Excerpts from Must Read Books & Articles on Mental Health Topics
Books, Part XXII

Psychiatric Interviewing: The Art of Understanding
Shawn Christopher Shea, M.D.

Chapter 3- Nonverbal Behavior: The Interview as Mime

"And now a dark cloud of seriousness spread over her face. It was indeed like a magic mirror to me. Of a sudden her face bespoke seriousness and tragedy and it looked as fathomless as the hollow eyes of a mask."
Herman Hesse, Steppenwolf

In this chapter we will explore the intricate processes known as nonverbal behavior. Few studies are more intriguing or more pertinent for the clinician. Our study will include not only body movements but also those elements of verbal communication that are concerned not with the content of spoken word but with how the words are spoken. Indeed, the noted social scientist Edward T. Hall has commented that communication is roughly 10% words and 90% a "hidden cultural grammar." He continues, "In that 90% is an amalgam of feelings, feedback, local wisdom, cultural rhythms, ways to avoid confrontation, and unconscious views of how the world works. When we try to communicate only in words, the results range from the humorous to the destructive."(1)
     The practical relevance of Hall's words can be readily seen in the following clinical vignette. During an afternoon of supervision, I had the opportunity to watch two interviewers interact with the same patient in back-to-back interviews. The patient, a male in his early twenties, sat with a slumped posture, his head seemingly pulled to his chest by an invisible chain. His legs were open, and his hands lay resting quietly on his lap. The interviewer was a young woman, who spoke in a quiet but persistent voice. The blending between the two was weak at best, provoking an occasional upward nod from the patient, rewarding the starved interviewer with a momentary scrap of interest.
     When the second interviewer entered the room, an intriguing process unfolded. Within 5 minutes the patient sat more alertly in his chair. Eye contact improved significantly and was accompanied by actual animation, albeit mild, in his voice. By the end of the interview, the conversation was proceeding naturally, and a reasonably good therapeutic alliance had been formed. Both interviewers were relatively young women, both of whom conveyed a caring attitude. One wonders what factors resulted in the clearly more powerful blending of the second interview.
     Some of the answers may lie in the communication channels each of these interviewers used in an effort to engage the patient. The first interviewer spoke in a quiet tone of voice intermixed with numerous nods of her head. Such head nodding frequently appears to facilitate interaction. Unfortunately, visual cues lose their impact if the patient refuses to look at the clinician. In short, her facilitatory efforts were on the wrong sensory channel. To the contrary, the second interviewer spoke in a more lively tone of voice, which appeared to perk the patient's attention. More important, her words were frequently punctuated with auditory facilitators such as, "uh huh" and "go on." In contrast, the first interviewer verbalized few such auditory facilitators. The patient had been stranded in the room, responding with detachment to the clinician's monotone voice. Like the first clinician, the second interviewer also utilized head nodding, but her nods became progressively more effective as the patient met her eyes more frequently.
     This example demonstrates the usefulness of flexibly employing different communication channels depending on the receptiveness of the patient. If the patient's head is down, one can increase the number of facilitatory vocalizations. With a deaf patient, one can increase head nodding. Perhaps more important this example emphasizes the overall influence of the interviewer's nonverbal communication on the patient. It suggests that we may be able to consciously alter our nonverbal style in an effort to create a specific impact on the patient.
     This fact brings us to one of the most important challenges of this chapter. In order to flexibly alter their styles, interviewers must become familiar with the baseline characteristics defining their own styles. From such a self-understanding, flexibility emerges.
     Thus a study of nonverbal behavior provides two distinct avenues of exploration. First, as the opening quotation from Steppenwolf suggests, one can learn an immense amount about the patient by studying nonverbal cues. This aspect of nonverbal behavior is the most commonly acknowledged. Hesse's protagonist quickly perceives his companion's change of affect as "a dark cloud of seriousness spread over her face." Second, as our clinical vignette illustrates, one can discover the impact of one's own nonverbal behavior on the patient and subsequently alter it as deemed appropriate.
     Before proceeding it may be expedient to examine the definition of nonverbal behavior, for this term can have different meanings. In their excellent book, Nonverbal Communication: The State of the Art, Harper, Wiens, and Matarazzo explore some of the ramifications of defining this term.(2) In the first place, it is of value to make a distinction between the terms nonverbal communication and nonverbal sign. Nonverbal communication consists of an actual attempt to communicate a message using an accepted code between an encoder and a decoder.(3) A nonverbal sign does not involve an attempt at communication but represents a nonverbal behavior to which the observer infers a meaning.
     Although developed as a refinement of research theory, this distinction between nonverbal communications and nonverbal signs can be adapted to provide a sound background through which to discuss clinical work. Specifically, in this book nonverbal behavior is viewed as the general category of all behaviors displayed by an individual other than the actual content of speech. In this context, tone of voice and the pacing of speech are also considered as examples of nonverbal behavior.
     This broad category of nonverbal behavior can then be split into two subcategories, nonverbal communications and nonverbal activities. In the first category, nonverbal communications, the patient is using a commonly accepted symbol associated with a specific meaning. An irate football fan "throwing the finger" to the quarterback of the visiting team is displaying a piece of rather vivid nonverbal communication. In the second category, nonverbal activities, the overt behavior does not have a single commonly agreed upon meaning, and the sender may not be consciously trying to convey a message. The act of chain smoking cigarettes would represent a nonverbal activity. This activity may indeed be usefully interpreted by the observer as having a meaning, perhaps indicating anxiety; however, this interpretation is inferred and may be wrong. In short, nonverbal activities may have numerous meanings.
     As clinicians we are interested in attempting to understand the significance of both nonverbal communications and nonverbal activities. It is important to keep in mind that nonverbal activities are generally multiply determined. It seems unwise to begin assuming that one "knows" exactly what any given activity means. In this regard Wiener and associates criticized some psychoanalytically oriented researchers as immediately positing unwarranted unconscious meanings to nonverbal activities.
     Considering this context one is reminded of the old psychoanalytic saw in which the astute clinician observes that the patient is experiencing severe marital discord because the patient is playing with her wedding band. Such interpretations of nonverbal activities are invaluable if kept in perspective. The clinician needs to think about other possible causes of the stated activity. For instance, this patient may be playing with her wedding band because she feels intimidated by the interviewer. She releases her anxiety by playing with objects in her hands. Normally she rolls a pencil back and forth, but because no pencil is available, she twists her ring. Other interpretations may be equally correct. To ignore these other possibilities while assuming the marriage is troubled is to ignore sound clinical judgment. On the other hand, having considered the various possibilities, the experienced clinician may gently probe to sort out which is correct and may indeed uncover marital discord.
     From this discussion, the following general principle emerges. Nonverbal communications are relatively easily deciphered, whereas nonverbal activities should be cautiously interpreted, because more than one process may be responsible for the behavior. This point deserves emphasis because both clinical literature and popular literature sometimes read as if the authors felt that they knew the exact meanings of nonverbal activities. They imply that one can read a person like a book. In a similar vein, the concept of "body language" suggests that nonverbal activities are more codified than behavior actually is.
     A similar element of caution emerges as one surveys the research concerning nonverbal behavior. The body of research appears both vast and promising, but there exist many limitations. Nonverbal interactions are so complex that it remains difficult to successfully isolate variables to study. For instance, suppose research was designed to prove that it was the paralanguage (how the words were said) of the second interviewer in our clinical vignette that directly increased blending. An attempt to isolate this single variable would prove difficult, for a variety of other variables could have had an impact, such as the interviewer's physical attractiveness, the distance between seats, and even the fact that there were two interviews.
     Even when one successfully isolates the relevant variables, the very act of isolation poses serious problems. Nonverbal elements seldom function as isolated units.(4) Instead, the various nonverbal elements exert their influences jointly, making the findings of research based on single channels such as paralanguage or eye gaze somewhat artificial. A different approach, the functional approach, attempts to study the various nonverbal elements as they function in unison.
     These research issues are worth mentioning because it is important for the clinician to realize that little knowledge exists on nonverbal activity that can be called "factual." It is safe to say that this body of exciting research is in its childhood. In this regard the material of this chapter is best viewed as opinion concerning an evolving craft or art. The subsequent material is culled from a variety of sources, including clinical work, supervision, research literature, personal communications, and even popular literature (5) if it seems to shed light on clinical issues.
     The following chapter is divided into two sections. In the first section three of the main categories of nonverbal behavior are briefly surveyed. As with the previous chapter, we shall develop a practical language through which to study the phenomena in question. Specifically, the following three areas are addressed: 1- proxemics (the study of the use of space), 2- kinesics (the study of body movement), and 3- paralanguage (the study of how things are said).
     In the second section we shall adopt a functional perspective, carefully investigating the interplay of these three areas as applied to clinical practice. The broad clinical tasks studied include assessing the nonverbal behaviors of patients, actively engaging patients, persuading and focusing patients, and calming hostile patients.


Edward T. Hall was quoted at the beginning of the chapter. Few people would be more suitable for introducing the topic of nonverbal behavior, because Hall literally coined the term "proxemics." It was in his book The Hidden Dimension that he defined proxemics as "the inter-related observations and theories of man's use of space as a specialized elaboration of culture."(6)
     Proxemics deals with the manner in which people are affected by the distances set between themselves and objects in the environment, including other people. As Hall notes, humans, like other animals, tend to protect their interpersonal territories. As humans move progressively closer to one another, new feelings are generated and new behaviors are anticipated. Hall postulates that people learn specific "situational personalities" that interact with the core traits of the individual, depending on the proximity of other individuals. This set of expected behaviors and feelings can be used by the clinician to improve blending. By observing the patient's use of space, the clinician may even uncover certain diagnostic clues.
     Hall delineated four interpersonal distances: 1- intimate distance, 2- personal distance, 3- social distance, and 4- public distance. With each of these distances different sensory channels assume various levels of importance.
     At the intimate distance (zero to 18 inches), the primary sensory channels tend to be tactile and olfactory. People feel at home with the specific scents they associate with lovers and children. At these close distances, thermal sensations also play a role, especially when making love or cuddling. Visual cues are of diminished importance. In fact, at the intimate distance, most objects become blurred unless specific small areas are focused upon. Voice is used sparingly. Even whispered words can sometimes create the sensation of more distance.
     As one moves to the personal distance (1'/z feet to 4 feet), kinesthetic cues continue to be used but olfactory and thermal sensations diminish in importance. With their decline the sense of sight begins to assume more importance, especially at the further ranges of this interpersonal space.
Upon arriving at the social distance (4 to 12 feet), we have reached the region where most face-to-face social interchange occurs. Touch is less important, and olfactory sensations are markedly less common. This region is the playland of the voice and the eyes. Most conversations and interviews unfold within the range of 4 to 7 feet. At the public distance (12 feet or more), vision and audition remain the main channels of communication. Most important, as people move further and further away, they tend to lose their individuality and are perceived more as part of their surroundings.
     A respect for these spaces is of immediate value to the initial interviewer. In general, people seem to feel awkward or resentful when strangers, such as initial interviewers, encroach upon their intimate or personal space. With this idea in mind it is probably generally best to begin interviews roughly 4 to 6 feet away from the patient. If an interviewer is by nature extroverted, by habit the interviewer may sit inappropriately close to the patient, intruding upon the patient's personal space. Obviously such a practice can interfere with blending and should be monitored.
     It should be kept in mind that patients do not determine a sense of interpersonal space by slapping yardsticks down between themselves and clinicians. As observed by Hall, it is the intensity of input from various sensory channels that creates the sensation of distance. An interviewer with a loud speaking voice may be invading a patient's personal space even when seated at 6 feet. Once again clinicians must examine their own tendencies in order to determine how they come across to patients.
     To emphasize the point that it is sensory input, not geographic distance, that determines interpersonal space, one need only consider the impact of a patient who seldom bathes. Such patients frequently create a sense of resentment, because, in essence, olfactory sensations are supposed to occur only at intimate and personal distances. These patients invade the intimate space of those around them even when seated at a distance. The same principle can explain why even pleasant odors such as perfume can also be resented if they are too strong.
     If a clinician intrudes into a patient's personal space, the clinician can set into motion the same awkward feelings and defenses commonly encountered in elevators. The artificial intimacy created by invading the patient's space results in a shutdown of interactive channels, so as not to further the intimate contact. Like a person in an elevator the patient will avoid eye contact and move as little as possible. The patient's uneasiness may even predispose the patient to decreased conversation. In effect, the clinician might just as well be conducting the interview on an elevator, hardly the image of an ideal office. This "elevator effect" can also occur if the clinician ignores cultural differences.
     Hall's distances were determined primarily for white Americans. These distances may vary from culture to culture. One piece of research found that Arab students spoke louder, stood closer, touched more frequently (7) and met the eyes of fellow conversants more frequently. Sue and Sue relate that Latin-Americans, Africans, and Indonesians like to converse at closer distances than do most Anglos.(8) They go on to describe that when interviewing a LatinAmerican, an interviewer may push away, because the situation may feel crowded. Unfortunately this need for distance by the clinician could be perceived as an element of coolness or indifference by the client. In a similar light, the clinician may make the mistake of immediately feeling that the client is socially invasive, when in reality the client is merely interacting at the appropriate distance for Latin-American culture.
     Race may also play a role during the interview. Research suggests that African-Americans may prefer greater distances than Caucasians.(9) Moreover, Wiens discusses the finding that the sexes of the participants can affect the preference for interpersonal distance.(10) One study demonstrated that male-female pairs sat the closest, followed by female-female pairs. Male-male pairs sat the furthest apart.

