Tuesday, September 23, 2014

Data mining protest violence: activist boon or social control?

There's no getting around the record: protesters and police have a long and storied history of conflict in these United States.

A few U.S. highlights: the May 3rd workers' rally in Chicago in 1886 that preceded the next day's Haymarket massacre; the 1965 civil rights march out of Selma known as "Bloody Sunday"; the Democratic National Convention of 1968; Seattle's WTO protests in 1999; the Occupy melees of 2011, most notably in Oakland, California ... and then there's last month's militarized suppression of protest in response to the fatal shooting of Michael Brown, an unarmed teenager, by a police officer in Ferguson, Missouri.

Via a headline story in the SF Chronicle, recent events in Ferguson surfaced ongoing research at UC Berkeley -- the Deciding Force Project -- in which sociologists have begun to use cutting edge data mining techniques to analyze police-protester interactions, and identify circumstances and tactics that lead to violent conflict. According to the article, the objective of this research is to reduce such violence, to the degree conscious and well-informed decisions on the part of police and of protesters can defuse volatile situations.

Let's say that the Deciding Force Project is successful in identifying ways to keep protest from boiling over into violent conflict. And let's say the project's research and analyses is made available to everyone, giving all parties access to information that describes conditions that lead toward and away from protests turning into riots.

Would activists and police/government benefit equally from this research? And is rigorous avoidance of violent conflict a goal that advances progressive political goals?

It's easy for most people to accept that non-violent exercise of democratic rights is 'better than' violent conflict. In general, I believe that is true. It's also easy to assume that more information is 'better than' less. It's hard to make a reasonable case for ignorance.

On the other hand, when civil discourse, electoral engagement, and peaceful protest fail to resolve weighty injustices -- what is to be done? And when information and the insights it facilitates are coupled with state and/or corporate power, many (including this writer) believe that its collection, analysis, and use become a risk to broadly-participatory democracy and to progressive political goals.

It's complicated.

Work like the Deciding Force Project could be a boon to activists, who might use it to base strategic organizing on new and deeper insight into the way crowds of protesters and battalions of police interact. Or -- with apologies for the hyperbole -- research in this vein could be developing a kind of information-based soma (à la Aldous Huxley's Brave New World), which might be deployed by the surveillance state to neutralize dissent. There's also the possibility that research of this sort won't deliver on its promise: that it won't predict the relationship of specific behaviors to on-the-ground outcomes any better than seasoned police and activists have done since time immemorial, on the basis of experience, familiarity with their own communities, and intuition.

'Big Data' and its analytical findings are part of the modern mix, whatever effect it might have. That's a fact. To my way of thinking, its introduction into political space demands attention and debate on the spectrum of possible roles 'Big Data' might play in relation to grassroots activism.

This post is not aimed at providing definitive answers. I do hope to raise questions and ideas worth examining.

Sociologists at UC Berkeley research police-protest interactions

Background first. From the San Francisco Chronicle on 22 Aug 2014, Police tactics often provoke protesters [print-edition headline]:
The violence that turns a small-town protest into a fiery national spectacle like the one that has played out this month in Missouri is often unwittingly provoked by police, according to researchers at UC Berkeley.

The research team, which studied clashes between police and activists during the Occupy movement three years ago, found that protests tend to turn violent when officers use aggressive tactics, such as approaching demonstrators in riot gear or lining up in military-like formations.

Recent events in Ferguson, Mo., are a good example, the study's lead researcher said. For nearly two weeks, activists angered by a white police officer's fatal shooting of an unarmed black teenager have ratcheted up their protests when confronted by heavily armed police forces.

"Everything starts to turn bad when you see a police officer come out of an SUV and he's carrying an AR-15," said Nick Adams, a sociologist and fellow at UC Berkeley's Institute for Data Science who leads the Deciding Force Project. "It just upsets the crowd."
On the day it appeared in the SF Chron, one activist friend responded to this article by posting it on Facebook, framed by the pithiest of snark:
Song in the key of duh...
Yup. If you've been around the activist block once or twice, you know this tune by heart. But there's more to the story than the article reported.

Applying 'Big Data' analytic techniques to police-protester interaction

What's new and perhaps significant about the Deciding Force Project is its focus on dressing up the obvious in scientific regalia, backed by the imprimatur of 'Big Data' analytic techniques of the kind employed of late by Facebook, Google, and the NSA.

Here from the SF Bay Guardian on 20 Aug 2014, Researcher explores police and protester violence in the Occupy movement:
Adams and the researchers trained computer programs to pick similar data from the over 8,000 news reports, automating the process. Articles from Boston, Chicago, New York, San Francisco, and more than 200 cities with Occupy movements are parsed for patterns. Did the police wear riot gear? What formations did they use? Were horses present? Assault vehicles? Was the crowd mostly Latino, black, white, Asian, or a mix? Were the Occupiers sitting or standing? These are [a] few of the hundreds of variables crunched by Adams' team.

