Command Sequence in Police Encounters: Searching

for a Linguistic Fingerprint



Julie Vandermay, MA, Minnesota State University, Mankato

Dan Houlihan, PhD, Minnesota State University, Mankato

Liesa A. Klein, MA, Minnesota State University, Mankato

William Lewinski, PhD, Minnesota State University, Mankato

Jeffrey Buchanan, PhD, Minnesota State University, Mankato

The analysis of language and word use is a fledgling area growing rapidly in the field of psychology. Whether the focus of language research is on purported emotions, personality traits, or the context of dialogue, it is apparent that the language we use and the words we choose to use in a given situation can have a significant impact on others. With a general dearth of studies in the area of language use, it is not surprising that language research is almost entirely missing from the field of law enforcement.


Officers are required to issue commands to citizens within the context of their professional duties. Law enforcement officers use commands to demand compliance (suggestive of a motoric or verbal response from the citizen); question or interrogate a citizen; and/or request a citizen to engage in, refrain from, or cease a behavior. Command types used by officers can vary greatly. Further compounding the issue is the notion that within the structure of the command types themselves, there exists a dichotomous relationship between those issued explicitly (alpha) and those issued implicitly (beta). This provides further confusion in the citizen in terms of understanding the officers’ commands and in the range of potential responses expected from the citizen by the officer (Peed, Roberts, & Forehand, 1977). Subsequently, the use of commands by officers may and do produce mixed results, leading to both compliance and increased resistance on the part of the citizen.


It is difficult to estimate the number of violent encounters occurring between law enforcement officers and citizens due to complicating factors in the studies such as population, geographical location, and citizen and officer demographics, among other issues. Nonetheless, few individuals would argue that the use of violence and force in police/citizen interactions is always possible and will likely always be with us. What appears more important than the use of or the justification for the use of force in these situations, however, is the ability of an officer to effectively de-escalate a potentially violent event through means other than the use of force.


Communication and Law Enforcement Officers


Although some police departments support language instruction, surprisingly, most offer no formal language training for patrol officers (Allread, 1999). This is true despite the fact that inadequate communication exchanges between law enforcement officers and the public not only jeopardize the efficiency of procedures but can threaten justice (Gibbons, 2001). Further, a lack of good communication skills can significantly increase an officer’s need to use force and subsequently increase the chance of injury for both the officer and the citizen.


With only a modicum of research in the area, it is difficult to do a diagnostic analysis to determine whether communication problems in these situations are due to honest mistakes, misuse of police power, personality types or behavior of the officer or citizen, or the nature and context of specific encounters.


Police officers are often required to force public compliance (Mastrofski, Snipes, & Supina, 1996). Officers using a more authoritative tone during interactions with the public tend to instigate the operations necessary for compliance behaviors (Sykes & Brent, 1983). However, it is also the case that at least some officers and departments receive training to be empathic toward citizens and deflect abusive remarks with professional behavior, allowing the officer to secure the citizen’s compliance through this means without the use of force (Johnson, 2004; Thompson, & Jenkins, 1993).


Frequently, law enforcement officers attempt to achieve the compliance of citizens simply through the issuance of various commands. Depending upon the situation, these commands may come very early or later in the encounter. There are several command types that have been identified and can be applied to linguistic exchanges between officers and citizens. These command types are regular, indirect, question, interrogation, stop, don’t, negative, and other (Bertsch, 1999). In addition, each of these command types can also be thought of as having an explicit (alpha) and vague (beta) counterpart within each category (Peed et al., 1977). Alpha commands contain descriptive components that adequately relay what type of compliance is being requested. For example, “Put your hands on the hood of your car” requests a specific motor response. Beta commands contain nondescriptive or incomplete components that do not sufficiently relay what type of compliance is being requested. For example, “Move” or “Knock it off” are vague and do not instruct the individual how to comply appropriately.


Although stress is something that affects everyone, law enforcement officers differ in that they are often under intense, negative pressures, including repeated exposure to pain, public hostility, threat of violence, and the constant presence of danger (Samaha, 1997). Operating under these high levels of stress might also have an effect on how officers react to and then communicate with the person they are addressing, subsequently having an effect on the nature and frequency of commands by the officer. It is known that individuals who are in distress choose how to participate in conversations with other people (Trees, 2005), all the while judging what they feel the other person wants and needs from them. Therefore, in order for there to be negotiations resulting in resolution, there must be mutual understanding and persuasion between the officer and the citizen which involves the induction of compliance (Emans, Munduate, Klaver, & Vandevliert, 2003). The psychological and cognitive adaptations that occur in high stress conditions might impair, in both the officer and the citizen, the underlying processes that facilitate this mutual understanding and, instead, generate a higher level of indiscernible commands by the officer.

Officers are expected to make split-second decisions, especially in the use of deadly force (Dunham & Alpert, 2001; Mastrofski et al., 1996) and also while engaging in communications in stressful conditions. To do this, officers have to maintain a significant level of emotional control while simultaneously engaging individuals who are emotionally out of control and/or violent. For, although information must be acquired quickly in order to appropriately address the situation, officers must maintain a measured pace of communication in order to allow the interaction to de-escalate emotionally (Mastrofski et al., 1996). Officers who themselves have lost emotional control cannot do this and, subsequently, lose the “high verbal ground” and the ability to verbally control or influence the citizen.


In further evaluating police communication, several studies have found that a majority of citizen complaints involved inappropriate or disrespectful verbal behavior by police officers (Dugan & Breda, 1991; Lersch, 1998; Reiss, 1971). The predominant problems cited were related not only to the words themselves, but also to how the officers conveyed their message to citizens (Johnson, 2004). It is unknown whether the officers in these studies intentionally engaged in this negative behavior; did so unknowingly because of a lack of insight, training, or ability; or did not have the emotional control to manage their own communication and verbal strategy.


