This article was published in Scientific American’s former blog network and reflects the views of the author, not necessarily those of Scientific American
Part 1: The Dancer.
It's called the Dancer. La Bailarina. Like some crazed Stravinsky diva, the Condor GL 310 gas grenade bobs and weaves when it lands, moving in random patterns, spewing noxious gas into the air. Its jittery swirls are specifically designed to cover large areas with gas, but also--critically--to prevent protestors from grabbing the thing and chucking it back at the police.
The 202, you see, is a long-range gas grenade--also quite popular--but the 202 lands and just lies there. It can be shot from further away (over distant barricades, for example, or down into metro stations, or into second-floor windows or terraces or hospitals but it gives no chase.
On supporting science journalism
If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.
The 310 dances.
Just weeks after Turkey's protest movement erupted, Brazilians have taken to the streets en masse. In a show of solidarity, Turks waved the Brazilian flag in Taksim square. The Condor company happens to be based in Rio de Janiero, Brazil, and is one of the two major suppliers of "non-lethal technologies" to both the Turkish and Brazilian governments. In a kind of bizarre, capitalist love triangle of tear gas, popular uprising and government repression, Turkey and Brazil have never been closer.
All along Istiklal, the pale, dented corpses of Condor grenades skitter into gulleys and cracks in the pavement, nestle in piles of burned trash: like detritus on a seascape, old shells, broken eggs.
~~~
The first time I saw the Dancer--really saw it, I mean, not just the haze of its after-effects--I was standing with my brother John on a side street of Istanbul. We'd come for a vacation, actually, on the way to the Netherlands to present at a scientific conference. I have friends that teach in universities in Turkey. One with a 5-year-old daughter I hadn't seen in two years.
Pleasantries, I mean. Socializing. And a bit of work. But then Turkey blew up. My brother started reporting for Science magazine. During the day, I worked on my data for the conference. At night I'd walk the city.
It was just getting dark. Maybe 9 pm. We didn't see the police. We knew the protestors had set up a barricade on Istiklal, the main pedestrian shopping street that reaches down from Taksim Square, one of the major arteries of the city's public life. They'd built the barricade from anything they could find--old metal siding, garbage bins, bits of stray lumber, metal ladders--and were holding it against the police in the distance. But from a side street, down a little hill, all you could see was the faint glow of garbage fires and people walking back and forth.
Then the grenade came. First the *pop* sound of a policeman firing his grenade launcher, then the zzzzzzzz as it flew through the air, then the tinny clatter when it hit the top of the hill and started to roll towards us.
The Dancer doesn't start releasing gas on impact--for most gas grenades, it's a time delay after firing. That's why you'll see streaks of gas falling down around civilians on TV news programs--that's usually the Condor 210, another Brazilian number, or the MP-40L, manufactured in Pennsylvania, fired from enough of a distance that the lozenges inside the grenade have started to release into the air. But the Dancer isn't fired from a great distance. Sometimes it'll roll a pinch before it starts to go off. And that's what I remember, in that flash before my brother and I started to run.
For a second or two, it rolled down the little stone street, picking up speed with gravity. Then the gas started. And that's when the stochastic movements kick in. It began dashing this way and that, spewing gas up and diagonally, semi-circles, jagged lines, at one point even pinging off the side of a building. It looked alive. Really alive. It was, to be honest, strangely beautiful.
The novelty quickly wore off--neither of us had gas masks--and the neurons in my brain sent off a cascade of firing, all of which reminded me, "Cat, it's time to run."
What I thought of at that moment--you think of strange things when you're running from a gas grenade near the Bosphorus--was stochastic movement. The rapid saccades of my eyes moving over the visual scene, the trill of my neurons firing--first in areas dealing with vision, then with novelty, then directly into recognition and fear. Underlying all these processes is the "random walk."
****
First named by Karl Pearson at the dawn of the 20th century, the random walk is essentially a mathematical tool that is now one of the most widely borrowed principles from statistics. It's used to model everything from crowd behavior to biological systems and back again, even reaching out to fluctuations in international markets. The idea applied in molecular science is fairly simple: molecules bump into each other--sometimes violently, sometimes gently. All of those countless interactions shape the path of a single molecule in space. That path is called a "random walk." Bobbing and jagging, the path looks cluttered and wild, but it's just how molecules move in space. Every molecule in the human body is, at every moment, buzzing and colliding with other molecules. We are an aggregate of movement. A density in a cloud.
The random walk is simply a statistical formulation of a series of random "steps" on a "walk". The size of those steps matter. For example, if step size is variable in a normal distribution, then you can use a random walk to model fluctuations in financial market data over time. Imagine the value of a stock over time on a graph. The value, for example, of Condor's stock on the international market is dependent on obvious factors--for example, how many of its usual buyers happen to be either actively using their products (and thus need to replenish their supplies) or are anticipating using them in the future.