Kinesics is the study of the body in movement. It includes "gestures, movements of the body, limbs, hands, head, feet, and legs, facial expressions (smiles), eye behavior (blinking, direction and length of gaze, and pupil dilation) and posture.(11) In short, kinesics is the study of how people move their body parts through space with an added attempt to understand why such movements are made. As a field, it is a natural companion to proxemics. Like proxemics, it had its own avatar of sorts, Ray T. Birdwhistell, who first elaborated his work in 1952 with the book Introduction to Kinesics: An Annotation System for Analysis of Body Motion and Gesture.(12)
     Birdwhistell is an anthropologist and emphasized understanding body movements in the context of their occurrence. He also pioneered the study of videotapes in an effort to decipher the subtle nuances of movement. Through his microanalysis he attempted to define the basic identifiable units of movement. For instance, he coined the word "kine" to represent the basic kinesic unit with a discernible meaning.(13)
     Albert Scheflen, a student of Birdwhistell's, expanded these notions to the study of broad patterns of kinesic exchange between people. In this context Scheflen postulated that kinesic behavior frequently functions as a method of controlling the actions of others.(14) By way of example, hand gestures and eye contact may be used to determine who should be speaking at any given moment in a conversation.
     Kinesics plays a role in all interviews. Specific activities may shut down or facilitate the verbal output of any given patient. Besides yielding information that may help the clinician to foster engagement, the study of kinesics can provide valuable insights into the feelings and thoughts of patients. Freud phrased it nicely when he stated, "He that has eyes to see and ears to hear may convince himself that no mortal can keep a secret. If his lips are silent, he chatters with his fingertips; betrayal oozes out of him at every pore.(15)

The study of paralanguage focuses on how messages are delivered. It may include elements such as tone of voice, loudness of voice, pitch of voice, and fluency of speech.(16) The power of paralanguage is immense and popularly acknowledged. Phrases such as, "It's not what you said, but the way you said it that I don't like," are considered legitimate complaints in our society. One can easily picture John Wayne snarling out such a phrase to some unruly bandit. Moreover, actors and comedians are well aware of the power of timing and tone of voice as it impacts upon the meaning of a statement.
By way of illustration, the phrase "that was a real nice job in there" appears complimentary at first glance. But one cannot determine its meaning unless one hears the tone of voice used in its conveyance. It could be far from pleasant if it was said with a sarcastic sneer by a displeased supervisor following an interview observed via a one-way mirror.
     Besides the tone of the voice, speech is characterized by a number of other vocalizations. Although not words per se, vocalizations can play an important role in communication. One set of vocalizations consists of "speech disturbances."(17) Under the heading of flustered or confused speech, these disturbances include entities such as stutters, slips of the tongue, repetitions, word omissions, and sentence incompletions, as well as familiar vocalizations such as "ah" or "uhm." Such disturbances occur roughly one time for every 16 spoken words. As would be expected, under stressful conditions these disturbances increase significantly. Thus they can serve to warn the clinician of patient anxiety as the interview proceeds.
     There is more to vocalizations than just their appearance or lack of it. Some vocalizations serve to enhance blending, as seen with the frequently used facilitatory statements "uh-huh" and "go on." But once again the way in which these vocalizations are used can significantly alter their effectiveness, as shown in the following vignette.
     The interviewer in question possessed a pleasant and upbeat personality. He was a caring clinician, but he found patients shutting down at times during his interviews. Videotape analysis revealed an interesting phenomenon. As he listened to patients, he frequently interspersed his silences with the vocalization "uh-huh." His "uhhuhs" were said quickly with a mild sharpness to his voice as if chopping off sausages. He also used vocalizations such as "yep" and "yea," also stated with a curt tone of voice.
     The net result was the creation of the feeling that he was in a hurry, wanting just the facts. And that is exactly what his patients gave him. This habit, combined with a tendency to overutilize notetaking, fostered a business-like persona, despite his natural warmness in daily conversation. It was a habit well worth breaking and once again highlights the power of paralanguage.
     Cross-cultural differences also affect paralanguage. Sue and Sue describe the variations in paralanguage that can interfere with the blending or assessment process when working with people outside the clinician's culture. For instance, silences are frequently interpreted as moments when the patient, for conscious or unconscious reasons, is holding back. Silence may also signal that the patient is ready for a new question. At other moments, silence can create a feeling of uneasiness in both interviewer and interviewee.
     But as Sue and Sue clearly state, the obvious may be too obvious.

     Although silence may be viewed negatively by Americans, other cultures interpret and use silence much differently. The English and Arabs use silence for privacy, whereas the Russians, French, and Spanish read it as agreement among parties. In Asian culture silence is traditionally a sign of respect for elders. Furthermore, silence by many Chinese and Japanese is not a flooryielding signal inviting others to pick up the conversation. Rather, it may indicate a desire to continue speaking after making a particular point. Oftentimes, silence is a sign of politeness and respect rather than lack of desire to continue speaking. A counselor uncomfortable with silence may fill in and prevent the client from elaborating further. An even greater danger is to impute false motives to the client's apparent reticence. (18)
Many other cultural subleties exist, but they are beyond the scope of this text. Clinicians frequently working with other cultures should make it a point to understand the cultural characteristics of their clients.

