Human Behavior in Military Contexts (2008)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

5 Nonverbal Communication Life-or-death decisions sometimes depend on subtle nonverbal signals: facial expressions, tone of voice, even the distance people maintain among themselves. According to Triandis (1994, in Carnevale and Choi, 2000), the first Gulf War could have been avoided if not for a misinterpretation of nonverbal cues: In January, 1991, James Baker, then the United States Secretary of State, met with Tariq Aziz, the foreign minister of Iraq. They met in an effort to reach an agreement that would prevent a war. Also present in the room was the half-brother of Saddam Hussein, whose role included frequent calls to Hussein with updates on the talks. Baker stated, in his standard calm manner, that the U.S. would attack if Iraq did not move out of Kuwait. Hussein’s half brother heard these words and reported that “the Americans will not attack. They are weak. They are calm. They are not angry. They are only talking.” Six days later Iraq saw Desert Storm and the loss of about 175,000 of their citizens. Triandis argued that Iraqis attend to how something is said more than what is said. He further suggests that if Baker had pounded the table, yelled, and shown outward signs of anger, the outcome may have been entirely different. Other examples abound. In a recent California murder case, for ex- ample, jurors pointed to the defendant’s physical demeanor when justify- ing their death penalty recommendation: “No emotion, no anything. That spoke a thousand words” (Dornin, 2004). Americans’ impression of the famous Kennedy-Nixon debate apparently depended on whether you heard the debate on the radio or watched it on television (Druckman, 2003). 46 

NONVERBAL COMMUNICATION 47 Nonverbal cues also play a key role in more mundane interactions. For example, Prickett, Gada-Jain, and Bernieri (2000) found that judgments formed in the first 10 seconds of a job interview predicted its outcome. When misunderstood, nonverbal signals escalate conflicts, deepen inter- cultural misunderstandings, and undermine leadership and team cohesion. By laying the theoretical foundation for more effective communication, classroom instruction, and organizational processes, nonverbal research ultimately will enhance soldiers’ ability to communicate, persuade, and avoid misunderstandings before they escalate. SIGNIFICANCE OF CONTEXT Nonverbal behaviors are different from language because they (e.g., gestures or facial expressions) rarely carry specific intrinsic meanings. Al- though nonverbal behavior can act as words—for example, the “thumbs up” gesture has a specific, though culturally varying meaning—most non- verbal communication is contextual, less conscious, and it performs a vari- ety of nonlinguistic functions. Nonverbal signals change depending on the social context; indeed, it is often through observing someone’s behavior that people become aware of the contexts: Is this person a leader? A follower? Surrounded by friends or enemies? Through nonverbal signals, people con- vey emotion (see Chapter 6), project power, manage interpersonal distance, modulate the flow of conversation, and construct ideas about how another person’s mind works. These processes largely proceed outside of conscious awareness, thus explaining the perniciousness of social biases that arise from subtle cues, such as a person’s appearance or accent. The automatic and contextual nature of nonverbal communication has been a thorny problem for study and analysis. Researchers have of- ten failed to properly distinguish between the production (“encoding”) and interpretation (“decoding”) of nonverbal signals. For example, early research by Paul Ekman (see, e.g., Ekman and Friesen, 1975) argued for the existence of discrete emotional states on the basis of the finding that widely disparate cultures could correctly identify an expression portrayed by an actor. However, subsequent findings showed that such experiments only assess people’s ability to recognize (i.e., decode) facial expressions, but provide little information about people’s behavior during actual emotional episodes. Indeed, there are well-known differences between the behaviors of actors and people in naturalistic settings (Coats, Feldman, and Philippot, 1999). Actors use stylized or exaggerated displays in an attempt to make their behavior easier to decode. Naturalistic behavior is far more complex and dynamic, and it often involves strategic attempts to mask or modulate nonverbal displays. Furthermore, even when there are reliable cues that 

