburg standard was adopted by the Supreme Court in 2011, in Arup Bhuyan’s Case. In that respect, it is something of a pity that Justice Nariman endorses “clear and present danger” instead of Brandenburg. Nonetheless, it is also arguable that his disquisition on the distinction between “advocacy” and “incitement” effectively speaks to a Brandenburg level of protection, going forward.
Meaning of Public Order
The term “public order” is a term of art. In Romesh Thappar, independent India’s first free speech judgment, it was defined as “a state of tranquility which prevails amongst the members of a political society.” In Ram Manohar Lohia vs State of Bihar (a different case that also involved Ram Manohar Lohia), the Court conceptualised three concentric circles: “law and order” being the widest, “public order” being narrower, and “security of the state” being the narrowest. Thus, a disruption of public order is something graver than merely breaking a law, or disrupting “law and order”. In the free speech judgments since Lohia, however, this definition has been largely ignored (the Delhi High Court case cited above is a classic example). As in the case of rejecting “public interest” as a ground of restriction, the Court’s endorsement of Lohia provides a crucial reaffirmation of the fact that constitutional terms – especially terms that limit fundamental rights, cannot be arbitrarily expanded, and the Court must adjudicate constitutionality specifically upon the touchstone of their defined meanings.
On a combination of all these factors – that the terms of S. 66A did not establish a proximate link with public order, as defined, the Court held that Article 19(2) did not save that section, at least insofar as public order was concerned. The same analysis was applied to defamation, and decency and morality – the Court reaffirming its recent judgment in Aveek Sarkar’s case. The government’s attempt to escape unconstitutionality by adding an exhaustive set of guidelines (see para 48) was correctly rejected, on the ground that this was tantamount to rewriting the whole section. Consequently, the Court struck it down.
In Grayned vs Rockford, the American Supreme Court defined a vague statute as one which ensured that persons of “ordinary intelligence… have no reasonable opportunity to know what is prohibited.” In Kartar Singh v. State of Punjab, the Supreme Court – citing this case – observed that “it is the basic principle of legal jurisprudence that an enactment is void for vagueness if its prohibitions are not clearly defined. Vague laws offend several important values… laws should give the person of ordinary intelligence a reasonable opportunity to know what is prohibited, so that he may act accordingly. Vague laws may trap the innocent by not providing fair warning. Such a law impermissibly delegates basic policy matters to policemen and also judges for resolution on an ad hoc and subjective basis, with the attendant dangers of arbitrary and discriminatory application.”
There are, therefore, two problems with vague statutes. One is a rule-of-law problem: citizens do not have a fair chance to plan their affairs. The second is a delegation problem: the executive is given far too much discretion to implement vague laws on the ground. We have seen both issues at play with the abuse of S. 66A over the years.
While in K.A. Abbas, the Supreme Court admitted that vagueness could be a ground for striking down a law, it did not do so (despite the Central Guidelines for film certification, which were at issue, being bizarrely overbroad.) In Baldeo Prasad, the Court struck down a law that criminalised goondas, but did not define who a “goonda” was. But the 66A judgment is the first time – to my knowledge – that the Court has struck down a speech-restricting law on the grounds of vagueness. Crucially, the Court observes that it is not possible for the legislature to cast “a net large enough to catch all possible offenders and leave it to the Court to step in and say who could be rightfully detained and who should be set at liberty.”
After citing a copious degree of American and Indian jurisprudence to establish the principle of vagueness, Justice Nariman correctly observes, in paragraph 69, that “judged by the standards laid down in the aforesaid judgments, it is quite clear that the expressions used in 66A are completely open-ended and undefined.” Further contributing to the vagueness are the absence of mens rea, and a series of terms (such as “obstruction”, “danger” or “annoyance”) which are ingredients of an offence in the Indian Penal Code (that of public nuisance), but have become offences in themselves in the IT Act (para 72). Justice Nariman distinguishes other IPC provisions that use identical terms (such as Ss. 294 and 510) on similar grounds, and ends by noting the sheer subjectivity of the words that constitute offences under the IT Act:
“… every expression used is nebulous in meaning. What may be offensive to one may not be offensive to another. What may cause annoyance or inconvenience to one may not cause annoyance or inconvenience to another. Even the expression “persistently” is completely imprecise – suppose a message is sent thrice, can it be said that it was sent “persistently”?”
