Vancouver’s Human Flesh Search Engine

Photo by Richard Eriksson

I don’t like violence, vandalism, or other actions that generally cause destruction. Certainly there are cases where violent social dissent is a sad but important final step to fulfil a much needed social change (e.g. overthrowing a ruinous dictator, tipping the scale to defend or secure essential civil rights) but riotous behaviour following a hockey game lacks any legitimating force. Unfortunately, in the aftermath of game seven between the Vancouver Canucks and Boston Bruins a riot erupted in downtown Vancouver that caused significant harm to individuals and damage to the urban environment.

The riot itself is a sad event. What is similarly depressing is the subsequent mob mentally that has been cheered on by the social media community. Shortly after the riot, prominent local bloggers including Rebecca Bollwitt linked to social media websites and encouraged readers/visitors to upload their recordings and identify those caught on camera. In effect, Canadians were, and still are, being encouraged by their peers and social media ‘experts’ to use social media to locally instantiate a human flesh search engine (I will note that Bollwitt herself has since struck through her earliest endorsement of mob-championing). Its manifestation is seemingly being perceived by many (most?) social media users as a victory of the citizenry and inhabitants of Vancouver over individuals alleged to have committed crimes.

Perhaps unsurprisingly, I have significant issues with this particular search engine. In this post, I’m going to first provide a brief recap of the recent events in Vancouver and then I’ll quickly explain the human flesh search engine (HFSE), both how it works and the harms it can cause. I’m going to conclude by doing two things: first, I’m going to suggest that Vancouver is presently driving a local HFSE and note the prospective harms that may befall those unfortunate enough to get caught within its maw. Second, I’m going to suggest why citizens are ill-suited to carry out investigations that depend on social media-based images and reports.

Riotous Behaviour Comes to Vancouver

Following the conclusion of Stanley Cup final, downtown Vancouver was overcome by riotous behaviour. Cars were set ablaze. Violent altercations between emergency services and rioters occurred. Citizens fought with one another, as some tried to protect property that was being targeted by looters and malfeasants. Riot police were out in force. The city sky was laden with the smoke of meaningless destruction.

Photo by Matthew Grapengieser

Initial reports suggested that only a handful of people were involved, but it was quickly evident that a substantial number of people were taking part in the mayhem. While a considerable amount of damage was done we can hope that companies’ and individuals’ insurance policies will cover the physical losses. While many were administered to hospital there seem to have been surprisingly few serious injuries, with most being administered to hospital because of exposure to pepper spray and tear gas.

During and following the riot, Vancouver Police Department (VPD) has requested that citizens provide photos and video of riotous behaviour so that the cops can conduct investigations and bring charges. Importantly (and responsibly) the VPD asked for images and video to be shared directly with them; they don’t want individuals publicly posting images and video. There are at least two reasons for this kind of request:

  1. Preserving the chain of evidence. When you upload photos to Facebook, as an example, it strips the metadata off the images. As a result there is a more labor intensive process to verify the legitimacy and accuracy of images located on Facebook. I can guarantee that clever defence attorneys will point to a flawed chain of evidence stemming from lack of metadata information as reasons that social media-based imagery and videos should be disallowed from the courts.
  2. Understanding the danger of the mob. The VPD clearly recognizes the potential harms that arise when a mob races off to identify and convict, in the court of public opinion, individuals who allegedly engaged in illegal behaviour. I say allegedly because even if an image shows a person seemingly lighting a car on fire, until the court has evaluated and verified the evidence, and the person in the image has a chance to defend themselves (or admit their guilt) they are allegedly guilty. This presumption of innocence before the courts, regardless of evidence that exists to ground a case, is a fundamental cornerstone of our democracy and we should be mindful and respectful of it.

Unfortunately, our politicians have been less thoughtful. We’ve seen them call on citizens to publicly publish images using social media networks such as Twitter. This isn’t particularly surprising: I doubt that many of the politicians making such calls are trained in the retention of evidence, nor do I expect them to have thought about the longer-term impacts of associating names and images with people who may be acting like idiots whilst simultaneously not breaking the law (e.g. posing before a flaming car may be stupid but ought not to be automatically interpreted as being responsible for lighting the vehicle aflame). Especially when you consider how rarely some people are tagged in photos on the public Internet, and that many websites are associating names with photos on sites that Google spiders crawl, permanent images of stupidity may be misread long into the future as images of illegality. Such contextual inferences may negatively impact people for the rest of their lives. This is not a good thing: the Internet doesn’t forget and you simply cannot assume that a border guard or employer is going to spend the time (or have the wherewithal) to search for the context of stupid but not illegal images.

The Human Flesh Search Engine

The Human Flesh Search Engine (HFSE) is an engine driven by humans to find humans. It often involves the collection of open- and closed-source intelligence to find individuals who have allegedly broken a law or violated some social norm. Open-source intelligence is “freely available on the Internet through groups with open memberships or simply posted to websites” whereas its closed counterpart “is not publicly available and is associated with information security operations, intelligence gathering, “softer” business, and professional relationships” (PDF Source). The HFSE is in regular operation throughout Asia (and especially China and South Korea) and is used for what may be seen as admirable aims (ferreting out government corruption and perpetrators of illegal actions) and far less admirable aims (identifying and shaming individuals who are violate social norms or break incredibly minor laws). Regardless of the end result – which one might consider positive or negative – the means of arriving at that end tend to do violence to the individuals under ‘investigation’.

