While visiting London in the summer of 1857, the Baron Carl de Gleichen, a man of complex nationality and advanced views, was set upon by denizens of the Victorian underworld and robbed. His assailants were caught and brought before the Marlborough Street Police Court. However, because the baron would not say he believed in a future state after death in which he would be “rewarded or punished according to his deserts,” they were set free. At English common law, the baron could not take an oath if he did not think a supernatural force would punish him for breaking it,1 and since he was the only witness, there was therefore no evidence with which to convict.

In 1992, an Ontario high school student – known to us as “N.S.” – told a trusted teacher that she had been repeatedly raped by her cousin and uncle from the age of six. Her family did not want to take any action and the police did not lay charges. It is hard to imagine N.S. had the self-confidence generations of privilege and freethinking had bred into the baron. Fifteen years later, though, she tried again. A Crown prosecutor was sufficiently persuaded of the plausibility of her evidence to allow charges to proceed. By this time, N.S. had developed the religious conviction that she must wear a niqab, a veil that covers her entire face other than her eyes, when in the presence of men outside her direct family.

Although they attended the same mosque as N.S., the accused men asked for a court order that N.S. remove the niqab while testifying. They argued that only by seeing her face could the judge or jury tell whether she was lying. As a result of a complex decision released by a divided Supreme Court of Canada in 2012, we do not know whether N.S. will be compelled to choose between obeying her religious convictions and testifying against her alleged assailants.2 On the basis of the Supreme Court decision, an Ontario Court judge has decided that N.S. must remove her veil to testify; she plans to appeal. In any case, it is clear that some Muslim women will not be allowed to testify in Canadian courts if they will not show their faces.

Religious belief and the competence of witnesses

Google remembers Baron de Gleichen today because his treatment by the English criminal justice system outraged John Stuart Mill:

This refusal of redress took place in virtue of the legal doctrine, that no person can be allowed to give evidence in a court of justice, who does not profess belief in a God (any god is sufficient) and in a future state; which is equivalent to declaring such persons to be outlaws, excluded from the protection of the tribunals; who may not only be robbed or assaulted with impunity, if no one but themselves, or persons of similar opinions, be present, but any one else may be robbed or assaulted with impunity, if the proof of the fact depends on their evidence … Under pretence that atheists must be liars, admits the testimony of all atheists who are willing to lie, and rejects only those who brave the obloquy of publicly confessing a detested creed rather than affirm a falsehood.3

Outlawery was the gravest punishment of the Anglo-Saxon legal order because an outlaw could be subjected to violence with impunity, and Mill knew the term resonated with his Victorian audience. He pointed out that the protection of the criminal law does not extend to a person who cannot give evidence in court. He had enough imagination to see that a judge or jury drawn from mainstream society will discount the honesty of a person who conscientiously holds minority opinions more than they should, and that there is therefore no danger that such a person’s evidence will be given excessive weight. By holding on to this “relic of persecution,” society injured both the religious liberty of its freethinkers and its own safety.

On Liberty is a favourite of undergraduate education in political theory and philosophy. It is clearly written and reasonably short, and easily inspires classroom discussion about the implications of the harm principle for censorship of pornography and hate speech. As an undergraduate, I must have been assigned it at least three times. But I don’t remember anyone lingering over the case of Baron de Gleichen. Classroom time is scarce, and I suspect it would have seemed too easy: an aftereffect of religious bigotry without any plausible secular justification that was anachronistic even at the time. Debates about the limits of freedom of speech were burning ones in the social science and humanities faculties of the early 1990s, but debates about the relationship between religious beliefs and the competence of witnesses in court proceedings, if we had thought about them at all, would have seemed about as relevant as the divine right of kings.

N.S.’s case shows we would have been wrong: de Gleichen’s case was not an easy one, and when faced with similar issues today, there is no guarantee we will get them right. With the exception of Justice Rosalie Abella, who consciously or unconsciously echoed Mill, the judges of Canada’s highest court decided that juries cannot be trusted to weigh the evidence of a woman who braves the obloquy of wearing a niqab. In Mill’s terms, they have declared such women outlaws, excluded from the protection of the tribunals. Two justices would never allow the evidence of a veiled woman to go to a jury; the majority would give trial judges the discretion to allow it when the evidence of marginal importance, but not when it is central to the case and credibility is in issue. The issues Mill dealt with remain living ones. Questions of who can give evidence go to the heart of who is included in a society both because a person who cannot give evidence is outside the protection of the legal system and because evidence depends on a consensus on procedures for determining who can be trusted. We are still a long way from those procedures being rooted in science, which means they must depend on cultural expectations, which in turn are more rooted in religion than secular undergraduates are likely to realize.