Over-breadth and the Chilling Effect
Over-breadth is a concept that is closely connected with – but not identical to – the chilling effect. A statute is over-broad if – in the words of the Indian Supreme Court in Chintaman Rao vs State Madhya Pradesh, “the language employed is wide enough to cover restrictions both within and without the limits of constitutionally permissible legislative action affecting the right. So long as the possibility of its being applied for purposes not sanctioned by the Constitution cannot be ruled out, it must be held to be wholly void.”
Over-breadth directly implicates the reasonableness requirement of Article 19(2). In State of Madras vs V.G. Row, the Supreme Court held that a “reasonable restriction” under Articles 19(2) to (6) would have to satisfy the requirements of proportionality: “the nature of the right alleged to have been infringed, the underlying purpose of the restrictions imposed, the extent and urgency of the evil sought to be remedied thereby, the disproportion of the imposition, the prevailing conditions at the time, should all enter into the judicial verdict.” It is clear that if a statute proscribes conduct that is much broader than what is permitted under Article 19(2), on the ground that there is some – tenuous – connection between the two, there is good reason to argue that the restriction is disproportionate.
While in Chintaman Rao, the Court struck down a statute for being over-broad, over-breadth as a constitutional concept has not yet managed to acquire a foothold in Indian constitutional jurisprudence. Here again, Justice Nariman’s judgment breaks new ground by expressly invoking over-breadth as a ground for striking down a speech-restricting statute. In paragraph 83, he observes:
“Information that may be grossly offensive or which causes annoyance or inconvenience are undefined terms which take into the net a very large amount of protected and innocent speech.“
And, in paragraph 86:
“[66A’s restrictions] fall foul of the repeated injunctions of this Court that restrictions on the freedom of speech must be couched in the narrowest possible terms.”
This is not all, however. Justice Nariman makes a further, crucial link: between vagueneness and overbreadth, and the chilling effect. The chilling effect refers to a situation where, faced with uncertain, speech-restricting statutes, which blur the line between what is permitted and what is proscribed, citizens are likely to self-censor, in order to be definitively safe. In the words of Justice Brennan, writing in New York Times vs Sullivan, citizens will “tend to make only statements which steer far wider of the unlawful zone… thus dampen[ing] the vigour and limit[ing] the variety of public debate.” In other words, the chilling effect – which applies across different areas of free speech law – ensures that self-censorship will extend even to entirely legitimate speech, and will impoverish the public discourse – the sustenance and enrichment of which is the entire point of free speech in the first place.
Yet again, it is crucial to note here that although past Indian cases have made vague references to the chilling effect (R. Rajagopal and Khushboo), again, the 66A judgment is the first that uses the concept to arrive at a positive legal outcome. In paragraph 83, after examining all the myriad kinds of speech that 66A will reach, Justice Nariman observes:
“Such is the reach of the Section and if it is to withstand the test of constitutionality, the chilling effect on free speech would be total.”
And, in paragraph 90:
“We, therefore, hold that the Section is unconstitutional also on the ground that it takes within its sweep protected speech and speech that is innocent in nature and is liable therefore to be used in such a way as to have a chilling effect on free speech and would, therefore, have to be struck down on the ground of overbreadth.”
Article 14 and differences by medium
One last point remains to be noted. The core of the government’s argument was that the internet is a very different medium from print or cinema, and that consequently, the government should be allowed greater leeway to regulate it. In paragraph 27, Justice Nariman lists some of the facets of the government’s contention: that the internet has a much greater (global) reach, it reaches both literate and illiterate people, even cinema has pre-censorship rules (but the internet doesn’t), rumours can spread to “trillions” (sic!) of people, there is much greater scope for invasion of privacy, the internet provides much greater shelter to anonymity, there are no internal regulatory norms, and that the spread is much more rapid.
In contradistinction, the challengers made precisely the opposite argument. They contended that since S. 66A lacked the kinds of procedural and other safeguards present for the regulation of print media, there was an Article 14 violation of equality. According to the challengers, a principle of equivalence must apply across media of communication.
Interestingly, Justice Nariman rejects both contentions. He rejects the Article 14 argument, holding that the internet is indeed a medium with some unique qualities, and that it is possible that there might be certain offences that can only take place online. In paragraph 28, he notes that the government is entitled to draft narrowly-drawn provisions that specifically speak to those offences (such as website blocking). But in the very same paragraph, he also notes:
“[the differential nature of the internet would not] relax the Court’s scrutiny of the curbing of the content of free speech over the internet. While it may be possible to narrowly draw a Section creating a new offence, such as Section 69A for instance, relatable only to speech over the internet, yet the validity of such a law will have to be tested on the touchstone of the tests already indicated above.”