Assessment of the Patient

Sir Denis Hill made the following observations during the 47th Maudsley Lecture in 1972:
     Many experienced psychiatrists of an earlier generation believed that they could predict the likely mental state of the majority of the patients they met by observations within the first few minutes of contact before verbal interchange had begun. They did this from observation of nonverbal behavior-the appearance, bodily posture, facial expression, spontaneous movements and the initial bodily responses to forthcoming verbal interaction.(19)
Sir Denis Hill was concerned that the ability to observe nonverbal behavior astutely represented a skill that had fallen by the wayside. Let us hope this demise is not the case, because experienced clinicians today as much as yesterday need to utilize nonverbal clues throughout their clinical work. The knowledge available today concerning nonverbal behaviors is significantly more advanced than 40 or 50 years ago. It is to this knowledge that we now turn our attention.
     To begin our discussion we will look at another statement by Sir Denis Hill: "An important difference between the disturbed mental states which we term 'neurotic' and those we term 'psychotic' is that in the latter, but not in the former, those aspects of nonverbal behavior which maintain social interactional processes tend to be lost."(20)
     An awareness of these potential deficits in the psychotic patient can alert the clinician to carefully probe for more explicit psychotic material in a patient whose psychotic process is subtle.
Perhaps an example will be useful at this time. I was observing an initial assessment between a talented trainee and a woman in her mid-20s. The patient had been urged to the assessment by her sister and a close friend. Apparently the patient's mother was currently hospitalized with major depression.
     By the end of the interview, the clinician seemed aware that the patient was probably also suffering from major depression or some form of an affective illness. But the severity of the patient's condition did not seem to have registered. Instead the clinician was about to recommend outpatient follow-up. The patient's nonverbal behavior was telling the clinician to take another look. In the second interview, which I performed, the patient disclosed a recent weekend brimming with psychotic terror. She had felt that her long-dead father had returned to the house to murder her. She was so convinced of this delusion that she had shared her secret with several young siblings, not a good idea if one is trying to get baby brother and sister to sleep. Eventually she ran from her house to escape her father's wrath. Even in the interview she could not clearly state that her father's return was an impossibility, although she hesitatingly said she thought it was.
     Let us return to the interview in order to uncover the nonverbal cues that suggested the possibility of an underlying psychotic process. The patient, whom we shall call Mary, answered honestly and appeared cooperative. She displayed no loosening of associations or other overt evidence of thought process disorganization, but she demonstrated some oddities in her communicational style. With regard to paralanguage, she demonstrated long pauses (about 4 to 8 seconds) before beginning many of her responses. This gave her a somewhat distracted appearance as if muddled by her thinking. This effect was heightened by a mild slowing of her speech and a flattening of the tone of her speech as well.
     As we have seen, silences, especially of this length, are generally avoided in daily conversation. Everyday social protocol would ordinarily pressure Mary to answer more quickly. This breakdown in normal communicational interaction was one suggestion that all was not well and represents a disruption of the empathy cycle. Her body also spoke to her internal turmoil.
     Although for the most part she had reasonably good eye contact, there existed protracted periods of time when she looked slightly away from the interviewer in a distracted fashion, whether she was talking or listening. This lack of "visual touching" during conversation is unusual.(21, 22)
Frequently, before beginning to speak, the intended speaker glances away briefly. As he or she looks back, speech will begin. While talking the speaker will frequently look away. But as the end of the speaker's statement is reached, the speaker will look towards the listener. This glance signals the listener that the speaker's message is over. The speaker and the listener glance at each other's eye regions in varying lengths, usually between 1 and 7 seconds, the listener giving longer contact. This complex eye duet was frequently missing with Mary. In depression the eyes are frequently cast downward, but it is the peculiar manner in which Mary tended to stare past the clinician that hinted at the possible presence of psychotic process. As Sir Denis Hill had suggested, Mary had lost some of the nonverbal cues that maintain social interaction.
     Other kinesic indicators of speech pattern have been called "markers of speech. "(23) For instance, hand gestures are generally made as one initiates words or phrases. As the speaker finishes commenting, the hands may tend to assume a position of rest. To keep one's hands upwards, in front of oneself, can indicate that one is not done speaking or will soon interrupt.
In Mary these markers of speech were generally diminished. She sat stiffly with her feet flat on the floor. Her head seemed to weigh her body down as she sat slightly hunched over with her fingers interlocked. She displayed little hand gesturing, leaving the interviewer with the odd sensation that it was not clear when Mary was going to start or stop speaking. Most likely, Mary's lack of movement was an associated aspect of her major depression, but it may also have been a ramification of her psychotic process.
     A more striking nonverbal clue to the degree of Mary's psychopathology lay in her method of dealing with unwanted environmental input, in this instance the questions of the interviewer. Apparently Mary had been concerned for some time that she might be "just like her mother," who was currently in the hospital. In addition, her sister had undergone a psychotic depression approximately 6 months earlier. Mary had been attempting to hide from herself the evidence of her own psychotic process, while the fear of an impending breakdown nagged at her daily. During the interview, as questions directed her back into her paranoid fears, she began to realize the extent of her problems. At this moment she did something out of the ordinary.
     Mary leaned forward slowly, her elbows perched upon the tops of her knees with her head cupped between her hands. In this position her hands literally covered her ears, as if keeping out unwanted questions or thoughts. All eye contact was disrupted. Mary remained in this position for a good 5 minutes, answering questions slowly but cooperatively. She appeared detached from the world around her. This type of behavior has been studied under the rubric of "cut-offs."(24) Cut-offs represent nonverbal behaviors made to dampen out environmental stress. When exaggerated to the degree of appearing socially inappropriate, as was the case with Mary, they may be indicators of psychotic process. Indeed, catatonic withdrawal represents a prolonged and drastic cut-off.
     One must also attempt to compare nonverbal activities to the patient's baseline behavior. Mary was normally a high-functioning secretary and most likely possessed better than average social skills. In this light, her preoccupied conversational attitude, and in particular her prolonged cut-off, represents very deviant behavior for her. A subsequent interview with Mary's friend revealed that Mary had been observed at work sitting and staring at the phone for hours.
     For a moment I would like to elaborate on the issue of cut-offs. We have been discussing dramatic forms of cut-off behavior, which may indicate underlying psychotic activity, but mild forms of cut-off behavior occur routinely in our work with nonpsychotic individuals. These more subtle forms of cut-off are not without meaning and warrant some discussion. Morris (25) described four such visual cutoffs, to which he attaches some descriptively poetic names.
     With the "Evasive Eye," the patient shuns eye contact by looking distractedly towards the ground, as if studying some invisible object. It can create the feeling that the patient is purposely not attending to the conversation and may frequently accompany the speech of disinterested adolescents. In the so-called "Shifty Eye," the patient repeatedly glances away and back again. With the "Stuttering Eye," the patient now faces the interviewer directly, but the eyelids rapidly waver up and down as if swatting away the clinician's glance. Finally, in the "Stammering Eye" the patient once again faces the clinician but shuts the eyes with an exaggerated blink.
     These four eye maneuvers represent nonverbal activities whose meaning may be multiple. They may indicate that the patient at some level no longer wants to communicate. Perhaps a specific topic has been raised that is disturbing to the patient, resulting in a nonverbal resistance. At such moments a simple question such as, "I am wondering what is passing through your mind right now,"
may uncover pertinent material. Such cut-offs may also represent objective signs of decreased blending and movement into a shutdown interview. Exaggerated examples of these cut-offs can also be part of a histrionic presentation and in this sense could also be seen in both wandering and rehearsed interviews.
     Investigators have also looked at the promising possibility that nonverbal activities could provide even more specific diagnostic clues, but at this point the research results remain tentative.(26,27) Moreover, the results appear to be in accordance with what common clinical sense would predict.
Concerning the diagnoses found on Axis I, schizophrenia appears to be accompanied by some distinctive nonverbal behaviors. Studies show that schizophrenic presentations are marked by a tendency for gaze aversion. A flattening of affect with decreased movement of the eyebrows was noted (which could alternatively be secondary to antipsychotic medication). Patients' postures were slumped, and they had a tendency to lean away from the interviewer. Naturally the type of schizophrenia and the stage of the process could significantly affect the type of nonverbal behavior present, emphasizing a cautionary note to these generalizations.
     Depression has also been investigated. Researchers have noted that nonverbal behaviors vary depending on whether one is observing an agitated depression or a retarded depression. In the agitated depression, patients demonstrated "a puzzled expression, grimacing and frowning, gaze aversion, agitated movements, a crouched posture, and body leaning towards the interviewer. Subgroup 2 (retarded depressives) showed some increase in gaze, slowed movements, self-touching, an emotionally blank expression, and a backward lean away from the interviewer.(28) In many respects, these findings have limited usefulness, because they simply seem to confirm the obvious.
     But at a different level, especially with depressive patients, these findings emphasize the importance of nonverbal behaviors as clinical indicators of improvement.(29) The return of routine hand gesturing may herald an oncoming remission even before the patient admits to much subjective improvement. As the clinician becomes more aware of such behaviors as spontaneity of facial expression, smiling behavior, and eye contact, the informal monitoring of such cues to improvement can become a routine element of clinical follow-up.
     With regard to Axis II, less research is available. Consequently, we will emphasize principles derived from clinical observations. Observations made during the first 5 minutes of the scouting period may provide important diagnostic clues. In this sense, these cues can help determine which diagnostic regions to emphasize in the body of the interview, for in the limited time available, it is generally not feasible to explore all areas of Axis II pathology. The following three clinical vignettes illustrate the usefulness of nonverbal activities in suggesting the presence of possible character pathology.
     In the first example, I was observing an interview performed by a psychiatric resident during morning rounds on an inpatient unit. The patient was an adolescent girl with a head of curly lightreddish hair. The interviewer was sitting on a couch in a group activity room. The patient pertly entered the room and promptly plunked down beside the clinician. At first she leaned towards him with her right arm straddling the back of the couch behind his shoulder, but she quickly withdrew the arm. Her final perch was with her right knee up on the couch resting a few inches from the clinician's body.
     In a proxemic sense, she had positioned herself well within the personal distance zone and actually very close to being within the clinician's intimate zone. Her speech was bright and snappy, percolating from a face rich with expressions and playful eyes. All this activity occurred in a matter of a few seconds. The clinician immediately responded by leaning away from the patient and crossing his legs by placing his left ankle over his right knee. This brief territorial excursion by this patient is not a typical initial interaction, even with adolescents who frequently feel more comfortable with "chummier" interpersonal distances. Instead, this type of interpersonal game may be seen in people with underlying histrionic personality traits or borderline personality traits.
     The second patient was a woman in late middle age, with graying hair pulled back in a bun. Before the interview, she had had to wait longer than usual before entering the room. Initially, the clinician gently apologized for the inconvenience with a warm smile on his face. She made cool eye contact. Her lips did not so much as consider returning his smile. She fluctuated between a baseline of mildly cooperative answers, with a reasonably lengthy duration of utterance (DOU), to brusque shut-down remarks.
     A peculiar piece of body movement gradually evolved as she continued with her acerbic tone of voice. She tended to lean back in her chair and gradually proceeded to stretch her legs out in front of her towards the interviewer. The movement was ingeniously slow but as steady as a barge pulling into a dock. As usually happens, the dock was gently bumped by her feet, at which point she did not pull away. Instead, the dock recoiled, with the interviewer quickly tucking his feet beneath his chair.
     Her nonverbal activities may be multiply determined, but one possibility well worth exploring would be underlying passive-aggressive traits. Later historical information from the interview tended to further substantiate this diagnostic hunch.
     The third and final patient carefully orchestrated a relatively unappealing opening gambit. She was a tall woman in her mid-20s with long black hair hanging limply about her body. She was dressed in jeans and a black pullover sweater. Her first noticeably unusual action consisted of reaching over to pull up a chair, which she promptly used as a footstool. She stretched her body out, making herself conspicuously at home. This settling in did not signify the beginning of an easy engagement, because she proceeded to visually cut the female interviewer off throughout most of the interview. She would look down at her hands, frequently using the Evasive Eye movement described earlier.
     All of this display was topped with a convincingly dour facial expression. Concerning paralanguage, she managed to push through her disinterested facial mask an equally disinterested and mumbling voice. Her attitude visibly disturbed the interviewer. She also demonstrated one other nonverbal communication with a set meaning. Specifically, she held her coat on her lap throughout the interview, perhaps communicating an eagerness to leave.
     Her collection of behaviors, all present during the first few minutes of the interview, suggested a variety of personality traits worth exploring later. Her lack of concern for making the interviewer feel more at ease could suggest a possible hint of antisocial leanings. Along similar lines her obvious attempt to display disinterest could be part of the manipulative trappings of a borderline personality or perhaps of a narcissistic personality. And as we saw with our previous example, some passive-aggressive tendencies may be present. Her behaviors in no way prove that she has any of these disorders, but they do provide suggestions of which disorders warrant additional consideration, further highlighting the importance of noting nonverbal behavior.
     Let us now move from away from diagnostic issues, and look at some of the nonverbal clues that may suggest that patients are feeling uncomfortable or anxious. One of the most well-known indicators of increased anxiety remains the activation of the sympathetic nervous system, the system geared to prepare the organism for fight or flight. During the activation of this system a variety of physiologic adaptations occur that can serve as hallmarks of anxiety. The heart will beat faster and blood will be shunted away from the skin and gut to be preferentially directed towards the muscle tissue that is being prepared for action. This shunting accounts for the paleness so frequently seen in acutely anxious people, who look like they have seen a ghost. Saliva production decreases, and the bowels and bladder are slower to eliminate. Breathing rate increases, as does the production of sweat.
     This last sign, increased sweating, reminds me of one of the more striking and humorous examples of autonomic discharge I have encountered. A medical student was doing one of his first physical examinations on a real patient, which can truly be an upsetting experience, as the student frequently feels painfully inept. In this case, the patient was a child about 9 years old, who could be generally classified under the label "brat." As the exam labored onward, with the worried mother looking increasingly fretful, the student began to sweat profusely. As the student leaned over to listen to the child's heart, a bead of sweat fell from his forehead directly onto the child's chest. Being a subtle kid, he immediately looked the student in the eve and in a loud voice said, "What's a matter with you, you're sweatin' all over me!"
     If the poor student was not already uptight, that little proclamation did it. He sheepishly turned to the increasingly upset mother and produced a quick-witted white lie, "Don't worry, I've got a thyroid condition." I know this story all too well because I was the poor panic-stricken medical student. It clearly shows the truth that the autonomic system does not lie. With our patients, subtle signs of anxiety such as sweating, damp palms, and increased breathing rate can help us detect anxiety. If the anxiety represents evidence of poor blending, we may be able to purposely attend to the patient's fears. If it represents the presence of unsettling thoughts, we may be inclined to probe deeper.
     If the sympathetic system is not presented with a chance to actually get the organism into action soon enough, the parasympathetic system may try to counterbalance with a discharge of its own. In these cases, one may find a sudden urge to urinate or defecate, as people frequently feel before public performances or job interviews. If a patient begins a session by immediately requesting the need for a restroom, this may represent a clue to a higher anxiety level than the patient may verbally admit.
     Other good indicators of anxiety are described by Morris under the rubric of "displacement activities."(30) These activities are those body movements that release underlying tension. I recently watched a businessman waiting for a meeting. As he sat in the lobby, he nervously tugged at his tie and picked at his clothes. He then hoisted his briefcase onto his lap and meticulously unloaded it piece by piece, after which he gingerly repacked the case, carefully feeling each object as he delicately reassembled his "peripheral brain."
     These behaviors were accomplishing very little in the way of needed physical functions, but they offered a calming effect of some sort for the businessman. Other typical displacement activities
include smoking, twirling one's hair, picking at one's fingers, nailbiting, playing with rings, twitching one's feet, tugging at the ear lobe, self-grooming activities, tearing at paper cups, and twirling and biting pens. The list could certainly be extended. For instance, Morris points out that serving drinks and holding them in one's hands at cocktail parties probably serve to decrease people's anxiety, as they "have something to do."(31)
     Clinically speaking, displacement activities are worth noting during both the initial interview and subsequent psychotherapy. Each patient seems to display a unique set of displacement activities. Once decoded by the clinician, these activities can be usually reliable indicators of patient anxiety. When suddenly increased, they may represent a more reliable indicator than the patient's facial expression or verbal response that an interpretation was on the mark.
     It is also of interest that anxiety will sometimes display itself not through the appearance of displacement activities but in their conspicuous absence. When engaged in an active conversation, most people will display a normal amount of periodic displacement activities. If these suddenly stop or are not present from the beginning, then the person may be experiencing anxiety. In a sense the person may be trying to avoid mistakes by doing nothing.
     This "still-life response" frequently appears when people are videotaped or interviewed in public. It seems to afflict interviewers even more than patients. Supervisors need to be aware that this response may be more of an artifact than a stylistic marker.
     Another area of interest revolves around facial clues that the patient is visibly shaken or on the verge of tears. I am sure the reader is well aware of the faint quiverings of the chin and glazed quality of the eyes that frequently indicate that a patient is close to tears. But a fact not as well publicized is the tendency for people to demonstrate extremely fine muscle twitches across their faces when stressed. These frequently occur beside the nostrils and on the cheek. In people who demonstrate this tendency, these fine twitches can be extremely accurate indicators of tension.
By way of example, I was working with a young businesswoman during an initial interview. She had been referred to me for psychotherapy. She was attractively dressed with a bright disposition and her speech was accompanied by a collection of animated gestures. When asked to talk about her history, she launched into a detailed review of her life since age 16. Of note was her striking avoidance of any events prior to age 16.
     When asked why she had done this, she responded that she did not know and had not noticed it. I asked her if any aspects of her life seemed different before the age of 16. She commented, "Not really, although I spent more time with my father back then." At that point a few muscle twitches appeared by her left nostril. I commented that I had a feeling she was feeling upset, and she burst into tears. Subsequent therapy revealed a complex and ambivalent relationship with her father and other male figures. Throughout therapy, these faint twitches were a sure sign of tension.
     This issue of tension leads directly to another important aspect of nonverbal behavior, the detection of deception. In one piece of research a group of nursing students were asked to participate in a study in which they would be asked to deceive a person.(32) They were told that gentle deceptions were sometimes needed in clinical work, as when comforting frightened patients. Thus the nurses felt a need to perform well in the testing situation.
     In the research itself the nurses were exposed to two different types of films. Some films were pleasant in nature, such as an ocean scene, and other films depicted unpleasant scenes such as a burn victim and a limb amputation. After seeing the pleasant film segments, the nurses were asked to describe their feelings to the listener. This task was obviously not problematic. But after viewing the unsettling film, in one experimental design, the nurse had to convince the listener that the gory film was pleasant and enjoyable to watch. This task was not so easy. Indeed, it so reproduced the sensation of lying that some nurses dropped out of the study.
     All of these interactions were videotaped. Segments of these videotapes were then shown to subjects, who were supposed to determine from the visual images who was indeed lying. It was an ingenious experiment and represents the foundation work upon which further research on deception proceeded.
     The original researchers, Ekman and Friesen, predicted that subjects would state that while lying they would focus on making their faces "look natural." This prediction proved to be true. The deceivers did attend to their faces more, which suggested that nonverbal activities from the neck down may provide a better lead concerning deception. Interestingly, trained observers could pick up clues of deception from videotaped facial expressions. These microexpressions represent accurate clues but are too difficult to pick up routinely.
     On the other hand, the body of the deceiver had a tendency to betray its own head, so to speak, and further research has substantiated many of these initial findings as described in Ekman's fascinating book Telling Lies.(33) Apparently, changes in below-the-neck movements may be of the most practical significance for accurately detecting deception. Direct communications or emblems, as Ekman refers to them, can sometimes be useful indicators of deceit. Emblems represent nonverbal behaviors that carry a distinct meaning, such as a yes or no head nod or pointing to an object. Just as slips of the tongue may betray hidden feelings, slips of the body can occur. With the nursing students in the above study, many felt a helpless sensation that they were not hiding their feelings well. This feeling of helplessness was sometimes inadvertently conveyed by a shrugging movement.
     When representing indicators of nonverbal leakage, emblems usually appear in part. Thus only one shoulder may partially rise or one palm may turn up during a shrug. Another good indicator that an emblem represents a deceitful mannerism is the display of the emblem in an unusual placement. An angry fist will not be raised towards an antagonist but will quietly appear by the side of the patient.
     Hand gestures, which people make while speaking, have been called illustrators by Ekman, and they tend to decrease when deceit is under way. This decrease is particularly true if the patient has not had time to rehearse the lie and must carefully attend to what is being said. The clinician can monitor behaviors such as those described above while exploring regions in which resistance and deceit may be high. For example, when eliciting a drug and alcohol history from a typically active interviewee, a sudden decrease in associated hand movements may suggest that deception is occurring. Several other studies have also found supportive evidence for the idea that below-the-neck clues are best for detecting deceit on a practical level.(34, 35)
     Besides kinesic indicators of deception, the clinician can look for paralanguage clues that deceit is occurring.(36) For instance, a higher pitch to the voice has been associated with deception as well as emotions such as fear. In a complementary sense, lower pitches have been associated with judgments by observers that the subject is more relaxed and sociable. Another possible clue to deception involves the response time latency (RTL). Deceptive subjects were found to demonstrate a longer RTL and to give longer answers when in the act of deceiving.
     It should be kept in mind that most of the kinesic and paralanguage clues to deception mentioned so far represent nonverbal activities, not nonverbal communications. Thus these behaviors may be multiply determined and do not in any way ensure that the patient is being deceitful. In many cases, they may simply indicate that the patient is feeling more anxious. Each activity must be interpreted in the interpersonal matrix in which it was born. By way of example, one researcher found that an increased latency of response could be interpreted in different fashions. If it was followed by a self-promoting comment, then it was often interpreted as being an indication of deception. On the other hand, if the pause was followed by a self-deprecating comment, it was often registered in the opposite direction as evidence of a truthful remark.(37)
     It is probably best to conclude the discussion of cues of deception at this point. Clearly the research is somewhat tentative, but it suggests that some changes in the baseline behavior of the patient may provide useful hints that deception may be at hand. Two practical points warrant mentioning. First, as the interview proceeds, it is generally a good idea to ascertain the baseline body movements that are typical of the patient. Second, during sensitive inquiries, it is best to avoid notetaking. Notetaking can completely eliminate the ability of the interviewer to observe the subtle nonverbal clues that may be the only warnings of deception.
     In the same sense that nonverbal activities may indicate that the patient may be deceiving the clinician, a variety of important mixed nonverbal messages may be sent to an interviewer. These mixed messages are not necessarily deceptions. Instead, they may represent hallmarks of patient ambivalence and confusion.
     In order to explore this fascinating area, the work of Grinder and Bandler (38) offers a wellspring of practical and sound clinical observation. Although controversy has arisen over their later work, their first two books provide some pioneering insights into engagement techniques.
     Their work follows naturally from the principles we have been discussing thus far. Put simplistically, they state that as a person communicates a message, the message is transferred through a variety of communicational channels simultaneously. The patient's message may be conveyed through the content of the spoken words, the tone of voice, the rate of speech, the amount and type of hand gesturing, the posture, and the facial expression. These messages are termed paramessages. When all paramessages have the same meaning, the paramessages are said to be congruent. But if some of the channels convey discordant information, then the paramessages are said to be incongruent.
     The underlying theory is simple; perhaps that is why it proves to be so powerful therapeutically. People who consistently communicate with an incongruent style can frequently create a confusing impression. Their incongruence may make the people around them feel ill at ease and uncomfortable. If the clinician can detect this selfdefeating interpersonal style, he or she may be able to help the patient modify it. In a more immediate sense, incongruent paramessages may indicate underlying mixed feelings of which the patient is unaware. Once again, the therapist may be able to cue off this incongruence, leading the patient into an exploration of the uncovered mixed feelings.
More germane to the topic of the initial diagnostic interview, episodes of incongruent communication may alert the clinician to areas worthy of more immediate investigation or perhaps regions pertinent to explore in later sessions.
     I am reminded of a woman in her early 30s who I was evaluating for possible psychotherapy or medication. Ms. Davis, as we shall call her, was coping with a variety of stresses, not the least of which was the loss of her mother several months earlier. For years she had been her mother's caretaker and verbal whipping post. Ms. Davis was mildly overweight with stocky legs, offset by a face embraced by a full head of black hair. As she spoke, her conversation turned to her bitter relationship with her boyfriend, who apparently enjoyed her sexually but found marital ceremonies not to his liking. She commented, "I hate him, I'll never go back to him. He's not worth it."
     Harsh words, but one should be wary of taking them too seriously, for Ms. Davis' body spoke differently. The words were spoken with a tone of pained resignation, not biting anger. They had the quality of the child-like pout, "Daddy's not bringing home a present from his vacation." Not only did her voice lack selfindignation, but her hands played a martyr's role. Rather than the more typical pointing and jerking movements of an angry accusation, they were held low towards her lap with the palms upwards. This type of hand positioning is frequently associated with a tone of supplication and need.
     Put more precisely, Ms. Davis was communicating with an incongruent set of paramessages. As Grinder and Bandler point out, all of these messages may have elements of truth to them. In Ms. Davis' case, she certainly did have angry feelings towards her boyfriend, as suggested by the content of her words. But she also had extremely powerful needs to be accepted by him; indeed, these needs bordered on a masochistic willingness to be verbally beaten by him. Her tone of voice and hand gestures suggested her strong need for acceptance. Even her breathing rate did not increase or become more spurt-like, as is frequently seen as someone becomes increasingly angered. This set of incongruent messages was one of the first clues to her deeply rooted problems concerning hostile dependence, which became central working issues in the remaining therapy. Indeed, her relationship with her mother was in reality no different from her relationship with her boyfriend.
     In any given initial interview, periods of incongruent communication may occur. If noted, they can serve as road signs that effectively guide the interviewer towards a deeper understanding of the patient.
     In a similar fashion the work of Scheflen, whom I mentioned earlier, deserves more detailed examination, because it too focuses on the nonverbal interactions that serve as communication scripts for people.(39) Scheflen discusses the idea that humans, like other animals, engage in certain shared behaviors that tend to escalate into specific actions. Such actions include fighting behavior, mating behavior, and parenting behavior. Frequently, these mutually arousing actions serve to eliminate the actual need to engage in the final activity. In such a manner animals will frequently avoid actual combat by undergoing a territorial display of sorts. Scheflen calls such escalating patterns of behavior "kinesic reciprocals."
     Kinesic reciprocals can frequently be seen in clinical interactions. If the patient begins the reciprocal, the clinician may inadvertently continue the process. I have certainly seen this process occur within the realm of the courting or mating reciprocal. I remember watching a videotape of a session of psychotherapy. The patient was a young woman interacting with her therapist, who was a relatively young man with about 7 years of clinical experience. The patient sat pertly forward, cigarette hanging aesthetically from her fingers. The therapist, who was dressed casually in a sport shirt, sat rakishly back, also with a cigarette in hand. Their voices possessed a spritely coyness.
It was unclear whether I was watching the beginning moments of a therapy session or the opening sequences of a grade B movie. In any case, the therapist and his patient were engaging in the courting reciprocal, otherwise known as flirting. Inadvertent participation in such reciprocals can create a variety of problems. Obviously it can stimulate an erotic transference. Moreover, if initiated unconsciously by the therapist and then reciprocated by the patient, it can lead the therapist towards the inappropriate perception that the patient is histrionic.
     I am reminded of one clinician who tended to be pleasantly flirtatious and buoyant with staff. She was surprised when male patients, following initial evaluations, would ask her out. On videotape the answer was obvious in that some of her flirtatious qualities appeared in her clinical work, albeit in a much toned down fashion.
     Scheflen provides a good description of kinesic behaviors utilized by both sexes in the courting reciprocal:

