How the disaster starts does not matter: ¹It could be a plane crashing into the World Trade Center, ⁵it could be the sea receding rapidly ahead of an advancing tsunami, it could be smoke billowing through a nightclub. Human beings in New York, Sri Lanka and Rhode Island all do the same thing in such situations. They turn to each other. They talk. ¹⁵They hang around, trying to arrive at a ¹⁰shared understanding of what is happening.

            ¹⁶When we look back on such events with the benefit of hindsight, this apparent inactivity can be horrifying. "Get out now!" we want to scream at those people in the upper stories of the South Tower of the World Trade Center, as ⁶they ¹¹huddle around trying to understand what caused an explosion in the North Tower at 8:46 on a Tuesday morning in September. ¹⁷"You only have 16 minutes before your exit will be cut off," we want to tell them. "Don't try to understand what is happening. Just go."

            ²Experts who study disasters are slowly coming to realize that rather than try to change human behavior to adapt to building codes and workplace rules, it may be necessary to adapt technology and rules to human behavior.

            For all the disaster preparations put in place since the attacks of Sept. 11, 2001, the behavior of people confronted with ambiguous new information remains one of the most serious challenges for disaster planners.

            ¹⁸Computer models ¹²assume that people will flow out of a building like water, emptying through every possible exit. But the reality is far different. People talk. They confer. They go back to their desk. They change their mind. They try to exit the building the way they came in, rather than through the nearest door.

            Building engineers at the World Trade Center had estimated that escaping people would move at a rate of more than three feet per second. On Sept. 11, 2001, said Jason Averill, an engineer at the National Institute for Standards and Technology who studies human behavior during evacuations, people escaped at only one-fitfh that speed. Although the towers were only one-third to one-half full, the stairwells were at capacity, he said. ³Had the buildings been full, Averill said, about 14,000 people would probably have died.

            That is because the larger the group, the greater the effort and time needed to build a common understanding of the event and a consensus about a course of action, said sociologist Benigno E. Aguirre of the University of Delaware. If a single person in a group does not want to take an alarm seriously, he or she can ¹³impede the escape of the entire group.

            The picture of what happened on Sept. 11 is very different from ¹⁴conventional assumptions about crowd behavior, in ⁷which it is assumed that people would push each other out of the way to save their own lives. In actuality, ⁴human beings in crisis behave more nobly - and ⁸this could also be their undoing. ¹⁹People reach out not only to build a shared understanding of the event but also to help one another. In so doing, they may delay their own escape. This may be why groups often perish or survive together - people are unwilling to escape if someone they know and care about is left behind.

            This may be why in fire disasters, Aguirre said, entire families often perish. "The most important factor for human beings is our affinitive behavior," he said. "You love your child and wife and parents; ⁹that is what makes you human. In conditions of great danger, many people continue to do that. … People will go back into the fire to try to rescue loved ones."

Adapted from the Washington Post Monday, September 11, 2006; Page A02

 "Had the buildings been full,... about 14,000 people would probably have died." (ref. 3) means the same as: