Rorschach blot
Lately, I've been thinking about the human predisposition for perceiving meaning in patterns.
The week before last, I gave a presentation at a local middle school about climate change. Before my talk, one of the teachers came up to me to ask if I thought that global warming could have triggered the earthquake. He whispered conspiratorially, "There have been a lot of earthquakes lately!" I said no, because (1) I couldn't think of a plausible mechanism for that to work in Haiti* and (2) there are perfectly good tectonic reasons for the Haiti earthquake, and (3) there haven't been more earthquakes lately than usual, at least not in any significant way.
As I type these words, the first snowflakes are falling in the second major snowstorm to hit Washington, DC, in less than a week. The first, dubbed Snowmageddon or the Snowpocalypse, dumped more than 18 inches of snow on the capital city. The city does not respond well to that amount of snow: schools shut, the government closed, and there were insane panicked runs on groceries at the area's stores. The new storm, dubbed Snoverkill, adds insult to injury with another 6 to 16 inches predicted. The schools I teach for, NOVA and GMU, have both been closed since last Friday. The way it's looking now, I'm not going to be working again until this coming Friday -- a full week lost due to the white stuff.
What's interesting to me is how people react to the snow. I mean this on two levels: one is the obvious fact that both the culture and infrastructure of Washington, DC, are quite poorly equipped to deal with a couple feet of snow. However, that's not as interesting as the way the snow serves as a reflection on people's mental states. Some people look at the two big snowstorms coming back to back and say, "Where's global warming now?" Others look at the same two storms, with fear that this is the new paradigm like The Day After Tomorrow, and say, "See what climate change hath wrought?" The thing is, they are both wrong.
Weather is not climate, even when two weather events occur in the same week. Tamino calculates that you need about 14 years of temperature data to tease out the long-term trend. Nine years isn't enough, and one week isn't enough. Yet people notice the "clumping" of these data: "two storms in one week! That's a pattern! It must therefore be significant! If it's significant, it's therefore reflective of a common cause, and that common cause is the one I have already decided to be true. Therefore these two storms are evidence of global warming / global cooling**" (depending on who you're listening to).
The thing is, data are clumpy. There is going to be noise in the data. When it comes to snowstorms, the noise is the weather. The noise is superimposed upon a longer-term trend. That trend is the climate. Likewise with earthquakes: no one expects earthquakes to occur with a periodicity regular enough to set to music. Earthquakes of a certain size have a certain probability of occuring in a given period of time, but there's no guarantee that one will occur. If you calculate the average number of earthquakes during a given period of time, and then compare any period of time to that average, your comparison time period will either have more earthquakes than the average or less earthquakes than the average. Ditto for snowstorms: some winters will be more snowy. Some winters will be less snowy.
There are exceptions, such as earthquakes triggered by other earthquakes, or a common cause. The recent earthquake storm at Yellowstone is sufficiently constrained in time and space to suggest that it indeed is a group related by a common cause (though it does not show a magmatic signature, another case of people seeing what they want in the data). Haiti's aftershocks play the same game, though not every earthquake that occurs in Haiti is necessarily connected to the big shock.
Anyhow, this stuff has been on my mind this week. The world is one big Rorschach blot, and humans see what they want in it. We are psychologically all too likely to jump from pattern recognition (noticing clumps in data) to conclusions (often pre-determined), without taking the time to really analyze whether there is in fact a trend present, and if that trend is significant, and if there is a logical causal mechanism to explain that trend.
We are a species that seeks meaning: some augur the future from tea leaves, while others engage in science and reason. Some methods of gaining access to meaning are themselves meaningful. Others are meaningless. Both surround us.
Stay warm out there.
________________________________________
* Though potentially up in Greenland it could work due to glacial melting causing unloading on the crust, changing the stress field that's keeping a fault locked in place.
** It's probably also worth pointing out that snow does not equal temperature. It's precipitation. People are visually susceptible to the sight of snow: it registers more than numbers.
Labels: analogies, climate change, dc, movies, nova, primary structures, science and society, snow




































.jpg)
.jpg)
.jpg)


