After the variable compiling, the computer's usefulness ends and the human element picks up again, as Adams and his sociologists then sift through the patterns to see what elevates conflict between police and protesters. In the end, he hopes to be able to show police departments what specific actions can de-escalate violent situations.
Adams describes his methodology, currently a hybrid of algorithmic and human analysis, in Researchers to Crowds to Algorithms: Building Large, Complex, and Transparent Databases in the Age of Data Science. From the conclusion (excerpted for brevity):
We offer RCA and Text_Thresher as enabling technologies researchers may deploy to capture, analyze, and interpret our world in all of its complexity [...] And we look forward to the day when [...]algorithms can collect and refine data from text automatically [...].

[...] We imagine a society where outcomes understood to result from “individual choices” or “ineluctable forces” — once they may be quantified in broader multi-level and temporal contexts — can be shown to result from situational and interacting factors, allowing policymakers to more appropriately calibrate solutions to the level at which human challenges emerge. [...]
Data Science -- in a nutshell -- is about using technology and statistics to tame intractably large bodies of data, extracting information and drawing conclusions from aggregations of text and/or instrument readings that are too large for one person or a group of researchers to analyze manually.

Adams and his colleagues aim to use these methods to understand how and under what circumstances the interactions of protesters and police become violent. To the degree they are successful, they will add scientific authority to conclusions drawn by experienced police and politicos. They may also surface patterns of interaction that haven't been identified before.

Quelling conflict as a limit on activist effectiveness?

From an Associated Press article of 20 Aug, NYC took quick precautions after in-custody death, have a look at what NYC Police Commissioner William Bratton has to say following Eric Garner's death-by-chokehold, about containing protest by calibrating government's response to it (emphasis added):
The relative calm in New York followed a carefully calibrated response by city and police officials intended to neutralize possible unrest. The response drew on the lessons from other high-profile use-of-force cases involving black victims that roiled the city in the late 1990s.

"What you want in a democracy is the ability to express your concerns, but you don't want it to spill over into disorder," Police Commissioner William Bratton said in a recent interview with The Associated Press. "I think we've had a very informed and reasonable response to the issues raised by everybody. There's been no violence."
I'm not sure whose democracy Bratton is referring to. In the democracy I live in, Eric Garner is dead. So is Michael Brown. And John Crawford. And Dante Parker. And Ezell Ford. And Trayvon Martin. And Oscar Grant. The list is all but endless. At what point is the informed and reasonable response praised by New York's Police Commissioner insufficient to address crises on scales like the plague of police killings his department and others are inflicting on our communities?

Here's the thing. Sometimes disorder is exactly what's needed to effectively push against the forces arrayed to maintain a status quo.

Consider the forces arrayed to ramp up the militarization of police and to maintain the criminalization of black skin  -- or of poverty, if you find a more thoughtful truth in positions about class warfare and disenfranchisement argued by Kareem Abdul-Jabbar (in Time Magazine) and Jelani Cobb (in The New Yorker), among others. How broad, how deep, and how long would we have been talking about Michael Brown's death if the people of Ferguson hadn't resisted the militarized response to their legitimate grief and despair viscerally and -- yes -- violently?

It's a counterfactual: one could argue that there's no 'true' answer to that question. But the question isn't new.

One historical example is the role Alabama's "Bloody Sunday" played in sparking national outrage and leading to President Johnson's statement, in March 1965:
Americans everywhere join in deploring the brutality with which a number of Negro citizens of Alabama were treated when they sought to dramatize their deep and sincere interest in attaining the precious right to vote.
... and then to introduce the Voting Rights Act to Congress later the same month.

It wasn't the first time that Civil Rights Movement activists and their leaders had crossed a boundary set by police and government authorities to contain, neutralize, or negate their 'orderly' protest; drawn police violence in response; and, after disorder and suffering, advanced the CRM's goals. Here is the Rev. Martin Luther King, Jr. on these dynamics, excerpted from his Letter from a Birmingham Jail (April 1963):
Before closing I feel impelled to mention one other point in your statement that has troubled me profoundly. You warmly commended the Birmingham police force for keeping "order" and "preventing violence." I doubt that you would have so warmly commended the police force if you had seen its dogs sinking their teeth into unarmed, nonviolent Negroes. I doubt that you would so quickly commend the policemen if you were to observe their ugly and inhumane treatment of Negroes here in the city jail; if you were to watch them push and curse old Negro women and young Negro girls; if you were to see them slap and kick old Negro men and young boys; if you were to observe them, as they did on two occasions, refuse to give us food because we wanted to sing our grace together. I cannot join you in your praise of the Birmingham police department.