Oaths and early modern English law

As is so often the case, condescension about the past blinds us to a more interesting story. In fact, easy as it is to forgive him, Mill was unfair to the authors of the common law rule he decried. They were relative cosmopolitans who acted not out of religious bigotry or hatred but from purely secular motives.

Like most legal rules in those days, the “rule in Omichund v. Barker” arose out of concrete litigation. Barker was an Englishman trading in Calcutta in the first half of the 18th century, before British dominance was established in India. He ran up a large debt with a local merchant, Omichund, and sailed home rather than pay it. In 1744, Omichund sought the assistance of the English courts in enforcing his debt. The English Court of Chancery allowed for written depositions under oath from residents of foreign countries, a process that gave it a lot of international commercial litigation. Omichund’s witnesses swore an oath in accordance with Hindu custom and Barker objected that only Christians could swear an oath that would be admissible in an English court.

The issue was of enough importance that the Lord Chancellor asked the chief justices of the other English courts to convene a panel to rule on it. Britain’s growing power depended in large part on the ability of its newly independent legal institutions to ensure commercial security, which would be undermined if they could not accept evidence in commercial matters from outside the Christian world. But secular law was still closely associated with Christianity. The Lord Chancellor himself, although no longer usually an ecclesiastic, still played a significant role in appointing bishops of the Church of England. More importantly, the law was extremely concerned about perjury and tightly restricted who could testify in a case. Parties and spouses of parties could not testify in their own case. People under 21 could not testify. Perhaps the most influential common lawyer of all time, Lord Coke, had specifically said that “infidels” could not testify.

The opinion that comes down to us was that of the Lord Chief Justice of the Court of Common Pleas, Sir John Willes. Willes was by all accounts a worldly Hanoverian gentleman. According to Horace Walpole’s catty but entertaining memoir, Willes’s passion for gaming “was notorious; for women, unbounded.” Willes recognized it was “greatly to the advantage of this nation to carry on a trade and commerce in foreign countries.” He also recognized how important it was that an oath be regarded as binding by the person swearing it. Willes had the education in classical and biblical literature to easily demonstrate that oaths are universal institutions long predating Christianity. What mattered was not whether Omichund’s witnesses’ beliefs were true, but whether they created a motive to be honest. Willes and his fellow chief justices told the Lord Chancellor that Omichund’s evidence should be admitted.

Indeed, Willes’s rhetoric sometimes reminds the modern reader of John Stuart Mill. Willes denounced the “little mean narrow notion that no one but a Christian can be an honest man.” Still, as a betting man he was quite concerned about probabilities, and he was prejudiced enough to give the prior probability of a Christian’s telling the truth a higher estimate. Willes said Christians were “under much stronger obligations to swear nothing but the truth” than those who subscribed to other religions. But he also assumed, no doubt correctly, that English judges and juries would think likewise and could safely be entrusted with the decision of how much weight to give infidel witnesses.

This was an important breakthrough in an era when the ability to give evidence was so restricted by rules based on reliability. Willes was the product of a legal system that took the propensity of human beings to perjure themselves for granted and did not particularly trust juries to figure things out: that is why the system would not let anyone with an interest in a case testify. Omichund actually represents a landmark in the development of the more modern principle of evidence according to which anything relevant can usually be admitted unless there is good reason to think people have cognitive biases that mean they will put much more weight on the evidence than they should. Of course, common sense is imperfect and subject to prejudices, which no doubt sometimes put Hindu creditors at a disadvantage relative to their defaulting Christian debtors. On the other hand, while they were at a disadvantage, they were not excluded. Non-Christians could continue to use English courts; British commerce could continue to build on reliable legal institutions.

Willes stipulated the limit that caught the baron a century later as follows: “Such infidels (if any such there be) who either do not believe in a god, or, if they do, do not think he will either award or punish them in this world or in the next, cannot be witnesses in any case.” As his parenthetical remark suggests, Willes was impressed by the universality of oath-taking and did not seem to think there was anyone who failed to give it supernatural sanction and would therefore be caught by the rule. In any event, he needed a limit because he needed the oath. The 18th century had no great faith in human honesty. The oath was both a ritual and a technology for distinguishing statements on the basis of which a person’s life or property could be taken.