The full-blown picture of the female courting posture is well known to us, for models and actresses simulate it continually in being seductive or attractive. The head is held high and cocked. The "mark" is looked at from the corners of the eyes. The chest is brought out so that the breasts protrude. And the legs appear "sexy" as the foot is extended and the calf musculature is tightened.... An actively courting woman may present her palm, a highly affiliative act, in many ways; e.g., when she pushes back her hair, when she smokes, or when she covers her mouth while coughing. The man's state of high tonus is evident most clearly in the thoracic- abdominal behavior. He moves from a slump, with abdominal protrusion to thoracic display by sucking in his belly and squaring his shoulders. A man may use some of the same behavior in courting that he uses in dominance. He may draw up to full height, protrude his jaw, stand in close, and display what is generally regarded as a masculine stance.(40)
Other reciprocal behaviors besides the courting reciprocal can occur in an initial interview. A striking example was provided by a videotape made of an initial interview for use in supervision.
     The interviewer was a young woman. Across from her the patient sat with eyes occasionally cast downwards. As the interview unfolded, the patient produced a folded piece of paper, and she
asked the clinician to read the paper before proceeding. Her voice seemed to step meekly away from her lips. In the meantime the patient began fumbling with the microphone. She had correctly wrapped it around her neck but had problems attaching it to her blouse. Noticing her problems the clinician looked over and asked if she needed help. The patient did not look up for a moment as she continued to fumble. Then with her head cocked downwards, she innocently glanced upwards shaking her head "yes." She gazed with the helpless eyes of a little girl and said not a word. The clinician promptly leaned over and fixed the microphone.
     The parenting reciprocal had emerged as naturally as if enacted between a true mother and her child. In this brief vignette, the power of the first few minutes of the scouting period to provide clues for further diagnostic probing is once again amply demonstrated. This patient's manipulative style and dependent behavior suggested the possibility of some form of character pathology. Indeed, further interviewing revealed a mixed personality disorder with histrionic, passive-aggressive, and dependent characteristics. Apparently this patient had perfected the art of eliciting parental responses as a method of garnering attention.
     This patient also displayed another type of nonverbal activity, auto-contact behavior. Auto-contact behavior consists of movements involving self-touching.(41)Such behaviors may consist of grooming behaviors, defensive-covering behaviors, and self-intimacies.
     Self-intimacies are defined as, "movements that provide comfort because they are unconsciously mimed acts of being touched by someone else."(42) These self-intimacies appear frequently during interviews. Patients may hold their own hands or sit with their knees pulled up to their faces, arms literally hugging their own legs. In regressed patients, one can see even more extreme forms of selfhugging as patients lay in tightly curled fetal positions.
     With regard to frequency, the most common self-intimacies in order of most to least frequent are as follows: 1- the jaw support, 2- the chin support, 3- the hair clasp, 4- the cheek support, 5- the mouth touch, and 6- the temple support. With hair touching there is a 3:1 bias in favor of women. Temple touching demonstrates the opposite bias with a preference in men of 2:1. Sometimes these kinesthetic comforters can be tied into other sensory modalities as well. I remember one patient who would pull her hair across her cheek. She would simultaneously gently sniff at her hair, which she related as being very comforting. Such activity was a sure sign of her underlying anxiety, much like a displacement activity.
     In this manner these behaviors may serve to alert the interviewer that the patient is feeling pained or anxious. It can cue the interviewer that the patient may need some verbal comforting, perhaps prompting an empathic statement. It can also alert the clinician that powerful affective material is being approached, possibly suggesting the need for further exploration.
     In summary, in the above material the focus has been on the power of the patient's body to convey information to the perceptive clinician. It is now time to explore the reverse situation, those moments when the clinician uses his or her body to affect the patient.