The sparse literature indicates a need for further investigation to allow for a greater understanding of the effectiveness and impact that a law enforcement officer’s language, word usage, dialogue, and command type and subtype have when issued in the context of a citizen encounter. The present investigation evaluated the pattern of command types unique to each individual officer during a violent encounter. The first hypothesis considered the specificity of command types used (whether alpha or beta) in an effort to assess the frequency of explicit and implicit commands from the time of the initial communication between an officer and a citizen. It was believed that the frequency of beta commands would increase closer to the occurrence of the violent event. The second hypothesis involved the frequency of beta subtypes within the dialogue structure of the interaction between an officer and a citizen during a violent encounter as occurring more often than those of alpha subtypes.


Methods Participants


In order to evaluate linguistic patterns and the command types and subtypes used by law enforcement officers in the context of violent encounters with citizens, archival data were analyzed. The data were extracted from two sources: (1) visual and audio recordings taken from “dash cams” in law enforcement vehicles and dubbed onto VHS tapes and (2) recordings (both visual and audio) from a video camera for the television show “‘Cops . . . Shots Fired®” in which all encounters between law enforcement officers and citizens resulted in violence. Each encounter was given a case number to allow for officer and citizen anonymity.


Overall, an archival pool of ten violent encounters between law enforcement officers and citizens was assessed in this study, with the total number of officers evaluated being 14. However, two encounters (and subsequently two officers) were not included in the analysis as the officer(s) involved issued few commands (< 10), leaving inadequate data for determining linguistic patterns.


Furthermore, several other cases also included inadequate data. For instance, there were three responding officers in violent encounter 1004 in which two of the officers uttered fewer than ten commands, leaving one officer evaluated in this case. Encounter 1009 also involved two officers, but only one of the officers issued ten or more commands during the dialogue, resulting in only that officer being included in further analyses. Therefore, the total number of violent encounters analyzed was eight, which included nine individual officers involved in the violent officer/citizen encounters being evaluated.


The officers were representative of law enforcement agencies across the continental United States and were employed by either a police department or sheriff’s department, or were employed as a state or highway patrol officer.




Data was gathered for each law enforcement officer in each of the violent encounters. Citizen dialogue with the officer involved in the encounter was recorded in the same manner as officer dialogue but was not included in the results as the focus of this study was on the linguistic patterns and command types (subtypes) used by law enforcement officers in violent encounters with citizens. The dialogues between the officers and citizens were recorded using a paper-and-pencil method. The officer’s dialogue data, once in text form, was then coded by command type and subtype, allowing for qualitative and quantitative analysis.


Command types and definitions were taken directly from Bertsch (1999) and applied to fit a law enforcement context (see Table 1). Regular commands were orders that were stated directly. Indirect commands were suggestions not in question form (allowing for nonresponse) to respond physically or verbally. Interrogations were statements in question form to which the only appropriate response was verbal. Question commands were statements in question form to which a motor response was expected, even though a verbal response was available but inappropriate. Don’t commands were instructions to terminate an ongoing behavior or a future behavior, generally preceded with the word “don’t.” Stop commands were instructions to terminate an ongoing behavior, generally preceded by the word “stop.” Negative commands were orders consisting of instructions to terminate an ongoing behavior which does not begin with the words “don’t” or “stop.” Other commands were defined as any order that cannot fit with the above categories or a command that can fit into two or more categories at the same time. Each command type was classified as either alpha or beta. Alpha commands were precise, descriptive orders in which a motoric response was appropriate and feasible. Beta commands were orders in which compliance was difficult due to vagueness or interruption.


For the purposes of this study, a violent event constituted any event in which the officer or citizen engaged in physical contact, either by force or will, which could be achieved through direct contact with another individual, the use of a weapon (e.g., firearm, Taser®, night stick, or vehicle), or the use of an object in such a way that it is meant to inflict physical damage to another individual, as well as acts of self-defense or attempts at suicide. All violent encounters were witnessed, and encounters were defined as the initial contact made between an officer and a citizen to the final contact.




The accuracy of dialogue recording and command type/subtype coding was checked by a second rater to determine if there were any discrepancies between the dialogue from the VHS tapes or the DVD recordings and the command codes determined by the primary researcher. Using a percent agreement formula (agreements/agreements + disagreements × 100), there was 98% accuracy in dialogue text and 99% accuracy in command type/subtype coding when a random 30% of the cases were evaluated by the second rater.

The data from the VHS tapes and DVD recordings provided the main source of dialogue and command type and subtype data. The resulting texts were evaluated first by encounter, and then by individual officer (if there was more than one officer on the scene participating in active dialogue with the citizen, each officer was analyzed separately). The data were then further assessed by looking at the specific language and words used by the officer, the command types and subtypes issued during the encounter, as well as the command type and subtype used in both pre-violent and post-violent dialogue.


Violent Encounter 1000


In violent encounter 1000, the officer involved in the dialogue arrived immediately after the violent event, which involved both an officer and a citizen being shot. A total of 29 commands were issued by the officer in violent encounter 1000. Regular commands were the most dominant in this encounter, accounting for 55% of the total commands issued, followed by indirect commands, which were used by the officer in 19% of utterances.


The number of command types and subtypes used in pre-violent and post-violent events for all encounters are listed in Table 2. Of the command subtypes, beta commands were used by the officer 79% of the time. More specifically, regular beta commands were issued the most (51%), followed by the use of indirect beta commands (19%). Overall, beta subtypes were used by this officer in 85% of the commands issued to the citizen.