In my December 2007 Test Connections column in Better Software, I discussed the problem of counting bugs, test cases, and other things that are mind-stuff, rather than physically constructed objects. I gave a number of examples, but I now have another compelling one.
I got the same Christmas gift—Steven Pinker’s The Stuff of Thought—from both my mother and my brother-in-law. (I guess they have me figured out.) In Chapter One, Pinker asks a question about the attack (or is it attacks?) on the World Trade Center in 2001. An airplane hit the North Tower at 8:46am. Seventeen minutes later, another airplane hit the South Tower. Now: was that one event or two?
You could argue that this was a single event, since it was part of a co-ordinated plan with a single agenda, organized by a single group. Or you could argue that there were two events here; two different buildings, two different airplanes, two different groups of hijackers, and two different times. Or you could argue that it doesn’t matter—that’s such talk is just nitpicking, or hairsplitting, or mere semantics, and that there’s no value in making such distinction.
As Pinker notes, though, we can put a precise value on such distinctions in this case: $3.5 billion dollars. That’s because Larry Silverstein, the leaseholder on the WTC property, held an insurance policy that paid out a maximum reimbursement of three billion and a half billion dollars for each destructive event. Several courts and several juries have come to different conclusions on the matter. If we use the formula of total insurance paid out (about $5 billion) divided by $3.5 billion dollars, it appears that there approximately one and one-half events that day—and if that doesn’t seem right to you, don’t worry; it doesn’t seem right to me either.
Our words, our ideas, and our systems of measurements are very complex and tangled. If we want to understand something, simple numbers simply won’t do the trick.