Utilization of Nonverbal Behavior to Engage the Patient

One of the exercises undertaken in our interviewing class concerns the use of seating arrangement. Two of the trainees sit in the middle of the room on easily rolled chairs. They are given a simple task, to situate themselves so that they feel the most comfortable with regard to conversing with one another. In about 90% of the cases, the participants choose a similar position.
     They sit roughly 4 to 5 feet apart. They are turned towards each other but do not quite directly face one another. Instead, they are turned about a 5 to 10-degree angle off the line directly between them, both in the same direction, as shown in Figure 3A. Only about 10% choose to face each other directly.
     If the participants are asked to turn directly towards each other, they complain of feeling significantly less comfortable. Some will even push their chairs back a bit. The discomfort is related as feeling "too close." More specifically, many of the trainees complain that the head-on position forces eye contact, making it difficult to break eye contact without undertaking a significant head movement. This head-on position fosters a sensation of confrontation.
     On the other hand, the preferred position readily allows for good eye contact but also makes it easy to break contact in an unawkward fashion. In my own practice, I have certainly found this position to be the most comfortable and the most flexible interviewing position for me. This last statement is important, because it emphasizes that the most comfortable position may be different for each interviewer and indeed for each interviewing dyad. Each clinician needs to discover a comfortable position, keeping in mind that the clinician must also be willing to alter this position depending on the needs of the patient.
     In addition to the nonconfrontational feeling provided by the position described above, another phenomenon may be enhancing its comfortableness. As discussed before, one of the key processes that enhances blending is the ability of the clinician to convey a sense of seeing the world through a shared perspective.
     If one looks at the actual fields of vision available to each participant in the interview, an important relationship readily becomes apparent. When two people are directly facing each other, the fields of vision exhibit little overlap. What overlap exists lies directly between the two participants. This situation tends to foster the sensation that "You are over there, and I am here." It seems to work against the sensation of "We are here together." On the other hand, when the two participants are turned slightly away from each other, so that they are subtly facing the same direction, then the feeling that "We are here, and the rest of the world is out there" naturally emerges.

     Thus, in a phenomenological sense, the feeling of confrontation is decreased, while the sense of blending is given a gentle boost, as illustrated in Figure 3B. It should be noted that the directly oppositional position may be preferred by some people. Indeed, some clinicians recommend it,(43) but I myself do not, for the reasons provided above.
     The concept of seating raises the more general issue of furniture arrangement. Some clinicians prefer setting, away from their desk, two large comfortable chairs. Another alternative is to utilize the desk creatively. In general, I believe a desk should not sit between the clinician and the patient, because this creates an authoritarian distance appropriate for chief executive officers, not therapists.
On the other hand, the desk can be placed as shown in Figure 3C with only a corner protruding between the clinician and the patient. If the clinician's chair rides on wheels, the clinician can move the chair and alter the resultant interpersonal distance either by increasing or decreasing the amount of desk between the participants. A paranoid patient may require more distance from the clinician, which can easily be accomplished by moving only a short way, because the desk quickly provides a protective barrier. On the other hand, the clinician can easily move to a point where essentially no desk intervenes.
     The overall concept of the clinical setting warrants attention. When designing a private office, an effort should be made to provide a comfortable and professional atmosphere. The office represents an extension of the clinician's persona, and the patient's first impression in the scouting period may be significantly affected by the decor of the clinician's waiting area or office. Calming prints or photographs, accompanied by several diplomas and shelves of books, provide a reassuring and pleasant environment.
     Trainees are faced with limited financial resources. But three or four unframed art posters and a few plants can be bought very reasonably, producing a sometimes startling change in the atmosphere of the room. There is no need for a trainee's room to look like a prison cell. On the contrary, part of the training experience is learning to consider the principles behind creating an appropriate private office.
     Outside the office, situations can be a bit more difficult, because the clinician faces crowded hospital rooms and disorganized emergency rooms. It remains important in these situations to consider the comfort of both the patient and the clinician. While performing a consultation in a crowded hospital room, there is nothing wrong with saying, "Before we start, would you mind if I slide your bed over, so both of us can have more room to talk."
     This discussion of seating arrangements leads to the issue of determining an optimum distance between the clinician and the patient, which will vary for each interviewing dyad. There does seem to exist a small region in which the clinician's presence respects the patient's sense of personal space while still allowing the movements of the clinician to have an immediate impact on the patient. This zone of effective interpersonal space may be referred to as the "responsive zone" (RZ). If the clinician moves out of the RZ towards the patient, then the interviewer risks frightening the patient or creating a sense of discomfort. On the other hand, if the interviewer leaves the RZ by moving too far away from the patient, then the movements of the clinician may have little impact on the patient. For instance, the act of gently leaning forward towards a patient, which can enhance communication during particularly sensitive moments of an interview, may have no effect if done outside the RZ.
     Two examples may help clarify the importance of establishing an RZ that seems most comfortable for each patient. First, if one intuits that a patient may be feeling paranoid, it is useful to remember that such patients may require a larger space around them in order to feel more comfortable. In these cases the RZ is larger and it may be wise to begin such interviews sitting further from the patient than one normally would sit or perhaps using a desk or table to help provide a safety barrier as mentioned earlier. As the interchange proceeds the clinician may find that the distance can be gradually decreased; hence the RZ frequently may change as the blending waxes or wanes.
     In the second example, one looks at the problem of accurately eliciting a formal cognitive examination in elderly patients who are seriously depressed and withdrawn. To attract and maintain their attention, the interviewer might need to sit considerably closer than normal. This more intimate RZ may help decrease the likelihood of obtaining poor cognitive results secondary to the patient's lack of attention or interest. If a patient is not interested in answering, then the risk of getting artificially low scores becomes very real indeed. In such cases the tendency to suspect a real dementia when only a pseudodementia is present can become a true dilemma.
     Another way of obtaining the withdrawn patient's attention during the cognitive examination is to speak more loudly, effectively moving closer but not moving one's chair. At times it is also important to ensure attention by literally asking the patient to look at the clinician as the questions are asked. For instance, the interviewer can gently but firmly make statements such as, "It may help you to do well on these questions if you watch me as I actually say the digits to you." In the last analysis, if a withdrawn patient is looking down at the floor as the clinician performs the cognitive mental status, the validity of the results are certainly questionable.
     The concept of increasing the validity of the cognitive examination also raises the issue of touching patients. Some clinicians seem to have a block against the idea of touching a patient. Although it is not frequent for me to touch a patient during an initial interview (except for handshakes), I sometimes find touching useful and poignant. With regard to the cognitive examination, some depressed and withdrawn patients may ignore the clinician's attempts to make eye contact and attend to the task at hand. In such instances, one can touch the arm of the patient, offering comments such as, "I know it is difficult for you to concentrate right now, but it really is important." At such points, the patient may glance up at the interviewer and more effective contact will have begun.
     Of course, touching, a method of entering the patient's intimate space, as described by Hall, may also be used at points at which the patient may benefit from some simple comforting. I am reminded of a sad, middle-aged man who I interviewed as he was entering the hospital. For all of his life he had been a kind and hard-working mill worker. Unbeknownst to himself he was being exposed to an extremely toxic industrial poison. Over the years he experienced gradual changes in his behavior, including irritability and occasional violent outbursts, which frightened him and produced extreme guilt. Simultaneously he underwent marked changes in his intellectual functioning, to the point that he had problems dealing with everyday activities. Only recently had he learned that his problems were secondary to brain damage.
     As we neared the end of the interview, he told me that he was afraid of the hospitalization because "people say mean things to me, they think I'm stupid. Please let me come in, I promise I won't hurt anybody, I promise, and I'm not that stupid." At which point he began to weep. It seemed only natural to reach over and grasp his arm while reassuring him that I believed what he said and that we would help him make the transition to the hospital.
     Outside of the types of situations described above, touching patients is not common during initial interviews, because touch is a powerful communication, that may carry numerous connotations, not all of which are appropriate. Patients may misinterpret touch as an erotic gesture or at a minimum as a sign of implied intimacy. Although the clinician may intend the gesture as a sign of caring, a psychotic patient or a patient with a histrionic personality may distort the message considerably. Indeed, if a clinician finds a routine need to touch patients during initial interviews, it would be wise for the clinician to determine why such a need is arising. Usually it is not from clinical considerations. Such clinicians frequently have a desire to be perceived as "comforting angels." Ironically, this drive to be perceived as "comforting" may get in the way of effective caregiving. Such self-exploration may also reveal flirtatious traits or histrionic qualities in the therapist.
     At this point we can turn our attention to another aspect of nonverbal behavior, which frequently emerges if the clinician has effectively determined the appropriate RZ for the patient. At such times the appearance of certain nonverbal behaviors can suggest that the blending process is proceeding well. As mentioned in Chapter 1, several verbal signs, such as an increased DOU, may indicate the presence of improved engagement. In a similar fashion nonverbal activities may also be used routinely to monitor the blending process.
     For instance, as blending increases, the patient may begin to make progressively better eye contact, while spontaneous arm gestures and "talking with one's hands" may increase. Along similar lines, if a patient in a shut-down interview begins to talk more with his or her hands, this may be a hint to pursue the present topic more fully in order to further strengthen the engagement process. The clinician can also frequently see the patient turn more towards him or her as blending increases. Relaxation is also shown by an asymmetry in posture, while tense posture is frequently seen with a person who feels threatened.(44)
     We have been discussing the nonverbal activities that may suggest powerful levels of blending. It is important to return to a topic approached earlier, namely, the differences seen cross-culturally. With regard to the African-American culture, eye contact is not considered as important in conveying attention to a listener.(45) Just being in the room or close to the speaker may be considered enough to convey that attention is being given.
     Direct eye contact may be considered disrespectful in certain cultures, such as with Mexican-Americans and with the Japanese. In this context, a clinician could be making a serious error in judgment by interpreting poor eye contact with members of these ethnic groups as an indication of rudeness, boredom, lack of assertiveness, or poor blending.
     Another process that may emerge more frequently when one has successfully found the RZ is the surprising phenomenon of postural echoing.(46) In postural echoing one finds that two people who are communicating effectively tend to adopt similar postures and hand gestures. At a cafe, two lovers may sit across from each other, both heads perched in their hands, as they animatedly stare into each other's eyes.
     A frequent phenomenon seen in interviewing occurs when one member suddenly shifts positions and relaxes. Simultaneously the other person will also shift and relax. Moreover microanalysis of videotapes has suggested that as blending increases, the minute movements of the interviewer and the interviewee tend to parallel each other as if a miniature minuet were being performed. During moments of discordant interchange this reciprocity decreased.
     At one level these findings suggest that the appearance of postural echoing may serve as a clue to the clinician that the blending process is on the right track. In a slightly different vein, the clinician can subtly match some of the patient's postures in an effort to actively increase blending. For example, if a male clinician is interviewing a steel worker who is crossing his legs with his ankle over one knee, the therapist may cross his leg in the same manner, as opposed to crossing his leg at the knees. The latter method could be misconstrued by the interviewee as "feminine." By adopting a style similar to that of the patient, the metacommunication is passed that "we do certain things similarly and we may not be as different as one might first suppose." This discussion of the use of postural echoing, in an effort to actively engage the patient, leads to a consideration of other methods of nonverbally increasing the blending process.