By the time Mill wrote, the era of parliamentary legal reform had begun. Protestants in the Anabaptist tradition who believed Matthew 5:34 prohibited oath-taking4 were a loyal part of the Liberal-Radical coalition, which had introduced legislation to allow for “solemn affirmations.” This legislation excluded atheists and did not change the common law requirement of belief in supernatural enforcement in relation to oaths. However, it was relatively easy to imagine extending affirmations to freethinkers and atheists, and by the end of the 19th century, this had occurred.

It remained open to 20th-century juries and judges to give greater weight to the testimony of witnesses who appear to think they will be subject to divine wrath if they do not tell the truth. In fact, difficult as it is for me to admit as a secular person, this may be a valid statistical generalization; the question of the impact of religious belief on truthfulness and other prosocial behaviour is a complicated one. While empirical sociologists in the sixties seemed to show a lack of correlation between antisocial behaviour and lack of religious belief, subsequent results are, well, complicated.5 Suffice to say that while Mill and Willes were definitely right about a lack of any necessary relationship between religious belief and honesty, Willes may have been on to something about the way to bet. There is no real reason to think that ordinary people are systematically deluded about these effects and, even if they were, long before the Charter of Rights, no one thought it would justify denying secular people the protection of the tribunals. Richard Dawkins himself can visit Toronto secure in the knowledge that if he is mugged on Philosopher’s Walk, he will be allowed to testify about it. We have made that much progress.

The Canadian conception of freedom of religion from the eighties to the present

At this point, the reader will already be anticipating that I will argue that the solution our great-grandparents came up with in relation to atheists and freethinkers in the 1890s ought to be extended to women whose interpretation of Islam requires them to wear the veil in court today, and may have framed some objections. Some may think there is just a more rational basis for worrying that testimony by veiled women will lead to inaccurate verdicts than was the case with enlightened barons. That objection will be addressed when I get to the Supreme Court’s own reasons. It turns out that the evidence, such as it is, says the opposite.

But a different objection demands a digression before we get there. A rule requiring a witness to swear to a belief in supernatural consequences for perjury differs from a rule requiring witnesses to reveal their faces when testifying. The question of supernatural consequences for perjury specifically refers to religious or metaphysical belief, while the question of whether one must reveal one’s face does not. On its own terms, the common law would not accept Baron de Gleichen’s evidence because of his sincere religious convictions. In contrast, the rule announced by the Supreme Court does not say N.S. must testify in a certain way because she is Muslim: rather, her interpretation of Islam means she cannot conscientiously follow the rule.

The most fundamental issue in the law of freedom of religion is whether this distinction matters. On one view, a law does not interfere with freedom of religion as long as it is written in general terms and is not motivated by religious bigotry: if a religious believer feels she cannot comply with it, that is her problem – the law is fine. In 1990, in Oregon v. Smith, a majority of the American Supreme Court endorsed this limited view.6

The case was brought by adherents of the Native American Church, which prescribes the use of peyote in its rituals. The adherents claimed an immunity from the general Oregon law prohibiting possession of peyote. The majority of the highest American court said the right of free exercise does not relieve an individual of the obligation to comply with a “valid and neutral law of general applicability on the ground that the law proscribes (or prescribes) conduct that his religion prescribes (or proscribes).” A more militant version of this objection holds that allowing exemptions from general rules based on religious beliefs is itself a violation of state religious neutrality – a concept associated with “strict separation” in the United States and laïcité in France.

Oregon v. Smith was immediately controversial. If it is right, then religious freedom has very little contemporary relevance, since laws almost never single out religious belief, but majoritarian practices often cause problems for believers. The U.S. Congress enacted the Religious Freedom Restoration Act (RFRA) in 1993 in specific response to Oregon v. Smith. The authors of the RFRA decided that free exercise of religious freedom needs to be protected by giving courts or other bodies the power to scrutinize general laws that have the possibly unintentional effect of interfering with sincere religious practice. A law stipulating that everyone must work on Saturday or accept blood transfusions is no big deal for a Methodist, but on this broader view would render Orthodox Jews and Jehovah’s Witnesses less free.

