His shoes glinting, his arms churning like a drum major's, Usain Bolt high-stepped away from the top sprinters in the world to win the gold medal in Beijing. The 100 meters produced one of the most memorable images of the 2008 Olympics: Bolt looking around for competition but finding none and then raising his arms to celebrate not just the victory but a world record, 9.69 seconds.
A year later, at the World Championships in Berlin, the Jamaican lowered the record again. Glancing to his right but not seeing America's best sprinter, Tyson Gay, anywhere near, Bolt must have realized or at least suspected that he was about to accomplish something extraordinary, because then he looked to his left, for the readout of the automatic timer, and his face registered not surprise or elation but a sort of just-as-I-expected satisfaction: 9.58 seconds.
For the most part, the world record had for many years crept slowly downward, like a slinky toy making its helical way down a flight of stairs. But then, first in Beijing and then in Berlin, Bolt took the stairs in bunches, three or four at a time, and in doing so he invited many to wonder just how fast a person might ultimately run? Where's the limit, the ceiling, and could somebody approach it in August at the London Olympics?
Mathematicians and statisticians have confronted the question of speed limits. Reza Noubary, for example, a professor at Bloomburg University in Pennsylvania, once predicted that the ultimate, or fastest, clocking for 100 meters would be 9.44 seconds or thereabouts.
But if you were to plot the progression in the 100-meters event, starting in 1912, when the International Amateur Athletics Federation (later to become the International Association of Athletics Federations) recognized Donald Lippincott's 10.6 seconds, and were to follow the trend through all 67 records ratified by the IAAF, you would see a distinct curve that suddenly crashes with Bolt.
The recent records simply defy the model and the curve, said Tatsuo Tabata, the director of the Institute for Data Evaluation and Analysis in Japan, in Wired Science. Over a period of 39 years, from 1968 (Jim Hines' 9.95) to 2007 (Asafa Powell's 9.74), the world's best sprinters lowered the record for 100 meters by 0.21 seconds. And then Bolt dropped the record 0.16 seconds in just two years.
So what's possible? How fast can a man run?
In preparation for the upcoming Olympics, Bolt recently recorded a 9.76 clocking in Rome, the fastest time of the year. And he didn't dismiss the possibility that he could lower the record, perhaps even to 9.4, in London.
"If it's possible," he told CNN, "I'll be the one."
But is it even possible? Surely there must be a limit to how quickly the legs can churn before they become twisted in their own turbulence, a limit to how hard the feet can hit the track before they crumble. How fast can a man run?
"I can't give... a good scientific answer," said Peter Weyand, an associate professor of applied physiology and biomechanics at SMU, "and neither can anybody else."
Science isn't very good, Weyand explained, who might be the country's foremost expert in the science of sprinting, at predicting outcomes that lie outside existing boundaries. And clearly Bolt has sprinted beyond the boundaries of expectation, prediction and even understanding.
Science, in other words, hasn't quite figured out how Bolt runs as fast as he does. Running fast isn't simply a consequence of moving the legs quickly.
In his research, Weyand has studied runners of various shapes and sizes, at various levels of performance, from world-class sprinters to recreational athletes. And the research led to a profoundly surprising conclusion: The time required to reposition a limb -- that is, to pick up a foot and put it down again -- is rather constant, about 0.33 of a second.
Although Bolt is the fastest person in the history of the world, he probably isn't moving his legs any faster than would, say, a backup wide receiver at TCU. What, then, gives the lightning bolt its speed?
"What distinguishes the really fast people," said Weyand said, "is how hard they hit the ground relative to their mass."
Hitting the ground so forcefully, Bolt can propel himself forward at a speed that defies prediction. And that, combined with his length -- he's 6-foot-5 -- gives him a stride that has been measured at 2.44 meters.
But it remains a mystery, Weyand said, how Bolt and other elite sprinters can generate so much force. The size and mass of a muscle determine its force. It's "non-intuitive," Weyand said. But when they're hitting their optimum speed -- and Bolt has approached 30 mph -- these elite sprinters have roughly twice the force-generating capacity of most people.
"Their muscle-extension force should be X, but it's almost 2X," Weyand said. How this happens isn't completely understood, he explained, but involves a transfer of energy.
Other factors, which seem forever evolving, also caution against setting a speed limit for man. Shoes, tracks and nutrition, not to mention pharmaceuticals and biotechnologies, could all contribute, quite conceivably to the development of even faster sprinters.
Then there are the burgeoning financial incentives to run faster and faster that attract more and more elite athletes onto the track. Through endorsements, most notably with Puma shoes, Bolt reportedly has become very wealthy.
How fast can man ultimately run?
A prediction would seem impossible. But Weyand said this: "People will continue to run faster."
Some people might even run faster in London, Weyand said, speaking as a "somewhat knowledgeable track and field enthusiast" and former competitor. Bolt seems to be in top form, Weyand said, and another Jamaican, current world 100 champion Yohan Blake, has indicated that he, too, might be capable of lowering a record.
How fast can man ultimately run?
We might not be close to finding out.