One collection of nonverbal behaviors potentially useful in the art of engagement consists of the so-called affiliative behaviors. Such behaviors include eye contact, smiles, and gesticulations. It has been shown that counselors who demonstrate these behaviors are viewed as significantly more persuasive than counselors who do not.(47) Another commonly encountered affiliative behavior consists of a body lean of about 20 degrees towards the patient.(48)
     One of the most well-recognized affiliative gestures is the simple head nod. Morris makes the interesting observation that the vertical head nod indicates a "yes" or "positive" response in all cultures and groups in which it has been observed, including Caucasians, African-Americans, Balinese, Japanese, and Eskimos. It has been observed in deaf and blind individuals as well as in microcephalic people incapable of speech. He relates that the head nod may convey different types of "yes" messages, such as the following:

The Acknowledgment Nod: "Yes, I am still listening."
The Encouraging Nod: "Yes, how fascinating."
The Understanding Nod: "Yes, I see what you mean."
The Agreement Nod: "Yes, I will."
The Factual Nod: "Yes, that is correct."49
Interviewers should make an attempt to learn the frequency with which they typically head nod. This frequency can vary significantly among interviews. From my own observations it appears that interviewers who are particularly adept at engaging patients tend to head nod numerous times during any several minutes of an interview. As obvious as the utility of the head nod may appear, I have found that approximately 20% of professionals I supervise tend to underuse it. A few barely head nod at all.
     The power of the head nod became apparent to me in an unexpected fashion during a session of psychotherapy. I had been working with a middle-aged male patient for several months. I decided to try a brief exercise in which I would purposely stop my typical head nodding for several minutes, in order to see what this practice would feel like to me. To my surprise I found it difficult to do, because it had become habitual. But more to my surprise, the patient broke off his spontaneous conversation after about 2 minutes and asked, "What's wrong? Somehow I feel that you don't like what I'm saying." This vignette emphasizes the power of nonverbal cues during clinical interaction.

In Chapter 1 we discovered that with guarded or paranoid patients, certain changes in approach could enhance engagement. In particular, certain verbal approaches that were effective with most patients could be potentially disengaging with guarded patients. For instance, guarded patients frequently respond better to basic empathic statements rather than to complex empathic statements. In a similar fashion, with certain patients, the clinician's nonverbal behavior may be too empathic or intimate.
     As mentioned earlier when discussing proxemics, guarded and paranoid patients may appreciate being provided with more space than most other patients. Along these lines, some of the affiliative gestures, when done too frequently, may prove disruptive to the guarded patient. I have heard paranoid patients comment that they have disliked frequent eye contact, perhaps twisting the attentive gaze of the "good listener" into the stark gaze of a potential persecutor. In this context, one may purposely break eye contact more frequently with paranoid patients, providing them with visual space.
     Even head nodding and arm gestures can be unsettling when done too frequently with guarded or paranoid patients. I vividly remember one patient whom I interviewed in an emergency room. He was an intoxicated male about 30 years old, who wore a frequent sneer. He challenged me frequently with not-so-subtle sniper's remarks such as, "I bet you think you're a good listener Doc." And at one point he suddenly began mocking my head nodding by aping it, with his jaw jutting outwards while grunting out loud "Uhhuhs." This was not one of my more rewarding interviews. He was the patient who later, while waiting for his disposition, spontaneously attacked one of our safety guards.
     This patient also illustrates the point that if the clinician finds a patient giving negative responses to typically engaging nonverbal behavior, then the interviewer should consider the idea that the patient may be guarded, hostile, or potentially violent.

Each clinician has a unique personality. In particular, clinicians will vary on parameters such as tone of voice, rate of speech, and loudness of voice. It is important for clinicians to discover their own typical way of coming across. This knowledge is of value, because certain patients may respond better to different approaches. An understanding one's own natural style offers the clinician the chance to modify it, if necessary, to enhance the blending process.
     With this idea in mind, it is useful for clinicians to practice exercises such as speaking more gently and slowing down their rate of speech. If an interviewer tends to speak loudly and quickly, a toning down of these parameters may prove more effective with a frightened or guarded patient. By way of example, my own personality is somewhat upbeat, with a mild pressure to my speech and a slightly louder voice than many people. When beginning interviews, I purposely adjust to a calmer middle ground until I understand the specific needs of the patient. Adjustments can then be made as deemed necessary. In instances when I have not made this adjustment, I have certainly come on too strongly for certain patients.
     There exists another area in which tone of voice can frequently disengage a patient. Specifically, when talking with geriatric patients, clinicians often unconsciously adopt a rather distinctive tone of voice. They talk as if they were speaking to a helpless child. This tone of voice, which is often mildly slowed, can easily be perceived as condescending. It is an extremely frequent phenomenon, and
clinicians must guard against it carefully. It is sometimes even done with psychotic patients and adolescents. In both cases the clinician is flirting with trouble.