Until the decision in N.S., it was uncontroversial in the Supreme Court of Canada, although not in the political sphere, that the broader view of religious freedom was the right one. If a general law or policy substantially interferes with a person’s ability to act in accordance with his or her sincere religious beliefs, then it offends that person’s freedom of religion under section 2(a) of the Charter of Rights and Freedoms. It cannot matter whether the sincere religious belief is a reasonable one from the point of view of a secular court and there need not be any religious motivation to the law or policy.

There is a catch, though – there has to be. Some sincere religious beliefs (for example, a belief in human sacrifice) would be completely intolerable if put into practice. Some in the media remarked that the N.S. case involved a “contest of rights” since the accused men complained that if N.S. could testify without showing her face, their right to a fair trial would be compromised. In fact, this problem is universal in freedom of religion cases. They are all contests of rights. Any law or policy has at least a perceived benefit, and therefore removing it, or even exempting a subsection of the population from its effect, has a corresponding cost. Someone must pay that price, and it is in the nature of freedom of religion cases that religious belief will be a determining factor in who that is. If my coworker cannot be required to work on Saturdays because she views them as sacred, then I must take her shift precisely because I do not share that belief. The problem is reasonably tractable in a society in which there is a secure and normative religious tradition that agrees to accommodate everyone else in exchange for its dominance. Canada is increasingly not that type of society.

Judges cannot solve this problem, but they can negotiate it. Canadian judges do so by invoking section 1 of the Charter, which affirms the protected rights and freedoms “subject only to such reasonable limits prescribed by law as can be demonstrably justified in a free and democratic society.” Courts take a look at the secular justifications for the policy and decide whether the burden imposed on the believer is greater or less than the burden on society of choosing some alternative. In practice there is no metric with which to measure the weight of a burden on a sincere religious belief, so the court really considers how compelling the secular justifications for the law or policy are.

The result is that freedom of religion cases do and should turn into grubby exercises in analysis of often technical policies. Those seeking grand principles will not find them. Sunday observance laws are not acceptable as criminal prohibitions, but are fine as labour regulation. Observant Sikhs need not wear hard hats on construction sites, and their children may wear ceremonial daggers in public schools, but Jehovah’s Witnesses under 16 cannot prevent medically necessary blood transfusions. Observant Jews can set up a succah on commonly owned property contrary to the terms of a declaration of co-ownership, but a Hutterite cannot refuse to provide a photo for a driver’s licence.7

The results in these cases are not easy to predict. Many of the decisions were accompanied by strong dissents, so that all we can say about the result is that it got the most votes in an electorate of nine. But any legal regime will have hard cases, and hard cases are the ones that will go to the highest court. The more fundamental objection comes from those who believe religious freedom requires a strict and formal theological neutrality from the state, a neutrality that is inconsistent with all this balancing and accommodation. But after being buried so overwhelmingly by the Supreme Court early on, is this alternative conception capable of revival?

The Supreme Court rarely explicitly refers to public controversies it may become embroiled in, but there is little doubt that it is aware of them. The basic framework of analysis for religious freedom in Canada dates back to the 1980s, when it had extremely low political salience. Economic forces and consumer demand pushed Canadians toward Sunday shopping faster than the courts did, and hardly anyone really cared about Sikh construction workers wearing hard hats.

Anxieties created by Canada’s changing demographics remained below the surface, at least from the perspective of elite opinion. September 11, 2001, brought to the fore an anxiety about Islam in particular that has always existed in Western civilization. Because the Canadian political system is not well designed for an open discussion of the optimal amount of assimilation of recent immigrants, the topic plays out in populist sociodramas.

The 2006 Multani decision permitting kirpans (ceremonial daggers) in schools triggered one such, when the town council of the village of Hérouxville, Quebec, enacted a charter for new immigrants forbidding both the use of kirpans and the wearing of the niqab and hijab. Premier Jean Charest reacted to the subsequent controversy by appointing sociologist Gérard Bouchard and philosopher Charles Taylor to head a public inquiry into “Accommodation Practices Related to Cultural Differences.” When Bouchard and Taylor reiterated much of the orthodoxy about religious accommodation in their 2008 report, it met with a cool reception from the public. All three provincial parties immediately rejected its recommendation that the crucifix hanging in the National Assembly be removed.