48 HUMAN BEHAVIOR IN MILITARY CONTEXTS encode a cognitive act, such as deception, observers often attend to irrel- evant cues. In the discussion below, it is important to keep in mind the distinction between three aspects of nonverbal communication: (1) how nonverbal messages are truly encoded, which is important for detecting deception or a person’s true emotional state; (2) how such messages are decoded, which is important for promoting efficient and persuasive communication; and (3) the relationship between these processes, which is important for under- standing the source of biases and cultural misunderstandings. Nonverbal behavior plays an important role in almost any face-to-face encounter; its absence in telecommunications can contribute to errors and misunderstand- ings. Not surprisingly, nonverbal communication research directly affects several areas that are important to the military, as discussed below. LEADERSHIP AND PERSUASION Nonverbal behavior plays an important role in the exercise of social power. Whether it is in formal leadership settings, as when a lieutenant com- mands a platoon, or less formal settings, as when a physician attempts to change the behavior of a patient, nonverbal signals vary dramatically with social role. Nonverbal cues may be valuable in predicting the effectiveness of attempts to exercise social power and influence (Tiedens and Fragale, 2003). In formal leadership settings, much of the research has focused on charismatic leadership and the role of nonverbal signals in conveying a leader’s sense of enthusiasm or confidence (Riggio, 1987). More generally, dominant partners in two-person interactions show distinctive patterns of facial expression, posture, and eye gaze. For example, dominant partners tend to use more relaxed facial expressions and more directed gazes. Outside of formal leadership settings, research has extensively docu- mented the effects of nonverbal behavior on persuasive relationships, par- ticularly the role of rapport and its nonverbal correlates. Cappella (1990, p. 303) states that rapport is “one of the central, if not the central, constructs necessary to understanding successful helping relationships and to explain- ing the development of personal relationships.” Rapport is correlated with characteristic nonverbal behaviors. Tickle-Degnen and Rosenthal (1990) equate rapport with behaviors indicating mutual attentiveness (e.g., mutual gaze), positivity (e.g., head nods or smiles), and coordination (e.g., pos- tural mimicry or synchronized movements). Rapport can be experimentally induced or disrupted by altering these nonverbal signals (e.g., Bavelas, Coates, and Johnson, 2000; Drolet and Morris, 2000), suggesting a causal relationship between such behavior and social effects. The benefits of rap- port are widespread, influencing esprit de corps, success in negotiations (Drolet and Morris, 2000; Goldberg, 2005), worker compliance (Cogger, 

NONVERBAL COMMUNICATION 49 1982), psychotherapeutic effectiveness (Tsui and Schultz, 1985), test per- formances in classrooms (Fuchs, 1987), quality of child care (Burns, 1984), and even susceptibility to hypnosis (Gfeller, Lynn, and Pribble, 1987). There are significant research opportunities for the U.S. military at the intersection of leadership and nonverbal behavior. The rise of network- centric operations has placed increased emphasis on the exercise of leader- ship “at a distance,” and research on nonverbal behavior has implications for the use of communication technology and could inform the design of more efficient command and control systems. Different communication settings (e.g., telephone, email, video link, or face to face) create dif- ferent styles of interaction and influence the content of communications (Parkinson, in press). For example, a reduction of nonverbal cues in email can reduce participants’ feelings of connection with their conversation part- ners (“social presence;” see Joinson, 2003), with the consequence that they show less concern for the emotional consequences of their communication. People tend to be more honest in emails, which can be an advantage in certain settings, but they often use intemperate language (e.g., “flaming”) with negative interpersonal consequences. By understanding the relation- ship between nonverbal cues and communication style, one could poten- tially design communication technology that is best suited to particular operational environments. Research on rapport can inform military training and operations, al- though current findings have to be further translated to the military context of formalized leadership structures, joint teams, and cross-cultural meet- ings. Understanding how leadership and rapport are nonverbally expressed (i.e., encoded) in such contexts could allow trained observers or even the use of automated techniques to better decode and identify tactically relevant information (e.g., the dominant partner in an interaction). Training indi- viduals to produce nonverbal indicators of effective leadership and rapport may have value as well, though basic questions remain about such learning (see below). NEGOTIATION Nonverbal behavior plays a significant role in negotiation and conflict resolution (see Brosig, Weimann, and Yang, 2004; Drolet and Morris, 2000; Frank, Gilovich, and Regan, 1993). For example, a situation in which people see the nonverbal behavior of a negotiation partner can lead to better negotiated outcomes for the former than when that behavior is hidden (Drolet and Morris, 2000). Positive nonverbal information seems to facilitate the establishment of rapport and social bonds and presumably facilitates partners’ understanding of each other’s goals and intentions. Nonverbal behavior can also be exploited for strategic advantage. For 

