Tom Vanderbilt's book "Traffic: Why We Drive the Way We Do (and What It Says About Us)" comes out on Tuesday.

Vanderbilt (right), 40, lives in Brooklyn, N.Y., and drives a 2000 Volvo. The Journal-Constitution's Richard Halicks spoke with him by phone last week about his three years of research and some of his most interesting findings. Here is an edited transcript of that interview.

Q: What is the two-second rule in traffic?

A: This came from someone at Virginia Tech based on an in-car study. They found that the majority of crashes involved distraction at a key juncture in time leading up to the event. You're talking about two seconds. One of my pet peeves in films is characters talking to each other when they're driving. Someone will look at their passenger four or five or six seconds. You can't keep the vehicle between the lines, much less deal with all the other things that are happening on the road. When you consider that it takes, what, a football field to stop at 55 miles an hour, assuming an ideal reaction time, which isn't often the case. If you think of how far we can travel in two seconds. ...

Q: Along that line, I wanted to ask about the role of "expectancy" in traffic. You illustrated this part of the discussion by talking about the gorilla video test.

A: The first time I did it, I of course was one of the more than 50 percent of people who don't see the gorilla. ... In the video, a group of people wearing white T-shirts are standing in a circle passing a basketball. And you're asked to count the number of passes they make. And about midway through this process, a person wearing a gorilla suit comes sauntering up at a fairly slow clip. He may even wave! I can't recall, but he just sort of walks through.

Q: Right through the circle?

A: Yes. And it's not a quick dash. It's almost like someone crossing a street. So this is surprising for a few reasons. No. 1 is that it's right in the middle of the scene that both your eyes are looking at and you don't see. And another thing that's quite surprising is that the gorilla is not the same color as the white T-shirts, and the thinking is that you've been trained to look for something so specific that what's different actually is screened out, rather than sticking out.

This relates to the road in something like motorcycle visibility. You've stopped at a T intersection, let's say, and you're looking left and right and you don't see any traffic, so you pull out. One of the main reasons motorcyclists die or are critically injured is that their right of way has often been violated by a turning car or a car pulling out because they didn't see it.

We mostly see cars, so a motorcycle violates our expectancy of what we were really thinking we would see. This is one thing that opened my eyes —- I guess that's a bad pun —- and this is not news, but it's good to be reminded of just how fleeting our impression is of the world. I've had a situation where I'm in New York streets looking for a parking space, and I become so trained to look for the gap in the parking, I've almost gone through lights. And you think about, "Oh, if someone had walked across at the moment."

Q: Why does it rankle us to see other people get ahead of us in traffic, and why does the other lane always move faster?

A: The first question relates to our everyday experience. Queuing experts say that, strangely, social justice underlies a lot of these issues. What underlies our notion of justice? Why do we get upset when we see someone cheating? It probably depends on our situation. Sometimes, you may feel like a sucker because you got into the "wrong lane."

There is this classic illusion that has been identified through simulations and videotaping of actual traffic —- just because of our forward bias in traffic, we do spend a lot of time —- 90 percent of our time on average —- looking forward. So we're spending a lot of time seeing vehicles pass us. We're less cognizant of how many vehicles we're passing, because of this shifting accordion effect of the way stop-and-go traffic tends to function.

It plays games with our sense of time and space. If you do pick that one car as a benchmark —- and I've found it quite striking —- you can wait 10 or 15 minutes in heavy traffic and then see that car again. It's like that phantom car you thought was long gone, and it's passing you again.

Q: A quote from Chapter 2: "The word accident ... seems to provide protective cover for the worst and most negligent driving behaviors. This in turn suggests that so much of the everyday carnage on the road is mysteriously out of our hands."

A: I'm not alone in suggesting this, far from it. Just recently, I saw a person struck by a car in New York City. They lived —- it was luckily a low-speed collision. I spoke to this person's doctor at Bellevue Hospital here, and he was reminding me of this whole perspective. "I, as a medical professional, just don't like to use that word. It suggests something that's out of our hands."

The language is quite sloppy on this, I find. It's still quite routine to read in newspapers the phrase "drunk driving accident." It's a bit of a sticky thing. Obviously, a lot of drunks do make it home every night. But the ones who don't, under what kind of semantic rubric can this fully be called an accident? To quote one scholarly paper, "some crashes are more accidental than others."

