The trial of Officer William Porter, the first of six officers charged in the death of Freddie Gray, ended in a mistrial last week after the jury heard sharply conflicting accounts about the inner workings of the Baltimore Police Department. The prosecution emphasized the department’s comprehensive and clearly written policies and procedures, while the defense asserted that, in practice, officers ignore the formal rules as a matter of course. The two accounts revealed the limitations of a legal approach to police reform, and the necessity for departments and officers to self-critique their practices.

Often, when another officer provides critical feedback that does not affirm the officer’s decisions, such feedback falls on deaf ears. Comments like “You weren’t there” or “You don’t know what you would have done in that situation” are a common refrain. In some respects, this makes sense. In a seminal Fourth Amendment case, Graham v. Connor, the Supreme Court held that an officer’s use-of-force decisions “must be judged from the perspective of a reasonable officer, rather than with the 20/20 vision of hindsight.” It would not be appropriate or fair, for example, to judge an officer based on information that the officer did not have and could not have had at the time he or she acted. But this sensible limitation has, at times, been used in ways that undermine the police mission of ensuring the community’s safety.

In many departments, officers have developed a pathological aversion to “second-guessing.” There is a pervasive belief that scrutinizing officer’s use-of-force decisions will lead officers to hesitate, exposing them to dangers that swift action might have averted. The result is a reluctance to engage in an in-depth, critical review of incidents in which an officer injures or kills a civilian and resentment when an outsider calls for such a review. That’s a problem. When an incident ends badly, it should be critically dissected to identify what contributed to that result, as is done when an officer is seriously injured or killed. The primary purpose is not to blame an officer, although poor judgment and failures to follow policy and training must be addressed, but to learn how best to avoid a similar situation in the future.

The aversion to what officers derisively refer to as “second-guessing” does not only make officers less receptive to a critique of their actions, it also makes them reluctant to provide their own complete and honest critiques. That, too, is a problem. Informal interactions between officers shape agency culture and affect officer actions at least as much, and often more, than formal rules. But while empowering officers to engage in peer conversations may help in the effort to self-critique, policing culture also needs to be addressed.

Much of the issue made evident in the Baltimore trial stems from a culture in which officers see themselves as soldiers on the front lines in a war on drugs, crime, and terrorism. Military equipment, weaponry, and tactics were introduced to policing in the 1960s in the form of Special Weapons and Tactics teams, but the so-called warrior culture did not become a visible mainstay of day-to-day policing until the massacre at Columbine High School in 1999.

Officers began wearing alternative uniforms, such as BDUs, that created a more military appearance. Police agencies increasingly acquired and deployed armored vehicles, high-powered rifles, and other military equipment. Officers were taught, formally and informally, to view themselves as warriors, seeking out enemies and ready to attack at the first sign of crisis. The warrior culture has since become more ingrained with every sporadic active shooter, act of domestic terrorism, and ambush that takes the life of a fellow officer.

But using these incidents to justify an aggressive, adversarial approach to policing is a dangerous trend. The warrior culture undermines police-community partnerships, public trust, and the level of cooperation that effective policing depends on. The result, therefore, can be increasing amounts of violent crime, the amount of force that officers use, and even police misconduct. In short, such a culture compounds the very problems that officers and police agencies are struggling to address.

In California, the Peace Officer Standards and Training Commission has used a series of PSAs to encourage officers to have what it calls “courageous conversations” about officers’ driving habits, seatbelt use, speeding, and fatigue. The goal is to reduce the number of officers injured or killed in crashes. This concept could be expanded to other areas where peer correction can be a powerful motivation to change. Telling a fellow officer that they created a needless confrontation with a suspect by using an inappropriately hostile tone or that they put themselves into an avoidable use-of-force situation with unnecessarily aggressive tactics may be difficult, very difficult. But as Travis Yates, a commander with the Tulsa Police Department and the director of training for a police training company, is quoted as saying in one of the PSAs,“If these conversations were easy, they would not be called courageous.”

The failure to engage in behavior reviews and the refusal to have these conversations sends the wrong message: that poor behavior or outright misconduct is tolerated or encouraged. Police agencies may sometimes blame a “rogue cop” or a “bad apple” for misconduct, but as Barbara Armacost has observed, sometimes apples go bad because the barrel—the culture of the police agency itself—is defective. Many reports on policing—the Wickersham Commission that investigated policing failures during Prohibition, the Knapp and Mollen Commissions that investigated corruption at the NYPD, the Christopher Commission that investigated excessive force in the LAPD, and so on—have recognized that misconduct flourishes in environments that fail to address it. Ignoring or individualizing problematic attitudes and actions is an all-too-common feature of a culture that resists second-guessing and peer correction, but doing so is the functional equivalent of tacit approval.

It is possible to build a better police culture, one that can set the stage for better police and community relations. At agencies with a culture of accountability, such as the Las Vegas Metropolitan Police Department and the Richmond, California, Police Department, officers use less force, particularly lethal force, without putting themselves at risk. And while training is an important component, other methods could also be adopted.

First, officers and agencies could embrace the concept of the courageous conversation and partake in them. Second, police officers could be trained to recognize the signs of problematic behavior in themselves and other officers, and to intercede before or even as it happens. For example, an officer who is dangerously stressed from a prior call is a liability, not an asset, and another officer’s intervention can avoid a potential tragedy. Third, officers could avoid what former Savannah, George, Police Chief Michael Berkow calls “officer-created jeopardy,” putting themselves in unnecessarily dangerous situations that increase the likelihood that an officer will use force. Fourth, police agencies could implement a real-time feedback loop for police-citizen interactions, so supervisors also have evidence of positive officer behavior. And finally, police agencies could work with their communities to create what Dan Flynn, the chief of the Marietta, Georgia Police Department, calls “A Public Bill of Rights Regarding Police Accountability.”

Law enforcement’s warrior culture has become a problem, and the solution requires changes to the culture of policing itself. The Police Warrior therefore could be replaced with the Guardian Officer. Guardians are protectors, driven to defend their communities from violence and oppression, including the violence and oppression that some communities have experienced at the hands of the police themselves. While they must acquire the skills and capacity to be a warrior in some occasions, their primary focus is to build relationships with the community they serve. Doing so requires officers to acknowledge errors, and to work actively to improve themselves, their agency, and policing as a profession.