50 HUMAN BEHAVIOR IN MILITARY CONTEXTS example, feigning anger can sometimes elicit greater concessions, even when the display is recognized as insincere (see van Kleef et al., 2004). Nonverbal cues are often misinterpreted, and these decoding errors increase when negotiation partners are from different groups or cultures, usually with negative consequences for the negotiation process: This outcome may result from systematic differences in nonverbal behavior, but it may also be the result of their preconceived notions of “the other” (see Chapter 2). Nonverbal behavior research can help military negotiators acquire better negotiation tactics and avoid the miscommunications that escalate conflict and undermine successful negotiated outcomes. CULTURAL FLUENCY Although some nonverbal signals seem universal, others differ dramati- cally across cultures, and these differences can contribute to cross-cultural misunderstandings (see Chapter 2). Nonverbal behaviors associated with language (e.g., gestures) can differ considerably, and in some cases the same gesture can have very different meanings. Even something as basic as a smile can be misinterpreted: in Japan a smile is a common indicator of discomfort or embarrassment. Other subtle cues, such as the use of personal space or gaze, can be misconstrued. For example, the more direct body language of Arabs may be interpreted as aggressive by Western observers (Watson and Graves, 1966). People can learn to recognize and compensate for these differences, and there is some evidence that explicit nonverbal training can facilitate the effectiveness of cross-cultural interactions when participants come from different cultures (see Collet, 1971; Garratt, Baxter, and Rozelle, 1981). Yet basic questions about the best ways to train remain unanswered. Moreover, the research has not been focused on issues that are necessarily relevant for the military. For example, the extensive research on cultural differ- ences in nonverbal behavior has emphasized business negotiations, typically between Western and East Asian participants. However, it is important to note that there is no research on using nonverbal behavioral cues to identify someone intending to carry out a suicide bombing or other attack, especially someone of another culture. Research on situations other than business and among many more cultural groups would be an important military investment for the long term. TRAINING AND LEARNING There has been relatively little research on the role of nonverbal com- munication in education settings, and even less is known about how to 

NONVERBAL COMMUNICATION 51 teach nonverbal skills. Perhaps the clearest effects of nonverbal behavior on learning have been demonstrated by research on interpersonal expectations, known as “self-fulfilling prophecies.” This research shows that teacher bi- ases can be clearly communicated to students through nonverbal behavior and eventually affect learning outcomes. For example, in Rosenthal and Jacobson’s (1992) classic experiment, teachers were misled to believe that certain students had higher aptitudes than others. Teachers used different nonverbal cues, as well as different overt behavior towards those students, creating a warmer socioemotional climate and providing them more feed- back and more time to respond. Perhaps not surprisingly, these students learned better, though subsequent research has questioned the generality of this effect. More recently and conclusively, Singer and Goldin-Meadow (2005) have shown that judicious use of gestures by teachers improved their students’ math scores. Despite the power of such communicated expectations, it seems dif- ficult for teachers to mask their biases. For example, Babad, Bernieri, and Rosenthal (1991) found that teachers actually compensated for their biases through their speech and facial expressions, but still “leaked” their biases through their expressive body behavior. These observations have provided some encouragement to those who would like to teach others how to de- code important nonverbal signals, like those associated with deception. In- deed, some progress has been shown in training people to do just that (Cao, Crews, Nunamaker, Burgoon, and Lin, 2004), and there have been some demonstrations that such decoding skill can smooth cross-cultural encoun- ters (see Collet, 1971; Garratt et al., 1981). How nonverbal communica- tion skills can be most effectively taught, particularly in a military context, remains a fundamental question for research. For example, it is possible to teach soldiers to replicate accurately the gestures of another culture, but it is not known if this ability leads to operational benefits. There is the pos- sibility that such attempts may be perceived by others as disingenuous or as mocking the other culture. Furthermore, it is unclear if one should focus on specific knowledge (what a particular gesture means in a given culture), or teach a general awareness that people may have different beliefs and goals (i.e., teach people to be open minded, ask questions, etc.). Recently, there has been considerable interest in the potential of new media and computer technology to overcome many of the challenges in teaching such skills but the effectiveness of such techniques is yet to be determined. This can be addressed by research carried out in the next 5-10 years. Addressing these fundamental questions would go a long way towards the translation of research findings into tangible results for soldiers in the field. 

