Looking Ahead

When I'm tired, or not paying attention, I tend to look ahead in only the most minimal way, by keeping my eyes fixed on the car ahead of me. That way of looking ahead is sufficient … up until the moment when something out of the ordinary happens and you're left with no time to respond to it. As I said in Via Tailgating, that's happened to me a couple of times in dense traffic, and it is a scary thing. The minimal way of looking ahead is also not ideal in that it makes you a sticky driver. In short, it works, sort of, but is not recommended.

If you want to look at more than just the car ahead of you, well, how about looking at the next car ahead? If the car ahead of you is small, you may be able to see right over it; if it's medium-sized, you can often look through it, that is, through its windshield; and if it's large … well, we'll come back to that later.

A friend of mine once pointed out that looking at the next car ahead is the same thing as a next-nearest-neighbor interaction in condensed-matter physics. Following that analogy, what you really want to do is develop long-distance interactions, that is, learn to look arbitrarily far ahead. For that, I think, you want your eyes' resting position to be at the road's vanishing point.

Speaking of nearest and next-nearest neighbors, Hofstadter once wrote a nice essay about (among other things) nearest-neighbor interactions among letters of the alphabet.

In particular, all a letter “knows” about itself is its predecessor and its successor …

:

There are two distinguished elements, namely, the endpoints a and z. These elements have identities on their own; they are somewhat like royalty. All other letters derive their identities, directly or indirectly, from these distinguished letters. Obviously b and y are like royal viziers, and c and x like vice-viziers.

:

In fact, any letters further in from the tips than c and x are pretty bland, and even those two aren't very exciting.

Amusingly, it's the same kind of nearest-neighbor interactions among numbers that makes, for example, fourth-neighbor interactions not very interesting.

Now, to un-digress, when I mentioned the road's vanishing point, above, I was of course thinking of the Platonic ideal of roads, a straight line extending to infinity. If, however, the road isn't perfectly straight and level, other ways of looking ahead become possible. If the road isn't level, you can look ahead from the hills and get an excellent view of the state of traffic ahead; similarly, if the road isn't straight, you can look ahead around the curves. In fact, looking ahead around curves is a good idea even when there's no traffic, because it lets you see obstacles in the road well in advance.

Speaking of obstacles, did you know that if you're really paying attention, you can dodge potholes, even on an unfamiliar stretch of road? That's pretty obvious, I guess, but it took me a long time to realize it.

I tend to generalize “looking ahead around curves” to include any kind of looking that's not just straight ahead. In that sense, when you're going down an on-ramp, you can look “ahead”—over to the highway—to see how you're going to merge with the traffic. That kind of thinking led me to a great triumph of looking ahead, which only applied at this one particular overpass that preceded a merge. There, I used to look ahead, try to figure out where I was going to fit into the stream of cars, and adjust my speed accordingly.

I mentioned, above, the problem of looking ahead when your view is blocked. You might, for example, be stuck behind a tractor-trailer, or an annoying SUV, and want to know how dense the traffic is, in order to judge how closely to follow. (If the car ahead of you is part of a dense packet, any crash would propagate backward very quickly, and you'd want to stay well back; but if there's a large gap in front of the car, it'd be reasonable to get close.) In so far as the traffic is uniform, you can get a rough idea by looking backward, or at adjacent lanes, but there's really no substitute for the actual information.

So, what can you do? For one thing, you can peek around the other car. Since the driver's seat is on the left side, it's easier to peek to the left, but peeking to the right is also possible, especially if the traffic is slow and you're next to the breakdown lane. Or, if you have a passenger, you can have ver look.

Or, if the traffic is so slow that it's completely stopped, you can peek by just getting out of your car and looking. In that case, of course, you don't care how dense the traffic is, but you can still look ahead to see why it's stopped.

Another possibility is that you can un-ask the question by not letting your view be blocked in the first place.

My favorite answer, though, is that you can look ahead using indirect observation. Sometimes, for example, you can see the shadows of the cars ahead of you, even when you can't see the cars themselves. Sun shadows are good, but you can also use moon shadows or streetlight shadows, or maybe even shadows from the headlights of cars going the other way. Shadows aren't the only possibility, either—I was once able to see under the car ahead of me via reflection in the wet road.

Indirect observation is good not just for seeing into the space obscured by the car in front of you but, in fact, for seeing into any space you can't observe directly, be it over a hill, around a corner, or wherever. I haven't found shadows and so on to be very useful for such observations; what I have found useful is the responses of other drivers. If, for example, you see everyone at the crest of a hill braking, you can bet there's something happening on the other side.

As another example, suppose you're on a two-lane road, and suddenly lots of people start getting into the right lane. That might mean only that they're all going somewhere you're not, and their exit is coming up; more likely, though, it means that they're all local drivers, all responding to some road condition they know about. Maybe the left lane is going to be forced to turn left, or maybe there's construction ahead and the left lane is closed. In either case, following the crowd and merging right is not an unreasonable approach. (It may not be the preferred approach, depending on what you think about how to merge.)

Finally, I sometimes like to think about looking ahead in time as well as space. That's what I was doing in the picture above, in fact. I often think about looking ahead in time when I'm approaching a barrier. I try to predict how the barrier is going to change, so that I can tell if it's solvable, if there's any way for me to get through it without getting stuck. If there isn't, well, that's when I fall back and wait.

Here's another example of indirect observation. One night recently I was stopped at a traffic light, but I couldn't actually see the light because of the SUV in front of me. However, right next to the SUV was a nice shiny car, and I could see the light in a reflection on its surface. I probably only noticed because it was dark out.