While the Charter and multiculturalism in the abstract remain enormously popular in Canada, the low salience of accommodating religious minorities and apparent consensus in favour of such accommodation no longer seem to exist. A 2008 poll by the Insitute for Research on Public Policy showed 53 per cent of respondents opposing accommodation of religious and cultural minorities in comparison with only 18 per cent supporting that approach.8 Although the heroic rhetoric of constitutional judicial review would hold that courts ignore public opinion, Multani remains a high-water mark for constitutional religious accommodation, which has not done as well in the highest court since then.

In N.S., the majority of the Supreme Court continues to analyze religious freedom issues in more or less the old way. The issue is still how big a deal it would be to change the secular rule, with the Chief Justice and three others deciding it would be too big a deal in most circumstances, while Justice Abella concludes that it would not. However, two justices, LeBel and Rothstein, seem to me to abandon the historic Canadian analysis entirely, albeit not explicitly. Their decision does not follow the familiar (to Canadian constitutional lawyers) steps of justification of limitations of rights, but simply asserts that court proceedings are a communication process and “wearing a niqab … does not facilitate acts of communication.” However, they do not really try to justify a new departure along the lines of Oregon v. Smith, a departure that would require backing away from three decades of decisions.

Of course, the reader of Inroads is not subject to the institutional constraints of Supreme Court justices, and can decide that we made a bad turn with the Sikh hard hat case. At least in the long run, supreme courts tend to follow national elite opinion, and I cannot claim that there is a drop-dead logical argument that compels us to avoid the path of Oregon v. Smith on pain of self-contradiction. But to me the lessons of Omichund v. Barker suggest that we should not follow that path.

It is obvious that Baron de Gleichen’s religious freedom was violated, as was the freedom of other skeptics who had to pretend to believe in a deity to get ordinary criminal and civil justice. It’s not that the rule that caught him up was explicitly theological in motivation; it’s just that the reasons for that rule were not really good enough when society was actually confronted with the need to think about them. And yet it’s only with the broader conception of religious freedom we have had since the eighties that the need to think about our reasons arises. And surely we owe N.S. very good reasons if we are to expect her to recognize the legitimacy of a system that allows violence against her to remain unpunished.

The broader conception of religious freedom at least provides a framework for requiring such reasons, but for there to be more than a framework, the court’s scrutiny needs to be a searching one. In cases like N.S., where the traditional rule was made by the courts themselves, we should be skeptical that such scrutiny will really be forthcoming. To see why, we need to turn to the decision itself.

Not necessarily removal,
but removal if necessary

Frank Scott famously suggested that William Lyon Mackenzie King would not let his “on the one hand” know what his “on the other hand” was doing. Chief Justice McLachlin would perhaps not appreciate the comparison, but she is comfortable portraying her decisions as treading a middle path between two extremes:

A secular response that requires witnesses to park their religion at the courtroom door is inconsistent with the jurisprudence and Canadian tradition, and limits freedom of religion where no limit can be justified. On the other hand, a response that says a witness can always testify with her face covered may render a trial unfair and lead to wrongful conviction.

In the end, the Chief Justice gives trial judges the discretion to order the niqab removed or not, depending on the importance of the evidence and the possibility of a case-specific compromise. Not necessarily removal, but removal if necessary.

As I hope the digressions about Omichund v. Barker and past religious freedom jurisprudence show, the Chief Justice was exactly right about the problems with such a “secular response.” However, in light of this, and even more importantly in light of the fact that excluding N.S.’s evidence renders her an outlaw in precisely Mill’s sense of subjecting her to violence with impunity, the burden of proof is surely on the Chief Justice to show when testimony with a face covered poses a risk of unfair trials and wrongful convictions.

In fact, as the Chief Justice acknowledges, the cognitive science evidence is the other way. The evidence before the court at first instance, and the bulk of the articles submitted on appeal by various interveners, supports the basic finding that people are likely to overestimate how much they can tell about a witness’s truthfulness by looking them in the face. The Chief Justice noted that the evidence before them was fairly weak, and I sympathize with a demand by the Court for better social science before we dispense with common sense and tradition.