It is true that the police have exercised a degree of discipline in hand[l]ing the demonstrators. In this sense they have conducted themselves rather "nonviolently" in public. But for what purpose? To preserve the evil system of segregation. Over the past few years I have consistently preached that nonviolence demands that the means we use must be as pure as the ends we seek. I have tried to make clear that it is wrong to use immoral means to attain moral ends. But now I must affirm that it is just as wrong, or perhaps even more so, to use moral means to preserve immoral ends. Perhaps Mr. Connor and his policemen have been rather nonviolent in public, as was Chief Pritchett in Albany, Georgia, but they have used the moral means of nonviolence to maintain the immoral end of racial injustice. As T. S. Eliot has said: "The last temptation is the greatest treason: To do the right deed for the wrong reason."
"Power concedes nothing without a demand" is one of Frederick Douglass's well-known aphorisms, taken from an address on West India Emancipation given in August 1857. But that's not all he said on that occasion. Here's more (emphasis added):
The whole history of the progress of human liberty shows that all concessions yet made to her august claims have been born of earnest struggle. [...] If there is no struggle, there is no progress. Those who profess to favor freedom, and yet depreciate agitation, are men who want crops without plowing up the ground. They want rain without thunder and lightning. They want the ocean without the awful roar of its many waters. This struggle may be a moral one; or it may be a physical one; or it may be both moral and physical; but it must be a struggle. Power concedes nothing without a demand. It never did and it never will. Find out just what any people will submit to, and you have found out the exact amount of injustice and wrong which will be imposed upon them; and these will continue till they are resisted with either words or blows, or with both. The limits of tyrants are prescribed by the endurance of those whom they oppress.
Disorder is a core element of social and political evolution (and devolution too). Politics are messy.

Information and state control

Some years before Facebook started mining what more than a billion of us 'like', Yale anthropologist and political scientist James C. Scott, in Seeing Like a State (1998) traced the history of permanent / inherited surnames, standardized weights and measures, population censuses, city planning, and scientific agriculture among a host of measures that render people and activity legible to -- and therefore governable by -- the modern state. From Professor Scott's introduction:
... much of early modern European statecraft seemed similarly devoted to rationalizing and standardizing what was a social hieroglyph in to a legible and administratively more convenient format. The social simplifications thus induced not only permitted a more finely tuned system of taxation and conscription, but also greatly enhanced state capacity. They made possible quite discriminating interventions of every kind, such as public health measures, political surveillance, and relief for the poor.

These state simplifications, the basic givens of modern statecraft, were, I began to realize, much like abridged maps. They did not successfully represent the actual activity of the society they depicted, nor were they intended to; they represented only that slice of it that interested the official observer. They were, moreover, not just maps. Rather, they were maps that, when allied with state power, would enable much of the reality they depicted to be remade. Thus a state cadastral map created to designate taxable property-holders does not merely describe a system of land tenure; it creates such a system through its ability to give its categories the force of law.
That is to say, those who wield state power have long history and deep practice of using information about citizens as an instrument of social and political control.

Sure, if information -- from the Berkeley research, for example -- is available to all, anyone can use it to advance their agenda. But only the state is empowered to join state power to information, and bring it to bear with the force of law.

Police departments function as hierarchies. Command and control are built-in. Yes, there are rogue cops and the unpredictability of on-the-ground events. But overall, rank-and-file police follow their sergeants' orders, who respond to their lieutenants, etc., up the chain of command.

Activists? Unity and discipline occur among crowds of protesters, certainly, sometimes. But, overall, not so much as in police organizations.

So when I think about whose goals will be advanced through coordinated application of an understanding of behavioral tendencies brought to light by sociological analysis, my gut tells me that the Deciding Force Project's research will favor the state over grassroots opposition.

On the other hand, the DFP's Nick Adams advances a legitimate argument about enforcing police accountability. From the SF Bay Guardian article quoted above:
But Adams' research isn't just about aiding police forces, it's about holding them legally accountable for esca[la]ting violence, he said.

"You can start to, from a legal standpoint, establish liability with research like ours," he told us. "If we reach out to police departments later on attorneys can hold them accountable for their actions."
Fair enough. Yet it's still up to police and government leaders to decide whether and when to deploy all that Homeland Security weaponry. Are William Bratton's goals representative of what that leadership wants? Again:
"What you want in a democracy is the ability to express your concerns, but you don't want it to spill over into disorder."
If so, I'm not convinced that accountability for instigating police riots will advance -- or be sufficient to retard state obstacles set in the way of advancing -- progressive political goals.

And therefore....?

The die is cast. Data mining is not going away anytime soon, and if Berkeley sociologists weren't doing research in Nick Adams' vein, somebody at the Dept of Homeland Security would be doing it off the public radar. It's probably a safe bet that DHS is on the case independently of the Deciding Force Project. And if it were only up to the Feds, it's likely that police would be coached on the lessons gleaned from their research, while activists would be left in the dark.

So how should activists think about development of techniques that better enable police to maintain order when push comes to shove, in circumstances where disorder is what's needed to push against state power? How can progressive activists employ those techniques in the service of our political goals?

I don't know yet. But I don't think these are questions we ought to ignore.



Related posts on One Finger Typing:
UC Berkeley's anti-apartheid movement: setting the record straight
When authorities equate disobedience with violence
The Occupy Movement and UC Berkeley's Free Speech Monument


Thanks to Loavesofbread for two photos of the Ferguson, MO protests uploaded to Wikimedia Commons; and to Steve Kaiser via Wikimedia Commons for the image of a police officer "applying" pepper spray to sitting protesters at the WTO protests in Seattle, 1999.