Mr. Levitis asked the professor for the full definition of a behavior. She referred him to their textbook, with its promising title, “Animal Behavior.” To his surprise, neither that textbook nor any other reference he consulted bothered to spell it out. “It was assumed that everyone knew what the word meant,” said Mr. Levitis, who is completing his doctorate at Berkeley.

Mr. Levitis decided to ask the people who should know best: working behavioral biologists. The provocative and crisply written results of his quest, carried out with his colleagues, William Lidicker Jr. and Glenn Freund, appear in the current issue of the journal Animal Behaviour. Among the highlights of the report: biologists don’t agree with one another on what a behavior is; biologists don’t agree with themselves on what a behavior is;

And all that money spent on an education. The problem is that behavior is not an exact science, as most people would like it to be. Most people would like everything to be an exact science, it makes everything so much easier.