As examples, after information about a person is made public they are often subject to illegal forms of harassment, including hate mail, threatening phone calls, death threats, and other forms of psychic and physical harms. To be clear: the HFSE operates on a fuel of vigilantism that is bereft of legal or operational legitimacy. To be yet clearer: the ends, no matter what they happen to be, do not justify the means of arriving at them. Conducting an illegal form of investigation that lacks legal legitimacy and that may in fact taint the ability to lay legal charges against potential criminals because of potential evidence misuse/mishandling/fabrication is not a good thing. The rise of smartmobs, while certainly an interesting phenomenon from a scholarly point of view, are not something that we as a society necessarily should be condoning, supporting, or encouraging.

Vancouver’s Newest Search Engine

Within hours of the riot in Vancouver, locals were turning to social media and began to collate images and videos of individuals who were seemingly engaged in riotous behaviour. In addition to posting images there are public efforts to both identify

From Vancouver Riot Pics Facebook Group

individuals in photos and take a direct form of justice on them: rather than simply providing information to the police social media users were actively punishing suspected criminals. Photos were transmitted to employers, or intentionally made public so that future searches for those individuals would bring up the images/videos along with uncorroborated commentary. One blog, titled ‘Public Shaming Eternus‘, saw its owner (who refuses to publicly reveal themselves and instead operates under the pseudonym ‘Captain Vancouver’) write the following:

These young men and women IF they get prosecuted will most likely be given probation, perhaps fines, community service etc. If they do end up with a criminal record, they will also (if given a savvy enough lawyer) be able to have that black mark removed.  At the end of it all, the consequences are minor.

Public shaming however reveals their faces, their names and ultimately has longer lasting effect.  There are those who would say that our court system is for the public record to view.  When was the last time the average person sat in on a trial and read court transcripts?  They are all there for the viewing to see.  Public shaming thru the use of social media and blogging shall place them into a world that is engraved into history.  It shall be chiseled into the hard stone of the internet and last eternally.

When a person’s name is typed into google or any search engine, especially if they are relative nobody’s, they do pop up on the first page during the search.  More and more of today’s employers are googling the names of the people who apply for jobs. I know when I run my company, I wouldn’t hire the people I saw rioting in the streets of my city.  So if they show up in a search 5 years from now, let them explain in their job interview’s how they’ve “all grown up now”. Roll the dice and see if they still get the job.

The politics of shame should not be the ‘solution’ to criminal activity as it strikes to the heart of a person’s dignity instead of striking towards the (lack of) integrity of their actions. Our system of justice is not designed to ‘brand’ individuals with social-media inspired scarlet letters. Those that actively attempt to cause harm to others outside of the judicial system should at the very least have the courage to conduct their shaming actions publicly so that they can ‘enjoy’ the prospect of future civil litigation. What’s good for the goose is good for the gander, right?

In addition to attempts to permanently ‘brand’ individuals who may or may not eventually be found guilty of criminal actions, there are calls for violence to befall individuals pre-trial. The below screenshot from the Vancouver Riot Pics Facebook group is commonplace throughout the social media landscape at the moment.

From Vancouver Riot Pics Facebook Group

Posing before wreckage is not in itself a crime, though citizens are clearly posting the images with the implied assumption that those posing are guilty of a crime. Further, one poster is advocating the hanging of the individuals in the photograph using language reminiscent of the Ku Klux Klan’s racially motivated hate killings. I ask: would Mr. Griffin use the language he did were he privately communicating with the police to convey information about people who he suspected of criminal behaviour? I suspect that he would not.

In some respects, of course, social media is behaving similar to the morality plays that go on everyday on talk radio: a key (if unspoken) element of the mob-usage of social media is to rearticulate the community’s social norms. This isn’t to suggest that local Vancouver norms are to hang prospective criminals, but instead that there is a permissiveness to “Othering” certain people who are merely suspected of wrongdoing, or who are perceived as ‘less real’ or ‘unclean’ in comparison to the community’s membership. The language of the noose is unfortunate in many respects, both in the sense that it articulates a glee for vengeance and the acceptability of drawing on racist overtures in the Othering of photographed individuals.

The language and articulation of shame towards alleged rioters is commonplace in social media environments right now, though it is perhaps best captured by Ms. Kashino’s decidedly non-unique statement:

From Vancouver Riot Pics Facebook Group

The celerity at which individuals engage in what is the digital equivalent of ‘water cooler talk’ – discussions that were historically informal, rarely thought out, hidden from a wider public audience, and non-permanent – and the longevity of such comments are common problems with social media. These are issues that have been discussed widely within academic, policy, and public circles and are far from being resolved. Personally, I tend to agree with Dan Solove in The Future of Reputation: online social networking environments lack traditional media’s norms of restraint (though I’m not arguing traditional media has been much better) and that the “less-well-developed-norms” of social networking need to be reconstituted with a widely held code of ethics that sees people reflect and filter their comments online prior to uttering/writing them. This isn’t meant to suggest that individuals should censor their language, but that they should be considerate in the language that they utter. If in their considered opinion racist, misogynistic, or homophobic language is appropriate then let them utter it, though one hopes that in their moment of reflection they realize and acknowledge the harms associated with such modes of verbal violence.

Citizens Are Not Trained Investigators

Statements made on social media have the ability to leave long-lasting scars on individuals as well as misdirect anger and the human flesh engine more generally. Consider the early reports of a Bruin’s fan being discovered bleeding on the ground:

 

 

 

While I would agree that this is a sad image, the full story is that the Bruin’s fan began a physical altercation and was subsequently knocked down by a pair of local security guards. Should the guards have their images circulated for using excessive force, given their own lack of legal authority to cause bodily harm? Or are they heroes for protecting property and individuals? Or, in fact, is the ethical situation around this image (along with many, many other photos and video clips) far more complicated? Are citizens ideally situated to evaluate the legalities and ethics of how these violent altercations were, and should have been, resolved?

While this was a case where the managing editor of a citizen-journalism site did a followup on claims, in doing so he demonstrated the value of journalistic ethics and integrity. Most citizen-journalists lack the ethical training and vetting processes that professional journalists go through. I’m not, of course, saying that to post something online one must have gone to journalism school: the point is simply that the alacrity at which citizen-journalists will press ‘post’, ‘like’, or ‘tweet this’ when the implications are potentially far-reaching will often diverge from the alacrity of a principled, ethical, and well-trained journalist. Further, one would not expect a journalism site to advocate for identifying and shaming individuals who have allegedly been involved in criminal acts. The same cannot routinely be said for citizen-journalists.

Beyond the notion of ethics and reserve that are often lacking when people turn to social media, there is the simple fact that citizens are not trained investigators. As such, it is entirely possible that in conducting their ‘investigations’ to trace and identify individuals in photos that they may behave in illegal or unethical manners themselves. Further, the processes used to identify individuals may in fact contaminate their very findings: while you may learn who John or Jane Doe are, if in the process you illegally collected information then those defendants are likely to argue that the means of identification are inadmissible and try to leverage that to avoid a guilty verdict. Of course the success of any such legal maneuver will vary, but that it is even possible is the result of untrained individuals using questionable methods as part of the Human Flesh Search Engine.

Finally, I would simply note that citizens are incredibly poor in conducting and acting on surveillance footage. In cases where normative or decisional actions are part of the surveillance process the untrained are often more of a hindrance than asset; they are truly best at identifying black/white kinds of evidence and then alerting trained investigators to the evidence. To clarify: watching a border or photos to identify illegal acts and individuals responsible involve normative and decisional processes that citizens are unqualified to responsibly and professionally engage in. Recognizing that a photo or video might be useful to authorities and passing it on – and letting them take action on the content – is a case of black/white identification. Effectively, if you outsource policing to amateurs you get amateur results that could bungle broader investigations and criminal proceedings.

Conclusion

I conclude by taking pains to largely reiterate the earliest statements of this post. I in no way support, condone, or like the riotous behaviours that were evidenced in Vancouver, and nothing in this is intended to absolve the individuals responsible for it of their guilt. Instead, I simply warn that a human flesh search engine that functions per the logics of vigilantism and shame are in clear violation of the norms and ethics of Canadian policing and judicial procedure, can cause harm far in excess of an action captured by a digital recording, and potentially undermine authorities’ capacity to prosecute alleged criminals. Further, citizens are not trained to analyze open source intelligence, nor are they ideally suited to know the full extent of what they can and cannot do to identify and name someone. The intent to cause harm in excess of the judicial system is not only detrimental to those targeted by members of the Human Flesh Search Engine but, potentially, to those members should they face tort-based legal challenges in the future. While the reaction of social media users to identify and shame alleged rioters is unsurprising this doesn’t mean that the practices should go unchallenged or left free of critique. The actions taken to identify, name, and shame alleged rioters is the beginning of a slide towards a state of mind and looseness of ethics that have been proven to cause harm abroad: I see no reason, based on those experiences, why we should import known, failed, modes of citizen surveillance and investigation into the Canadian experience.

Christopher Parsons

Managing Director of Telecom Transparency Project at Citizen Lab
I’m a Postdoctoral Fellow and Managing Director of the Telecom Transparency Project at the Citizen Lab in the Munk School of Global Affairs at the University of Toronto. I'm also a Principal at Block G Privacy and Security Consulting. My research interests focus on how privacy (particularly informational privacy, expressive privacy and accessibility privacy) is affected by digitally mediated surveillance and the normative implications that such surveillance has in (and on) contemporary Western political systems. I’m currently attending to a particular set of technologies that facilitate digitally mediated surveillance, including Deep Packet Inspection (DPI), behavioral advertising, and mobile device security. I try to think through how these technologies influence citizens in their decisions to openly express themselves or to engage in self-censoring behavior on a regular basis.