Q: You make the point that in a lot of cases it's a deliberate behavior, whether drinking or simply speeding, that results in a crash. But we still call that an accident.

A: It's difficult to get inside the head of a person at that moment to know whether they were really cognizant of the risks they were undertaking. But to be combining these high-risk behaviors into one. ...

And in this day and age it is somewhat shocking to read of fatalities that occur for non-belted drivers. The statistics in Canada have fallen 50 percent from 1979 to 2004. If you paired up the U.S. numbers, the drop-off was nowhere near 50 percent. It was barely 15 percent, I think.

The number of people who die every year because they're not wearing a belt is shockingly high, in my opinion. So getting back to the use of the word accident, these so-called acts of God out there are somewhat few and far between.

Q: Why did you study ants, and what did you learn from them about traffic?

A: It's hard not to look at a stream of traffic from a plane, from high above, and then to see similar footage on PBS of these coursing ant trails, and then draw an instant parallel, to feel some kind of kinship with that, as strange as it sounds. I was particularly drawn to the Eciton burchellii, the New World army ant, which is one of these raiding, foraging creatures that kind of rebuilds their highway infrastructure every day, and reshapes their commute every day and are just incredibly efficient in doing so.

The secret seems to be utter cooperation. Everyone is working for the queen of the colony.

Q: So what we really need is a traffic queen.

A: Exactly.

Q: You use the example of the salad bar at the Beijing Pizza Hut to talk about congestion pricing. [To quote from the book]: "I watched with some wonder as patrons at the salad bar carefully arranged towering piles of salad on their plates, then carefully walked away with mounts of teetering greens."

A: The salad bar was a one-trip salad bar. If it was all you can eat, there would have been a lot more traffic back and forth. But it was a one-trip, so people were trying to maximize that one trip, and I guess that the analogy might be paying $8 to get to New York. You might think very carefully about how you were going to get into New York, or whether you really need to go to New York.

We like to think of roads as public space and taxpayer-funded, and everyone should have equal access, but in other ways that's sort of a terrible way to run anything. There's no way to distinguish between different times of travel, different ways of travel.

Traffic congestion is a democratic institution. It hits us all. The wealthy person suffers the same as the poorer person. The interesting thing about congestion pricing that has emerged is how people across the political spectrum have responded to it positively. It unites the Milton Friedman right, if you will, with the progressive transportation people on the left. The Milton Friedman types just want a more efficient, faster ride into town, and the left people have environmental and other issues in mind.

In the U.S., I'm just not sure how we're going to get past this democratic notion of the roads. You can be angry sitting in traffic, angry about the delay, but who are you angry at, exactly? But if someone comes along and passes a congestion charge, then you know exactly who to be angry at. The mayor of the city that's charging you. It's politically much safer to just let people sit and stew in congestion.

Q: You talk about Hans Mondermon saying that time and distance in traffic have a lot of influence on how we behave.

A: It gets back to that idea of being surrounded by people you know or you don't know. You don't want to do anything too bad in your neighborhood because word is going to get out and your reputation is going to suffer. People aren't going to be as nice to you if they hear you're a jerk. But if you're driving in someone else's neighborhood, who cares? You'll never see them again. And the highway is sort of the ultimate in anonymous, nasty behavior. Certainly no one lives on the highway, so there's no incentive to act human at all.

Q: This was something that really surprised me —- that the roundabout was so much safer than the four-way intersection.

A: Each week I'm reading about a new U.S. municipality that is installing a roundabout, or thinking about installing a roundabout, and inevitably there's a town meeting, and a harried engineer has to show up with a PowerPoint and lay it all out, and people are instinctively skeptical.

A lot of this is simply geometry. You've eliminated in a roundabout the left turn, which is the most dangerous maneuver in an intersection. Crashes happen in or near intersections in 50 percent of cases in the U.S. You have 32 points of potential conflict in a standard four-way intersection, and in a roundabout there's only 16. There's no such thing as a left turn across another stream of traffic in a roundabout.

Older drivers in particular have trouble judging the gaps in approaching vehicles. And what you often find is people pulling out to make a left turn and misestimating how far away and how fast that approaching car is coming. We somehow think that four-way intersections are not that complicated or unsafe, but when you really analyze the statistics, it's a no-brainer. Roundabouts win the day every time.