Drive around Seattle long enough and the pattern starts to stand out: new traffic-calming bumps have appeared in many parts of the city over the last year, including places that do not always look like obvious candidates. Some are substantial and unmistakable. Others are small enough that many drivers barely notice them unless a tire catches the raised section.

Part of the confusion is that Seattle is using more than one design. Traditional speed humps run across the full street and are common on smaller residential roads. Speed cushions use cutouts so fire trucks and other emergency vehicles can pass through with less delay, and they are more often used on arterials.

The minimal appearance of some newer installations reflects that accommodation for emergency vehicles. Unlike a full-width hump, a cushion is not necessarily intended to give every vehicle the same jolt. Ideally, regular drivers still stay centered, pay attention and keep their speeds down. The cutouts leave a path for emergency vehicles to pass without the same delay.

Results appear to differ from street to street, leaving the cushions' effectiveness in dispute. On some streets, they appear to slow the flow enough that the whole line of traffic moves more carefully. On others, the gaps seem wide enough that larger vehicles can straddle them, or drivers drift toward the center line or bike lane to avoid the raised sections altogether.

Rather than whether Seattle should calm traffic, the question is now which tools actually do the job. There is support for raised crossings, lane narrowing, cameras and larger humps as alternatives. The smallest cushions draw the most skepticism when they look cheap, sit alone in the roadway or seem easy to bypass.

Many of these projects have been planned for years and are now reaching construction. The buildout reflects the city's continued reliance on street design to slow drivers on corridors where speeding has remained a problem for people walking and biking, even as questions persist about the smaller cushions.