``You started it -- you pushed me.'' ``You put your books down first -- you wanted to fight.''
Two boys sit well apart from each other, exchanging accusations in the John F. Kennedy Middle School office. Soon they are telling their story to Assistant Principal Sylvia Goff, who makes each listen to the other's version and decides the blame is equal. Both students get a three-day suspension from school as a penalty.
Within minutes this administrator is on the phone on another case, urging that a school-bus driver see to it that two students involved in a fracas the day before are seated as far apart as possible on the ride home that afternoon.
``I do miss teaching,'' says Mrs. Goff, who describes her job, like that of most deputy school-building administrators as ``100 percent discipline.''
As in many junior high schools, fighting has long been this school's major discipline problem. But, unlike many, Kennedy has been making a concerted effort to control and reduce it, with some success.
Kennedy has implemented a data-gathering system that enables school administrators to distinguish between incidents that violate school rules, those calling for discipline, and those requiring police involvement.
The Rockford school district (of which Kennedy is a part) and the school systems of Jacksonville, Fla., and Anaheim, Calif., are part of the National School Crime and Student Misbehavior Project. The project is jointly funded by the Department of Education and the Department of Justice.
Although none of the three has an unusually difficult discipline problem. Each was seen to have good potential for testing new strategies.
Chief technical architect of the research effort is Robert J. Rubel, author of ``Violence and Crime in the Schools'' and director of the Texas-based National Alliance for Safe Schools.
Dr. Rubel has designed a statistical profile for the schools focusing on where, when, what, and how serious aspects of incidents occur, rather than on who was involved. ``Dealing with who did it is a cop problem,'' he says. ``This is pure crime analysis.''
By pinpointing locations where most fighting occurs, administrators can learn where more supervision or other attention is needed. In one Jacksonville school, administrators discovered that persistent trouble in a locker area had to do with the way lockers were placed -- at right angles to rather than lining the wall. The lockers have since been rearranged.
Dr. Rubel hopes the schools will not only develop more effective strategies to deal with discipline problems, but that broad support networks also will be developed. The networks would be made up of teams of students, parents, custodians, and teachers at each building. Combining cooperation with community agencies, the networks can keep criminal assault under control and add to the principal's ``leverage.''
``It's a new way of looking at an old problem,'' says George Aschenbrenner, assistant superintendent for secondary schools in Rockford. ``I think many of us had never really stopped to think that fighting is [not just] two kids duking it out, but that an assault is criminal activity and involves someone who doesn't want to participate.''
By categorizing incidents in the reports they keep, project administrators are forced to differentiate between disciplinary and criminal incidents. For a mix of reasons, including concern about bad publicity, many school officials have long regarded anything happening within their walls as a school problem, calling police only reluctantly. The hope with the federal project is that school officials will seek outside help.
Broadening the circle of those trying to resolve school violence problems, administrators have said they feel relieved from the burden and also get new ideas.
In Rockford's East High School it was a parent on the crime team who had the idea about how to prevent vandalism to a series of wooden doors near the entrance.
The doors had been sent out for refinishing no less than three times. The parent, noticing that one busy hallway featured a student-painted mural that had never been defaced, suggested an art-class contest. The winning panels were put on the doors more than a year ago and have never been touched. ``The key is that the kids did it, and that the other kids know they did,'' says East High Principal Bill Bowen.
Each school or district in the project picked one major topic and tries to involve as many as possible to resolve it. The Anaheim schools all focused on class cutting and tardiness. Rockford's Kennedy focused on fighting.
Although the accumulated statistics are primarily for in-house use, Kennedy Middle School Principal John Costello says they show a 20 percent drop in fighting during the three weeks before and after an all-school seminar on the subject last fall.
Students were encouraged to think about why fights occur and alternatives they might recommend. One suggestion that was instituted, was penalizing fight-watchers who don't disperse at a teacher's request. Mr. Costello and other Kennedy officials say there is now more talking through of arguments, often with a teacher present.
Counselor Kristin Wilson, a member of the Kennedy team, says students often want an adult involved in their disputes. Many fights, like the one she broke up at the cafeteria recently, occur only where the combatants are sure to be seen by peers and stopped by a teacher before things go too far.
``There's a lot of showmanship involved,'' says Mrs. Goff. She says that if the discipline structure at home is strong, school officials usually need make only one call to a student's parents.
Rockford's efforts have been bolstered by a long-term liaison program with the police department and a reward effort piggybacked on the community's crime-stopper program. Dr. Aschenbrenner recalls that when one school announced on its public address system that it would give a $100 reward for information on the loss of two expensive two-way radios, there was a line of 15 students within minutes, and the radios were found within a half-hour. ``I'll never forget it -- it solidified my faith in rewards,'' he says.
The Florida and California experiments, while following the same statistical method, have focused more strongly than Rockford on developing a close personnel network among community agencies, according to Charles Tremper of California's Ursa Institute. Mr. Tremper was the official evaluator of the $600,000 federal project now in its second and final year. ``There's been some remarkable progress,'' he says.