52 HUMAN BEHAVIOR IN MILITARY CONTEXTS TECHNOLOGICAL OPPORTUNITIES AND CHALLENGES Recent technological developments hold promise for transforming re- search on nonverbal behavior and providing new vehicles to translate this research into practice. For example, research advances in artificial intelli- gence, computer animation, and computational linguistics have enabled the creation of realistic “virtual humans” that can approximate human verbal and nonverbal interaction (Gratch et al., 2002; Swartout et al., 2006). Vir- tual human technology creates the opportunity to transform both the study of nonverbal behavior and the teaching of nonverbal skills. Virtual humans can address one of the many methodological challenges in nonverbal research: demonstrating a causal as opposed to correlational relationship between nonverbal behavior and its presumed social effects. Due to the rapid and automatic nature of nonverbal communication, it is difficult to experimentally manipulate people’s nonverbal behaviors, some- thing that is necessary to show a causal relationship. Although some clever manipulations have been developed to get people to alter their nonverbal behavior—for example, Bavelas et al. (2000) had people listen to a narrator while subtracting by sevens—these kinds of manipulation, while effective in disrupting normal behavior, are inadequate for replacing normal behaviors with believable alternatives. Rather than using human confederates who at- tempt to change their nonverbal behavior and are then quickly perceived as unnatural, virtual humans can precisely and consistently modulate the non- verbal behavior they present to interaction partners. In one study, Gratch et al. (2006) showed that a “listening agent” that gives rapid nonverbal feedback to speakers dramatically increases speaker fluency and engage- ment in comparison with a less responsive virtual character. Such tech- niques have already proven successful in testing theories of communicative efficiency, learning, trust, mood, impression formation, and social influence (Bailenson, Beall, Loomis, Blascovich, and Turk, 2004; Blascovich, 2002). But more than testing theory, virtual humans have the potential to teach nonverbal competencies. A number of systems, some with branching video but increasingly using advanced character animation and game technology, have been developed primarily to teach decoding skills and cultural flu- ency. For teaching decoding skills, several systems have been developed to teach “shoot/no-shoot” decision making to law enforcement officers, and a number of systems attempt to train interrogators on how to recognize deception (e.g., Cao et al., 2004). For cultural awareness, the Tactical Lan- guage System attempts to teach culturally specific gestures in the context of an Arabic language training system (Johnson, Vilhjálmsson, and Marsella, 2004), and Ward (Bayyari and Ward, 2007) attempts to teach nonverbal “active listening” behaviors in Arabic. 

NONVERBAL COMMUNICATION 53 The promise and rapid advancement of this technology can be a two- edged sword. Lush virtual environments and the present hype surround- ing “edutainment” have created enormous interest in rapidly moving the technology to training applications. The danger is that the advances in the underlying technology may outstrip the science of how to use the technol- ogy effectively. Furthermore, the primary driver of this technology, the game industry, is working at cross-purposes to the goal of effective training. By focusing on the goal of entertainment, game characters, much like good actors, emphasize engaging and easily decoded behaviors that are quite different from the way people act in real situations. Using such technol- ogy without care might easily result in “negative training,” in which one performs worse after training than before. It remains unclear how to mimic the rapid, subtle, and interactive nature of human nonverbal behavior and exactly what level of detail is needed to provide effective skills training. MOTION ANALYSIS AND MULTIMODAL DATABASES A major impediment to research on nonverbal behavior is its reliance on coded data. Participants in a study are video recorded, and the result- ing data are laboriously hand-coded for their nonverbal content. Methods for coding nonverbal behavior, such as the Facial Action Coding System or Laban motion analysis require extensive training and multiple coders to achieve reliability. Yet when data are collected, they are rarely shared among research groups—because there are insufficient incentives to do so–and so the research is time-consuming and expensive. Research is needed in order to create tools that rapidly construct mul- timodal databases and to create mechanisms for collecting and distributing multimodal databases, particularly ones that emphasize military-relevant data, to the research community. Research on sensing technology that can automatically detect and characterize nonverbal communication would also mitigate the data collection bottleneck and promote rapid advancement in the fields. Such methods must be sensitive to the rapid and dynamic nature of nonverbal communication, since it is often changes in behavior rather than static poses that convey information. Distributing such databases, whether manually or automatically created, would facilitate rapid advances in nonverbal research. Other fields, such as verbal communication and machine translation, have seen dramatic progress as a result of the wide availability of machine- readable data, which can be analyzed by computational methods. For example, The Linguistic Data Consortium, which grew out of a project funded by DARPA (the Defense Advanced Research Projects Agency), sup- ports language-related education and research and technology development 

54 HUMAN BEHAVIOR IN MILITARY CONTEXTS by creating and sharing linguistic data, tools, and standards. Funding the development of such tools and shared databases, particularly ones that emphasize military-relevant data, presents an opportunity to expand the utility of nonverbal research and to direct it to military applications within the next 5-15 years.