While working as a law enforcement officer, I was exposed to a number of officer safety training materials that described nonverbal cues indicative of an impending assault. As a young officer I tried to memorize these cues and observe for them when dealing with people on the street. I found, however, that every time I read another book on officer safety there was a new, and different, list of nonverbal cues to watch for. I was also discovering that watching for these cues and reacting to them with harsher verbal commands and a defensive body posture, was making people more hostile towards me and generating citizen complaints. Therefore, when I became a researcher I placed nonverbal cues related to violence on my list of topics to study.

My first step was to find all of the existing scientific research on interpersonal social cues related to aggression. After all, most of the officer safety training materials I had previously read stated that these nonverbal cues were based on scientific research findings. I was surprised to learn that no such research existed. I was able to find only three studies that examined interpersonal social cues related to hostility and verbal aggression, and all of these three studies dealt with early elementary school children. There was no published research on this topic. I then examined more than 200 books, lesson plans, and websites related to officer safety or self-defense that discussed interpersonal cues that indicated a person was about to make a physical attack. These materials revealed more than 30 different interpersonal cues, some of which were contradictory. For example, some stated that pacing or fidgeting was a sign of impending attack, while others warned that when the person became still he was about to attack. Some suggested avoiding eye contact was a danger sign, while some of the same materials also said staring one in the eyes indicated an impending attack (Johnson & Aaron, 2013).

An experiment was conducted to explore the accuracy of these various interpersonal cues at predicting an assault. A sample of 178 university students was given a survey. In the survey the students were told to imagine that they were in a heated verbal argument with another person and both of you are becoming very angry. They were then given a list of 23 behaviors they might witness in the other person and were asked to indicate how concerned they would be that the person was about to become physically violent. For each behavior they were asked to respond on a scale from one (no concern at all) to seven (extremely concerned). The 23 behaviors were selected from the behaviors cited most often in the officer safety materials I reviewed for the study. These behaviors were: frowning, yelling, crying, sweating, increasing respiration, making verbal threats, clenching hands, violating personal space, tensed posture, removing excess clothing, blinking eyes, tense jaw muscles, flushed face, hands on hips, exaggerated hand gestures, stretching arms or shoulders, stretching neck, glancing around, hands in pockets, pacing, staring in the eyes, avoiding eye contact, and taking a boxer’s or fighter’s stance (Johnson & Aaron, 2013).

The responses of the college students revealed that only six of the 23 behaviors caused the majority of the students to be concerned that the person was about to become violent. These behaviors were:

  • Assumes a fighter’s stance
  • Invades personal space
  • Clenches / balls hands into fists
  • Makes verbal threats
  • Looks around the area (for potential witnesses, backup, or escape routes)
  • Tenses jaw muscles

None of the other 17 behaviors made a majority of the university students concerned for their safety. No significant differences were found among the students with regard to race or sex; they all generally held similar views about these cues of impending violence. University students, however, may not have much experience with real violence and may not have accurate views of nonverbal associated with violent attacks. Therefore, the study was repeated with a sample of 129 law enforcement officers, all of whom had previously been assaulted at least once on the job. The officers were given the same questionnaire but were asked to imagine that they were dealing with a verbally hostile suspect during a street encounter (Johnson, 2015).

The law enforcement officers, all of whom had previously been a victim of a real attack, scored almost all of the nonverbal behaviors higher than did the students. In order from most concerning to least concerning, these 11 behaviors are the behaviors the majority of the officers identified as predictive of an attack:

  1. Assuming a fighter’s stance
  2. Invading personal space
  3. Placing one’s hands in one’s pockets
  4. Clenching / balling hands into fists
  5. Making verbal threats
  6. Looking around the area
  7. Head rolls / neck stretches
  8. Tense jaw muscles
  9. Pacing back and forth
  10. Sweating profusely
  11. Stretches arms / shoulders

The majority of these officers also scored the following behaviors as of little or no concern when predicting an attack:

  • Blinking eyes rapidly
  • Avoiding eye contact
  • Placing hands on hips
  • Making exaggerated hand or arm gestures
  • Crying

The officers were inconsistent about the remaining 7 behaviors: breathes rapidly, yells, removes excess clothing, angry expression, stared into eyes, face becomes flushed red, and tenses up whole body. Some thought these were somewhat predictive of an attack, while others did not think they predicted an attack at all (Johnson, 2015).

While these are the opinions of officers who have actually experienced a real violent attack at some point in their career, their responses might also reflect their training as much as their actual experiences. Furthermore, due to tunnel vision or surprise during an attack, officers might not have accurate recollections or perceptions of the predictors of an attack. Therefore, I am currently studying video tapes of real assaults on police officers to try to determine what interpersonal social cues actually correspond with an attack on an officer. Unfortunately this research is still ongoing and it will be several months before I have any solid conclusions to report. In the meantime, however, we can use the perceptions of these veteran law enforcement officers as a starting point for evidence-based officer safety training.


Note:  Court holdings can vary significantly between jurisdictions.  As such, it is advisable to seek the advice of a local prosecutor or legal adviser regarding questions on specific cases.  This article is not intended to constitute legal advice on a specific case.



Johnson, R. R., & Aaron, J. L. (2013). Adults’ beliefs regarding nonverbal cues predictive of violence. Criminal Justice and Behavior, 40(8), 881-894.

Johnson, R. R. (2015). Perceptions of interpersonal social cues predictive of violence among police officers who have been assaulted. Journal of Police and Criminal Psychology, 30, 87-93.

Print Friendly, PDF & Email