The problem, though, is that the Chief Justice ignores the possibility that a contextual case-by-case balancing can be done by the jury. They are the ones chosen for common sense and the ability to distinguish lies and errors from truth. Juries are not perfect, and cognitive science can reveal biases that really do cause wrongful convictions. The most important is probably sincere but mistaken identifications. Lots of evidence suggests that juries and judges give far more weight to these than they should (unfortunately, the court system has been very slow to correct this bias). Another is forensic science, where juries (and, unfortunately also judges) have a propensity to be bamboozled by confidently presented work that may be subject to all sorts of biases.

However, even if the Chief Justice was unwilling to accept the evidence that juries and judges are overconfident about their ability to tell whether someone is lying by looking at them (demeanour), there is absolutely no reason to think they err in the opposite direction. There is no reason to think that an ordinary jury, in a country where a majority of people would ban the niqab if they could, will give too much weight to the words of a woman who will brave the obloquy of publicly confessing to a detested version of a marginalized creed.

Defence lawyers can, and will, draw the attention of juries, or of judges sitting alone, to the fact that they could not see the complainant’s face in sexual assault trials. It is far more likely that this will raise a “reasonable doubt” when it should not than that it will fail to do so when it should.

As Justice Abella pointed out, we allow people whose facial features have been frozen by accident or illness to testify. The Charter itself guarantees the right to an interpreter, even though that obviously reduces the ability of juries and judges to “read” a witness. The Court has relaxed the traditional rules against hearsay, effectively allowing out-of-court testimony that cannot be seen at all when there is no better evidence. In the absence of a clearly demonstrated cognitive bias, none of this is unfair if the defendant can poke holes in it.

While I think we should be glad that the Court’s majority has left the door open for future courts to look at better cognitive evidence and let women wearing niqabs testify, I fear this will be unrealistic in sexual assault cases. The effect will be to render those women outlaws – cruelly ironic in light of the usual objection to the niqab as oppressing those very same people.

Chief Justice McLachlin is a reasonable, moderate person, as was Chief Justice Willes three centuries ago. In both cases, the decisions could have been much worse, but both have had the effect of creating outlaws based on religious convictions. I hope the next generation of legal leadership can do better.


1 Omichund v. Barker, (1744) 125 E.R. 1310 (Ch.). Only Protestants who objected to oaths on religious grounds could give their evidence in any other way. The same rule prevailed in Canada until the end of the 19th century (Bell v. Bell, N.B.R. 615 ). Affirmations by those who would not fit the Omichund rule were allowed for criminal law purposes by section 23 of the Canada Evidence Act, 1893.

2 R. v. N.S., 2012 SCC 72.

3 John Stuart Mill, On Liberty, ch. 2, “On Liberty of Thought and Discussion.”

4 “You have heard that it was said to those of ancient times, ‘You shall not swear falsely, but carry out the vows you have made to the Lord.’ But I say to you, Do not swear at all, either by heaven, for it is the throne of God, or by the earth, for it is his footstool, or by Jerusalem, for it is the city of the great King.” (Matthew 5:33–35, New Revised Standard Version)

5 Travis Hirschi and Rodney Stark, “Hellfire and delinquency,” Social Problems, Vol. 17, No. 2 (1969), pp. 202–13. Later, Rodney Stark reviewed the state of the empirical research, and decided that there is an effect where religious belief is locally dominant; see his “Religion as context: Hellfire and delinquency one more time,” Sociology of Religion, Vol. 57, No. 2 (1996), pp. 163–73.

6 494 U.S. 872 (1990).

7 R. v. Big M Drug Mart Ltd., 1 S.C.R. 295; R. v. Edwards Books and Art Ltd., 2 S.C.R. 713; Bhinder v. Canadian National Railway Co., 2 S.C.R. 561; Multani v. Commission scolaire Marguerite-Bourgeoys, 2006 SCC 6, 1 S.C.R. 256; A.C. v. Manitoba (Director of Child and Family Services), 2009 SCC 30, 2 S.C.R. 181; Syndicat Northcrest v. Amselem, 2004 SCC 47, 2 S.C.R. 551; Alberta v. Hutterian Brethren of Wilson Colony, 2009 SCC 37, 2 S.C.R. 567

8 The poll and other surveys are discussed in Stuart Soroka and Sarah Roberton, Literature Review of Public Opinion Research on Canadian Attitudes towards Multiculturalism and Immigration, 2006–2009 (Ottawa: Citizenship and Immigration Canada, 2010), retrieved from here

Niqab photo courtesy Aulia Halimatussadiah/Flickr