In effect, what kind of speech might be restricted is agnostic to media. Furthermore, when it may be restricted (i.e., the 19(2)) principles, and the relationship of proximity) is also impervious to the difference in media. The only place where this difference might be relevant is where the medium itself allows for certain kinds of offences (such as spam, or phishing) that were not possible before, and in such circumstances, the State may frame a law, which will nonetheless be tested on the anvil of Article 19(2).
This raises the following question: in K.A. Abbas, the Court held pre-censorship to be valid in the case of cinema on the ground that films had a much more striking impact upon the average “illiterate” Indian viewer. In other words, the difference in medium was held to justify a difference in the form of the restriction – in particular, doing away with the proximity test, or at least, diluting it to an unrecognisable degree. The logic of Justice Nariman’s judgment, I would suggest, knocks the bottom out of the prior-restraint regime of film-censorship. It is not as if films permit the commission some specific kinds of offences that were not possible before (unlike the internet). In fact, the Court’s reasoning in K.A. Abbas was specifically based on an (unproven) assumption of how an “illiterate” audience reacts differently to the written word and the visual image. Cases after K.A. Abbas (such as Anand Patwardhan’s Case) have rejected the “illiterate Indian trope”, thus undermining the foundations of that holding. The 66A judgment, however, rejects that form of reasoning entirely.
Constitutionally, what follows? I would suggest the following. Ever since Ram Manohar Lohia, there has been a gradual, incremental chipping away at the philosophical foundations of some of our most regressive, speech-restricting laws. Consider the following:
(a) Ramji Lal Modi upheld 295A on the ground that proximity was irrelevant, specifically rejected an over-breadth argument, and held that deliberate insults to religion had a “calculated tendency” to cause public disorder.
(b) Kedar Nath Singh upheld sedition (124A) on the ground that disaffection against the government had the “tendency” to public disorder.
(c) K.A. Abbas upheld pre-censorship of films on the ground of how the medium differently affects viewers; it rejected a challenge to the Censorship Guidelines on the ground of vagueness, and did not even consider an over-breadth argument.
(d) Contempt of court judgments (unfortunately, rather consistently) have held that certain forms of criticism against the court have a “tendency” to undermine justice.
(e) Apart from Rajagopal’s Case, which is limited to public officials, the Supreme Court is yet to examine defamation law on the touchstone of Article 19(1)(a), and a criminal law of defamation continues to exist on the statute books. In countries such as the US, Canada, South Africa, and England, the “chilling effect” of the common law of defamation (strict liability) has been expressly invoked to limit its reach, and its propensity to be used as a tool of harassing journalists and investigative reporters.
At the same time, however:
(a) Ram Manohar Lohia, S. Rangarajan and Arup Bhuyan (the latter two coming after Modi and Kedar Nath Singh) have insisted on a proximity requirement between speech and disorder.
(b) Chintaman Rao and Kameshwar Prasad have struck down statutes on over-breadth grounds (while not directly invoking the concept).
(c) Rajagopal’s case has incorporated the Sullivan rule to protect writers from civil defamation claims brought by public officials, and indirectly invoked the chilling effect.
This latter group of judgments, it is clear, have indirectly, implicitly undermined the foundations of the former. The 66A judgment makes it direct and explicit. Proximity, over-breadth, vagueness and the chilling effect are all expressly invoked to strike down a statute. They have been specifically incorporated into Indian free speech jurisprudence, and may be invoked in future free speech claims.
So perhaps, at long last, the time has come to rethink fifty-year old judgments upholding blasphemy and sedition laws, rethink criminal defamation, throw off the oppressive fetters of civil defamation and contempt of court, and attack the censorship guidelines of both cinema and cable TV.
This judgment lays the constitutional, legal and philosophical terrain to do all of that. It creates the ground upon which we can erect a progressive, liberal understanding of free speech, based upon a vision of individuals as responsible and autonomous beings, who do not need “protection” from moral corruption, and who do not need a paternalistic State or court telling them what they need to read, see or here. Individuals, one and all, worthy of dignity and equal concern and respect.
This is the first judgment since 1960 that unequivocally affirms every one of those propositions.
It is the judgment that has given Indian free speech law its soul back.