But just like the weather, predicting the future value of a stock is extremely hard. (If anyone could actually predict the stock market beyond a two-day horizon, that person would be extremely rich.) The best you can do, like weather, is statistical. By looking back over the years, a June day in Istanbul will probably be beautifully warm and breezy. But you could also get a rain storm.
So while it's certainly the case that economists can tie real-world events to fluctations in Condor's stock, their power of prediction falls away to nothing over time. Yes, it's true that last November Condor secured a contract with the Brazilian government for 49 million dollars' worth of rubber bullets, gas grenades, "flashbangs" (sound and light grenades), and pepper spray. (That contract isn't even for the current protests--that's just for upcoming sporting events like the World Cup.) But the day to day fluctuations in the stock value, taken in aggregate, are best modeled with a random walk. That's why stock market graphs actually look a lot like data graphs of eye saccades, or neuronal activation patterns in a brain.
Human behavior, at various scales, from cellular activity to large social groups, can also be approximated by a random walk. But it's not truly random. And for scientists who try to model it--which way a crowd will move, which way a stock will go, the international valuation of the Turkish lira or the Brazilian real--that's the rub. To model, one can assume randomness as a given, but when thinking about what those graphs actually represent, one also has to assume that there's much more going on than you can know.
The purpose of making a model is to predict. Many models rely on the assumption that some random process or mechanism determines human decisions. If you want to predict what will happen in a crowd when a tear gas grenade goes off, you can treat people as particles in a thermodynamic system. And you get a lot of powerful mathematical tools when you do that. You can even predict the rate at which people will flow through passageways, the likelihood of trampling, the likelihood of crush and compression deaths. Diffusion, at its root, is powered by random movement, and many modelers will use a random walk to build their computations.
****
After wandering the winding streets and alleys of Istanbul's Beyoğlu district, we found ourselves back on Istiklal in the middle of the crowd. I climbed up on a tall metal utility box to get a better view. My brother followed. I hung onto a street light for balance. We were maybe three feet up. I wanted to get a handle on what was going on.
The most dangerous thing about a crowd is usually the crowd itself. In a completely open space--say a very flat plane that runs for miles, unbounded on any side--crowds aren't that bad. Trampling isn't usually the problem. Sure, some people trip. But crowds don't usually gather in wide-open spaces. More often, they gather in cities and stadiums. Places that are bounded. Crowds flow like a liquid. They will fill every available space. And if a crowd panics and rushes into an enclosed space that's too tight--a bottleneck, or a police blockade, an alley that's too narrow for them--the pressure from the back can crush the people in the front and on the sides. Horribly, in those moments, no one can even hear the people screaming, because their lungs are compressed.
The biggest advocates of "non-lethal technologies" like tear gas say that the real challenge for governments is controlling large crowds. If you can disperse them before they reach critical mass, lives can be saved. At low doses, by that logic, tear gas saves lives.
Of course, that says nothing about why those crowds might be forming in the first place. Condor's company profile says that their products are an "effective instrument" that are used "without infringing upon individual Human Rights." But it's well known that the Dancer is being "weaponized" in Turkey, Brazil and Syria right now. Canisters are being fired directly at protestor's faces at close range, leaving grotesque wounds as they cut through flesh and bone. People are being heavily gassed in closed spaces. Doctors worldwide are concerned about the long-term lung and nerve damage that may result from inhaling that much gas over time.
Nevertheless, at that moment, my brother and I were more concerned with the crowd than our burning lungs and eyes.
I was the first to see it. The front line. Where the flags stopped, and just beyond, the glare of headlights from water cannon trucks. Maybe half a mile up, some thousands of people between it and us. And suddenly, it changed. The entire front wall of people turned and started running towards us. People in the street behind them also started to run, a ripple, like a shockwave. The start of a stampede. And it was coming in our direction.
I didn't stop to think. I lept from the sill, barely pausing to hit my brother on the arm on the way down, and started sprinting up a side street. A couple of blocks out I heard Bro behind me calling out, "Sis? What are we running from exactly?" He hadn't seen it. Though others had seen me jump, and now they were running too--the way all of Istanbul has learned to run, now, when they see others running, even if it has nothing to do with them. If one runs, you run.
I stopped to catch my breath and looked behind us. Some dozens ran past us. Some stopped, like us, to turn and see what was happening through a narrow opening onto Istiklal.
It was maybe 60 seconds before the rest of the crowd slammed past our alley: a wall of panicked humans, wild-eyed and yelling. And another 60 seconds before we saw the gas.
(to be continued...)