We are generally well trained to observe the behavior of others, but the value of self-observation is frequently underplayed. As we have seen, the interview represents a dyadic process in which an understanding of one component depends on an understanding of the impact of the other component. The clinician's nonverbal activity always has the potential to significantly alter the behavior of the patient, as we have seen in our discussion of reciprocal behaviors.
     With regard to gestures, as with paralanguage, clinicians need to develop a sound sense of their natural nonverbal style. One exercise that helps clinicians in developing self-awareness consists of repeatedly picturing a mirror descending during the interview itself. This mirror is to drop into place between the clinician and the patient. Such a visualization exercise rather rudely awakens clinicians to the fact that their every move is potentially an object of scrutiny to an inquisitive patient. As a complement to this visualization exercise, videotaping provides invaluable objective self-observation.
In any case the clinician should foster an awareness of those nonverbal activities that may inadvertently decrease blending. I am reminded of an interview, which I supervised, of an adolescent boy. The patient sat in a pool of brooding preoccupation. He wore a worrisome expression more suited to a 60-year-old man coping with an agitated depression than to a boy beginning adolescence. Curiously he had referred himself to the evaluation center and did not want his mother to be contacted.
     During the interview he moved about anxiously in his chair and had considerable difficulty looking at the interviewer. He had a rounded face framed by a bowl of sandy hair, which was neatly clipped around his ears. It was about one of these ears that his discussion soon focused. Apparently he had the misfortune of watching a television documentary on cancer several days earlier. Since then he had become fixated on a small bump on his right ear, to which he gingerly pointed. He was convinced that he had developed a malignant tumor. This gnawing obsession, which may very well have reached delusional proportions, was nestled amidst a variety of depressive symptoms and difficult life circumstances.
     As the interview proceeded the boy became progressively more ill at ease. At several points he stopped talking, asking the interviewer, "You don't understand, do you?" To which the interviewer responded in a reassuring fashion that he was trying to understand and wanted to hear more. This type of response generally might have decreased the tension, but in this case it seemed of no avail.
What the interviewer did not realize was the message conveyed by his own face. Each time the boy discussed his "tumor" the clinician furrowed his brow in a not-so-subtle fashion, forming two small vertical lines between his eyebrows. Apparently the patient interpreted this facial gesture as a look of disbelief or condemnation. The clinician had no conscious awareness of this particular expression, which frequently cropped up as a habit during his interviews. It is just this type of habit that can lead to recurrent problems with poor blending.
     These habits are difficult to recognize unless the clinician is directly supervised or videotaped. They are also sometimes hard to accept. The clinician above seemed unimpressed with my explanation for the poor engagement until several weeks later. He then approached me sheepishly and said, "You'll never believe what a patient just did. In the middle of the interview he cut me off and asked me why I was frowning. My God, I must actually do it!"
     One of my own habits illustrates another category of clinician movement that can become problematic. As I become anxious, I begin to twist my hair behind my ears. This nonverbal activity represents what we have discussed earlier as a displacement activity. These displacement activities can be used to monitor patient anxiety, but on the flip side, they can be a useful self-monitor, indicating anxiety in the clinician.
     The clinician may not even have been aware of the presence of stress, but the appearance of numerous displacement activities warns that anxiety is present. At such points of self-awareness in the interview the clinician can explore the origins of the tension. Sometimes the interviewer is concerned about personal matters not related to the interview, including countertransference tensions. At other times the clinician may be intuitively registering patient hostility or even well-hidden psychotic process. In any case, the recognition of clinician displacement activities can provide yet another avenue for understanding.
     Another good reason for studying displacement activities concerns the eradication of potentially disengaging gestures. For the most part, displacement activities are natural and help to create a feeling of spontaneous communication. As such there is no need to eliminate them; indeed, they may actually foster good blending. But there exist certain displacement gestures that are probably best eliminated. We can return to my own habit of twisting my hair. This displacement activity has the potential to be disengaging. To some patients it may appear effeminate, because as mentioned earlier women touch their hair three times more frequently than do men. To others it may simply be distracting. In either case it serves no purpose and is probably best discarded.
     Similarly, certain categories of patients may not respond well to demonstrations of increased anxiety in the clinician. The immediate category that comes to mind includes patients escalating towards violence. These patients are frequently frightened that they are about to lose control. If they see the clinician becoming progressively more tense as well, they may become even more agitated. The same holds true for paranoid patients, who may appear almost ludicrously hyperattentive to their environments. I remember an older man with marked paranoid process who once asked me why I had just scratched my head. When I said I had an itch, he did not seem particularly reassured.
     Two other clinician displacement activities warrant discussion. The first activity is smoking. I personally do not believe that clinicians should smoke cigarettes or even the proverbial "Freudian pipe" while interviewing patients. My bias evolves from the feeling that smoking, at the very least, represents a possible distraction to the patient. More likely, it may sometimes actually function as an irritant. Even if one asks permission from the patient, many patients who do not like smoking may find it difficult to convey such concerns. Pipe smoking is so stereotypic of "a shrink" that it may bias transference or simply turn some patients off.
     The second displacement activity is much more of a mixed blessing, because it clearly serves some useful purposes. I had never even viewed it as a displacement activity until I had asked one student what his most common displacement activities were, and he replied, "That's easy, I'm constantly scribbling notes."
     There exist many good reasons for taking detailed notes, such as making process notes to be shared with a psychotherapy supervisor. On the other hand, in initial interviews I have become more and more convinced that much of notetaking represents a displacement activity that frequently distracts both the clinician and the patient. No matter how one views it, a clinician looking down at his or her clipboard while actively composing sentences cannot possibly be attending to the fine nuances of patient behavior available to the clinician with undivided attention.
     Once again I am sharing a bias that some clinicians would disagree with, but I feel that notetaking should be minimized in the initial interview. It should be utilized to jot down hard-to-remember details such as dates, medication dosages, and family trees. Instead of meticulously making a transcription of the patient's words, the clinician can carefully attend to the patient directly. In particular, during the early scouting period, I believe it is much better to do little, if any, notetaking. At this early stage the emphasis should be on actively engaging the patient. To this end, I find that patients are more responsive to clinicians who seem more interested in them than in the clinician's clipboard.
     I frequently do not even pick up a clipboard until well into the interview. When I do begin to write, as a sign of respect, I often say to the patient, "I'm going to jot down a few notes to make sure I'm remembering everything correctly. Is that all right with you?" Patients seem to respond very nicely to this simple sign of courtesy. This statement of purpose also tends to decrease the paranoia that patients sometimes project onto notetaking, as they wonder if the clinician is madly analyzing their every thought and action. Along these lines, notetaking should be avoided with actively paranoid patients.

Interacting with a patient who is escalating towards violence presents the clinician with one of the most difficult of clinical situations. Although it would be nice to think that violent interactions are rare, the facts speak otherwise. Tardiff reports that approximately 17% of patients reporting to an emergency room are violent. He further reports that roughly 40% of psychiatrists have reported being assaulted at least once in their careers.(50)
     Obviously, it is to the clinician's benefit to review the various approaches that may de-escalate an angry patient. In particular, the nonverbal characteristics of potentially violent dyads are of considerable importance, because issues concerning proxemics, kinesics, and paralanguage can all be of value in handling these situations. The interaction with the potentially violent patient provides an excellent topic with which to close this chapter, for the craft of utilizing nonverbal behavior is seldom put to a more critical test.
     I would also like to emphasize that violence is frequently a dyadic process. The clinician and the patient represent a two-person system, and it is this system that becomes violent. Clinicians may inadvertently, with their nonverbal behavior, further escalate an already agitated patient. Fortunately this cycle, representing a violence reciprocal, can frequently be broken.
     To begin with I am reminded of a curious story related by an anthropology professor during my undergraduate education. He described an interspecies encounter in which violence was averted by the quick thinking of a field anthropologist. This anthropologist had been extensively studying the behaviors of a baboon troop. One day he accidentally startled a mother baboon and her baby. Within seconds the squawkings of the alarmed mother attracted a swarming bevy of guard males. One can assume their intent was not of a social variety. Indeed, baboons are both intelligent and ferocious when provoked. The appearance of an ugly white ape with a mustache and safari hat was more than ample stimulus to prompt a display of their virility. Indeed, the baboons could have quickly disposed of the anthropologist.
     Having observed baboons demonstrating submissive behavior within the troop, he purposely replicated their submissive gestures, which apparently involved lowering oneself and making certain jaw movements. To his relief, the baboons grunted and snarled but waved off their attack.
Besides representing a delightful tale for college professors to relate to wide-eyed undergraduates, the above story has a valuable message. A group of animals were about to interact violently. The violence was prevented by the use of specific nonverbal behaviors, which functioned as actual nonverbal communications. Similar to these baboons the human animal possesses a repertoire of nonverbal activities and communications that signal the intent to attack and the intent to submit.
     For the clinician, the signals of impending attack, when recognized in a patient, can quickly alert the clinician that something needs to be altered in the interpersonal dyad before a violence reciprocal ensues. Through a knowledge of the signals of submission, the clinician may alter behavior in a fashion that appears less threatening to the paranoid or intoxicated patient. In many instances these alterations can break the dyadic cycle of violence, as effectively as the anthropologist supplicating the baboon warriors. It should be kept in mind that in some instances no matter what preventive actions are undertaken, violence will erupt. The goal is not to eliminate violence but to decrease its likelihood.
     Towards this endeavor the clinician should consider whether the clinical environment suggests that violence may be a possibility. In the first place, diagnosis can alert the clinician to an increased likelihood of aggression. Most psychotic patients are not violent, but psychotic process as manifested in schizophrenia, bipolar disorder, paranoid disorder, and other atypical psychoses may predispose the patient towards aggression, especially when paranoid delusions are simmering beneath the patient's social facade. If frightened, these paranoid patients may go to great extremes to protect themselves, as we would if we shared their vision of the world. It is always important to remember that such patients may believe that they are literally fighting for their lives.
     Other types of psychosis or poor impulse control may presentproblems. For instance, patients suffering from organic brain disease,as seen in frontal lobe syndromes, deliriums, and various dementias,may be predisposed towards aggression. A particular red flagshould arise in the clinician's mind when interacting with peopleunder the influence of various drugs, including speed, Quaaludes,and PCP. Alcohol intoxication remains a major area in which violence erupts, especially in settings such as emergency rooms. Because we frequently deal with alcohol intoxication in social settingsin our culture, it is easy to be lulled into underestimating thepotential for violence when dealing with an intoxicated patient.Such patients can quickly move from jovial jesting into a fit of rage. Diagnoses do not tell the clinician that any specific patient isabout to be violent. Most people suffering from schizophrenia arenot violent, but the diagnosis does alert the clinician to the possibil
ity of aggression. This consideration may represent the first step in preventing violence. In addition, the clinician may note that a patient has a history of assaultive behavior. In such instances, the clinician is well advised to take appropriate precautions, such as having safety officers unobtrusively nearby and aware of the situation.
     Besides diagnostic and historical factors, the clinician may be part of a situation in which violence is more likely. If the clinician has been asked to participate in the evaluation of a patient who is being committed involuntarily, then caution is always advised. There are probably few life situations more frightening than to have one's freedom taken away. In this situation patients should always be considered as potentially violent.
     I remember one instance in our emergency room late at night. The patient, an agitated woman of about 30 years of age, was being committed. Safety officers had been called down and were appropriately nearby. The patient appeared to have calmed and was quietly sitting with family members by her side. Everything seemed in control. The clinician began to move away from the patient and turned her back as she headed for the staff room. In a matter of seconds the patient was ferociously choking the clinician, for no apparent reason. I mention this vignette because it highlights the need to think cautiously while evaluating committed patients. It also reminds one of the old adage that one should never turn one's back on a patient, an adage as true today as when it was first coined.
     One other clinical situation to keep in mind arises when patients are agitated and accompanied by family members. In such situations the clinician should attempt to determine quickly whether the family member is calming or upsetting the patient. In emergency rooms a common mistake is to not separate feuding family members until it is too late. It is often best to separate the antagonistic family members quickly, while allowing different staff members to attempt to calm and understand the perspectives of both parties.
     I have strayed from the topic of nonverbal behavior. However in a practical sense, the first step in utilizing nonverbal behavior with violent patients consists of recognizing the violent situation in its infancy, not its adolescence. If the clinician is aware of the potential for violence, then the following nonverbal techniques can be brought into play.
     We will first look at various nonverbal activities that may alert the clinician that violence may be incubating. Subsequently we will look at ways in which to change our own behaviors in an effort to avoid confrontation. The signs of impending aggression can be loosely grouped into two categories--early warning signs and late warning signs. Although it is extremely difficult to predict whether a patient will engage in violence in the future, it is not particularly difficult to tell when a patient may be headed towards immediate violence.
     The early warning signs consist of behaviors that suggest emerging agitation. In the simplest examples, one may notice the patient beginning to speak more quickly with a subtly angry tone of voice. These paralanguage clues may be augmented by a display of sarcastic statements or challenges, such as, "You think you're a big shot, don't you!"
     These types of early warning signs may appear obvious, which is the exact reason why they warrant mentioning. As clinicians we may inadvertently ignore these signs, in the process unintentionally escalating the patient. This seems to occur during periods of intense time pressure or when the clinical situation has become increasingly hectic, as in a busy emergency room. Such obstinacy can unfortunately return as an unwanted gremlin. When these early warning signs are present, it is very important to crystallize in one's mind what the patient's needs may be. If the clinician can move with the patient's needs, hostility will frequently decrease.
     Kinesic early-warning signs consist of actual evidence of agitation, such as pacing and refusing to sit down. If patients refuse to sit, it is frequently useful to gently request them to return to their seat. One can use phrases such as, "It might help you to relax some if you sit over here," or "Let's sit down and see if we can sort some things out." If comments such as these fail to elicit compliance, one can more firmly state, "I'd like you to sit over here so we can talk." Some clinicians might quietly add, "It's difficult to have to keep staring up. I think we'll both be more comfortable if we sit." If these maneuvers fail, then it is probably best to let the patient walk around freely, while recognizing that this patient may be seriously impaired with regard to impulse control. In short, the patient may be on the way towards violence, and appropriate steps should be taken. If no one is aware that the clinician is alone with such a patient, it is generally best to let someone know what is going on. It is relatively easy for a clinician to make an excuse for leaving the room at such points. It may not be so easy 10 minutes later. Along these lines, if the clinician is at all suspicious of possible violence, the clinician should carry a "safety button" or know where the safety button is located in the interview room, so that other staff can be alerted if problems arise.
     Other kinesic early warning clues include rapid and jerky gesturing. Of particular note is the action of vigorously pointing one's finger at the clinician to "make a point." Such a gesture may be a harbinger of impending hostility. Increased and intense staring may also suggest anger. Finally, the appearance of suspiciousness or other increases in psychotic process, such as an increasing disorganization, should alert the clinician to the possibility of violence.
     As a person comes closer to overt violence, specific behaviors may serve as reliable indicators that aggression is imminent. Just like the charging guard baboons with their bared teeth, humans have evolved symbolic signs of threat. Morris has described behaviors known as intention movements.(51) These intention movements consist of those small gestures that suggest impending movement. For instance, as persons intend to rise from a chair, they frequently lean forward grasping the arms of the chair. This is a clear signal that they want to rise, signaling that the conversation is about to end. The intention movements suggesting possible violence include activities such as clenching of the fists, whitening of the knuckles as one tightly grasps an inanimate object, and even a snarling as the lips are pulled back from the teeth. People may not be as different from baboons as we would like to think.
     Perhaps the most common intention movement of attack is the raising of a closed fist over the head. Overhand blows delivered from this position are the most frequent blows seen in street brawls and riots despite the unlikelihood of hurting one's opponent in this manner. This behavior may be instinctual in nature, because it is frequently seen in children who are fighting.
     Morris also describes vacuum gestures, which represent completed actions but are not actually carried out on the enemy. Frequent vacuum gestures include shaking the fist, assuming a boxing stance, gesturing as if strangling the opponent, and the pounding of the fist into the opposite palm. All of these intention movements and vacuum gestures serve as late warning signals that violence is near at hand.
     It should also be noted that verbal threats or statements that one is about to strike out often accompany the nonverbal behaviors described above. When the above late warning signs are present, violence is a distinct possibility. At this point an application of nonverbal skills may help to prevent aggression.
     Earlier, reciprocal behaviors were discussed, such as the mating reciprocal, in which two organisms engage in an orderly sequence of events leading to a final outcome. Scheflen describes dominance and submission reciprocals.(52) In our story of the baboons, the anthropologist refused to participate in the dominance reciprocal. If he had, he might very well have been killed. Instead he chose to begin the submission reciprocal, which his would-be attackers fortunately agreed to follow. In a similar fashion, humans can engage in either of these reciprocals.
     When faced with a hostile patient, the trick is to avoid engaging in the dominance reciprocal while utilizing some submissive behavior. One avoids the dominance reciprocal by not demonstrating any of the early or late warning signs of aggression. Although this appears to make an obvious point, it is striking to watch the maladapative behavior of clinicians when faced with an agitated patient. The fear generated by the patient's hostility frequently results in unconscious behaviors that may threaten the patient. The clinician's voice may be raised. At times, the actual movements of the clinician speed up as the waiting area is hurriedly cleared of furniture and other patients. Even frankly antagonistic remarks may emerge. In this respect, it is not an exaggeration to say that clinicians can actually precipitate violence.
     There exist no absolute rules for interacting with a patient on the verge of violence, but some principles seem relevant. In the first place, the clinician should appear calm. The speaking voice should appear normal and unharried. It is particularly important to avoid speaking loudly or in an authoritarian manner. With regard to kinesics, the clinician wants to avoid an excessive display of displacement activities, which may be misinterpreted as aggressive displays. Moreover, exaggerated displacement activities may create an increasing atmosphere of fear, stoking the patient's own fears of an impending loss of control.
     Eye contact should probably be decreased, and the hands should not be raised in any gesture that may signify an intent to attack or defend oneself. Curiously, some clinicians will place their hands behind their backs, a situation that may raise fears in the patient that a weapon is being hidden. With regard to posture, one can purposely stoop one's shoulders slightly in an effort to appear smaller, because humans, when about to attack, frequently raise their shoulders and chests, a bit of gorilla-like display. It is probably also wise to remain in front of the patient, because an approach from behind or from the side may startle the agitated patient.
     One of the most important points concerns an issue mentioned earlier when discussing proxemics. At least one study has suggested that potentially violent patients may have significantly altered buffer zones.(53) Specifically, they will feel that their intimate body space is being invaded at distances much further away than for most people. These patients may feel that the interviewer is "in my face" while standing a full 6 feet away. In general, the agitated patient needs more room and interpersonal space. This can be a tough principle to remember, because some good-hearted clinicians feel a desire to calm the angry patient by touching them. This desire usually goes away after a few unfortunate encounters with feet or fists.
     If these principles are followed, accompanied by an intelligent use of safety officers and medication as needed, many violent encounters can be avoided. With regard to avoiding dangerous situations, another point warrants mentioning. When sitting in a room with a patient who one does not know, it is probably wise to arrange the chairs so that the clinician is closer to the doorway, while not obstructing the patient's pathway to the doorway. With this arrangement one can always get away if the patient becomes threatening or produces a weapon. It is naive to think that these situations do not arise, especially in emergency rooms. To pretend that they do not probably represents a defensive denial that prevents the clinician from fully thinking about these situations in a manner that could help prevent them in the first place.
     In conclusion, nonverbal processes are core elements of human communication during violent interactions. A sound knowledge of these processes can help the clinician to calm the angry or frightened patient. Helping patients to regain a sense of internal control remains one of the fine points of the art of interviewing. It also increases the chances that the clinician will be around to practice his or her art.

