Does anybody really know what time it is?

Time is such a contradictory parameter. It is at the core of most scientific analysis and engineering design, we can measure and slice it with femtosecond precision, and we effortlessly take it for granted. After all, everyone knows what time—and the time—is. Yet, no one really understands it. Scientists and philosophers have for thousands of years debated its meaning (Reference 1). Can it go backward? Can you travel through time and, if so, change the past and thus the present and future? What does "now" mean? What does it even mean to say that time can slow down? When Einstein conjectured, as a consequence of his relativity theory, the now-proven fact that time does indeed slow down as you move faster, the time-related questions and answers became more puzzling.

Beyond the broad question of the meaning of time, a new problem with its definition is facing the scientific community (Reference 2). Atomic clocks at the NIST (National Institute of Standards and Technology, www.nist.gov) specify the legal and technical standard for the second (Reference 3). Yet, every few years, scientists must "correct" this definition because a second is also a fraction of the time it takes for the earth to rotate once on its axis: 24 hours equals 86,400 seconds. Because we can so precisely measure time and the Earth's position, we know that this rotation is slowing down due to the influence of the moon, global displacements, and other factors. The correction requires adding a leap second to official time to synchronize the atomic-clock-based time with global motion.

And that's the problem. Adding this leap second is neither trivial nor free of unintended consequences for systems that use precise clocks, especially embedded systems. Well-documented cases show that the leap-second addition causes large-scale system crashes, losses of synchronization, and many other problems. For these reasons, the United States has proposed to, by 2007, standardize the definition of a second as atomic, not astronomical. Over time, astronomical- and atomic-based time would diverge slightly, though enough to cause problems with precision applications.

This seemingly simple proposal sparks controversy. The astronomy community points out that this change would make the tracking of heavenly motion and objects inaccurate. It would even adversely affect the tracking of launched satellites and guided missiles. The astronomy community has a deservedly large and historical presence at the ITU (International Telecommunications Union, www.itu.int). The ITU directs the IERS (International Earth Rotation and Reference Systems Service, www.iers.org). IERS, in turn, is the keeper of the official astronomical clock and defines Greenwich Mean Time. The ITU also tells the IERS what the change of definition for "second" should be and makes a strong case for the proponents of the change. Meanwhile, one could also make a strong case for the other view, which favors atomic-clock time.

No easy answer exists for resolving this issue. Both sides have good arguments. Our ability to precisely measure time, independently of the sky, conflicts with the measurements we make with our telescopes. One is neither better nor more correct, because how you view the conundrum depends on your perspective, definition, and application. Our tangible world of sun and sky clashes with the invisible world of atomic motion, which leads to some difficult and somewhat unsatisfactory decisions. Perhaps it's another manifestation of the conflict between our clock-driven schedules and a sunrise- and sunset-driven life (Reference 4).