In this chapter we have reviewed the basic principles of proxemics, kinesics, and paralanguage. It can readily be seen that these processes are at the very root of communication. As such integral parts of human interaction, they remain pivotal in bringing the initial interview to a successful conclusion.
In these first three chapters, we have reviewed many of the basic principles of both verbal and nonverbal behaviors as they apply to the initial interview. Before proceeding much further, the important topic of gathering and organizing information for treatment planning warrants a thorough discussion. Such a discussion will quickly move us into some of the most complex and fascinating aspects of assessment interviewing.

1. Hall, E. T.: Excerpts from an interview conducted by Carol Travis. GEO 2-5(3):12,1983.
2. Harper, R. G., Wiens, A. N., and Matarazzo, J. D.: Nonverbal Communication: The State of the Art. New York, John Wiley, 1978.
3. Wiener, M., Devoe, S., Rubinow, S., and Geller, J.: Nonverbal behavior and nonverbal communication. Psychological Review 79:185-214,1972.
4. Edinger, J. A., and Patterson, M. L.: Nonverbal involvement and social control. Psychology Bulletin 93(l):30-56,1983.
5. Morris, D.: Manwatching: A Field Guide to Human Behavior. New York, Harry N. Abrams, 1977.
6. Hall, E. T.: The Hidden Dimension. New York, Doubleday, 1966.
7. Watson, O. M., and Graves, T. D.: Quantitative research in proxemic behavior. American Anthropologist 68:971-985, 1966.
8. Sue, D. W., and Sue, D.: Barrier to effective cross-cultural counseling. Journal of Counseling Psychology 24(5):420-429, 1977.
9. Baxter, J. C.: Interpersonal spacing in natural settings. Sociometry 33:444-456, 1970.
10. Wiens, A. N.: The assessment interview. In Clinical Methods in Psychology, edited by Irving Weiner. New York, John Wiley, 1976.
11. Knapp, M. L.: Nonverbal Communication in Human Interaction. New York, Holt, Rinehart, and Winston, 1972.
12. Birdwhistell, M. L.: Introduction to Kinesis: An Annotation System for Analysis of Body Motion and Gesture. Louisville, KY, University of Louisville Press, 1952. 13. Harper, R. G., 1978, p. 123.
14. Scheflen, A. E.: Body Language and Social Order. Englewood Cliffs, NJ, PrenticeHall, 1972.
15. Freud, S.: Fragment of an analysis of a case of hysteria. In Collected Papers, Vol. 3. New York, Barri Books, 1959 (originally published in 1925).
16. Carmier, W. H., and Carmier, L. A.: Interviewing Strategies for Helpers: A Guide to Assessment, Treatment and Evaluation. California, Brooks/Cole Publishing Company, 1979.
17. Wiens, A. N., 1976, p. 27.
18. Sue, D. W., 1977, p. 427.
19. Hill, Sir Denis: Non-verbal behavior in mental illness. British Journal of Psychiatry 24:221-230, 1974.
20. Hill, Sir Denis, 1974, p. 227.
21. Wiens, A. N., 1976, p. 33.
22. Morris, D., 1977, p. 75.
23. Scheflen, A. E., 1972, p. 46.
24. Morris, D., 1977, p. 164.
25. Morris, D., 1977, p. 165.
26. Pansa-Hendersen, M., De L'Horne, D. J., and Jones, I. H.: Nonverbal behavior as a supplement to psychiatric diagnosis in schizophrenia, depression, and anxiety neurosis. Journal of Psychiatric Treatment and Evaluation 43:489-496, 1982.
27. Jones, I. H., and Pansa, M.: Some nonverbal aspects of depression and schizophrenia during the interview. Journal of Nervous and Mental Disease 167(7):402409, 1979.
28. Pansa-Henderson, 1982, p. 495.
29. Jones, I. H., 1979, pp. 402-409.
30. Morris, D., 1977, p. 181.
31. Morris, D., 1977, p. 109.
32. Ekman, P., and Friesen, W. V.: Detecting deception from the body or face. Journal of Personality and Social Psychology 29(3):288-298, 1974.
33. Ekman, P.: Telling Lies: Clues to Deceit in the Marketplace, Politics, and Marriage. New York, W. W. Norton, 1985.
34. Littlepage, G. E., and Pineault, M. A.: Detection of deceptive factual statements from the body and face. Personality and Social Psychology Bulletin 53(5):325-328, 1979.
35. McClintock, C. C., and Hung, R. G.: Nonverbal indicators of affect and deception in an interview setting. Journal of Applied Social Psychology 5:54-67, 1975. 36. Edinger, J. A., 1983, pp. 42-43.
37. Kraut, R. E.: Verbal and nonverbal cues in the perception of lying. Journal of Personality and Social Psychology 36:380-391, 1978.
38. Grinder, J., and Bandler, R.: The Structure of Magic II. Palo Alto, California, Science and Behavior Books, 1976.
39. Scheflen, A. E.: Body Language and Social Order. Englewood Cliffs, NJ, Prentice Hall, 1972.
40. Scheflen, A. E., 1972, p. 16.
41. Morris, D., 1977, p. 102.
42. Morris, D., 1977, p. 102.
43. Egan, G.: The Skilled Helper: A Model for Systematic Helping and Interpersonal Relating. California, Brooks/Cole Publishing Company. 44. Wiens, A. N., 1976, p. 35.
45. Sue, D. W., 1977, pp. 420-429.
46. Morris, D., 1977, p. 83.
47. La Crosse, M. B.: Nonverbal behavior and perceived counselor attractiveness and persuasiveness. Journal of Counseling Psychology 19:417-424, 1972.
48. Hasse, R. F., and Tepper, D.: Nonverbal component of empathetic communication. Journal of Counseling Psychology 19:417-424, 1972. 49. Morris, D., 1977, p. 68.
50. Tardiff, K.: The violent patient. In Manual of Psychiatric Consultation and Emergency Care, edited by F. Guggenheim and M. Weiner. New York, Jason Aronson, 1984.
51. Morris, D., 1977, p. 173.
52. Scheflen, A. E., 1977, p. 173.
53. Wiens, A. N., 1976, p. 28.