Sage Journals: Discover world-class research

Abstract

It has been shown previously that distance mitigates the extent to which nonverbal cues such as direction of eye gaze, facial expression, sex, and posture communicate threat. This article extends the relevant body of knowledge by investigating a new factor: proximity of threat. Theory predicts that environments with two stimuli, such as people, would be perceived as more threatening if the nearer (proximate) person exhibits hostile cues and the farther (distal) person exhibits benign cues rather than vice-versa. The results, based on an experiment with 24 scenes and 24 participants, supported the predictions for the visual cues of sex and posture at distances from 5 to 20 m.

This article reports findings on how distance and proximity mitigate perceived threat as communicated with the non-verbal cue of posture. This particular topic was chosen because of its relevance to a new theory of environmental perception and preference—permeability theory—and a gap in the existing literatures. In order to clarify the context of this work, we begin with a brief summary of permeability theory.

Permeability Theory

The quintessential scenario for permeability theory was written by the biologist Heini Hediger in the following language:

The most important biological activity of all animals, for that matter of all living organisms, and to which all their biological functions and behavior are directed, is without any doubt the preservation of the species… The satisfaction of hunger and sexual appetite can be postponed: not so escape from a dangerous enemy. As far as the higher animals are concerned, escape must thus at any rate be considered as the most important behavior biologically. The primary duty of the individual, to ensure its own existence, and thus the preservation of the species, lies in being prepared to escape. By far the chief occupation of the free wild animal, therefore, is constant watchfulness: eternal alertness for the purpose of avoiding enemies. (Hediger, 1955, p. 39).

Hediger worked with animals. Other researchers have applied similar concepts to people. One example is Goffman's (1971) Relations in Public. Goffman was concerned with face-to-face interactions in public spaces, with particular emphasis on the “un-safety and incivility of our city streets” (Goffman, 1971, p. ix). Goffman's theory of relations in public was explicitly derived from Darwin (Goffman, 1971, p. 239) and Hediger (Goffman, 1971, pp. 30, 251, 252). The basic scenario in Goffman's theory is this:

Individuals, whether in human or animal form, exhibit two basic modes of activity. They go about their business grazing, gazing, mothering, digesting, building, resting, playing, placidly attending to easily managed matters at hand. Or, fully mobilized, a fury of intent, alarmed, they get ready to attack or to stalk or to flee…. The individual mediates between these two tendencies with a very pretty capacity for dissociated vigilance. Smells, sounds, sights, touches, pressures—in various combinations, depending on the species—provide a running reading of the situation, a constant monitoring of what surrounds. But by a wonder of adaptation these readings can be done out of the furthest corner of whatever is serving for an eye, leaving the individual himself free to focus his main attention on the non-emergencies around him. Matters that the actor has become accustomed to will receive a flick or a shadow of concern, one that decays as soon as he obtains a microsecond of confirmation that everything is in order; should something really provide to be ‘up’, prior activity can be dropped and full orientation mobilized, followed up by coping behavior. Note, the central thesis here is Darwinism. If individuals were not highly responsive to hints of danger or opportunity, they would not be responsive enough; if they carried this response far on every occasion of its occurrence, they would spend all their time in a dither and have no time for all the other things required for survival. (Goffman, 1971, pp. 238–239).

In permeability theory, detection of safety is the most important function an organism can have in an environment. This implies that when a person (or animal) inspects an environment, features of that environment are first and foremost evaluated in terms of how they indicate safety or its converse, threat. Consequently, as pointed out by Hediger, perceived safety is intimately entwined with the abilities to see or to move. Another term for “abilities to see or move” is permeability. The ability to see through is visual permeability. The ability to move through is locomotive permeability. These two types of permeability are important because both influence perceived safety, which again, is taken to be the most important function of environmental perception and preference. This implies that detection of the ease and range of either vision or locomotion should be major components of a theory of environmental perception and preference. In other words:

The most important function an environment can provide for its inhabitants is safety.

Safety depends on how far and how easily one can perceive or move through an environment.

Ability to pass through something is called permeability, from which comes the name of this theory, and

Therefore, properties of physical environments that mitigate either perceiving through or moving through those environments should be strongly related to responses of perceived safety.

It is possible to conduct formal experiments on the premises of permeability theory, e.g., Greene and Oliva (2009) measured how quickly people used two types of criteria when evaluating environments. One type of criteria used constructs that clearly indicate ability to see or to move. Those constructs were concealment, mean depth, navigability, and openness. The other type of criteria was the identification of what kind of landscape was being shown: lake, forest, desert, etc. The pertinent point is that evaluations of permeability were faster than evaluations of identifying the kind of landscape. In terms of correlations, the effect size was r = −.26. This result suggests that the first premise of permeability theory—the primacy of detecting threat—obtains: threat detection actually does have temporal priority over classification of environments. Many researchers have explored evolutionary theories of non-verbal communication because, as Montepare (2003, p. 61) noted, “Evolution and nonverbal behavior are undeniably linked.” Öhman (1986, p. 123) suggested that “… animal fear originates in a predatory defense system whose function is to allow animals to avoid and escape predators,” thus echoing the point made by Hediger. Many other examples could be found. Patterson (2003) used evolutionary theory to explain nonverbal mimicry. Other interpretations and implications of evolutionary behavior with respect to nonverbal communication are reviewed in Floyd (2006) and Anderson (2008, pp. 139–140). Additional examples can be found in Le Doux (1996; 1998; 2003), Öhman (2008), Öhman and Mineka (2001), Lundqvist and Öhman (2005) and Hansen and Hansen (1988). On the whole, it seems safe to assert that Darwin's insights regarding nonverbal cues and Hediger's insights into the importance of fear and distance are worth inquiry.

Non-verbal Communication by Posture and Distance

The fundamental reference for communication by nonverbal cues is Darwin's The expression of the emotions in man and animals (Darwin, 1872). More recent work on the cues of posture and distance is summarized in the following paragraphs.

Posture.—Reviews of how posture communicates are given in Argyle (1988), Bull, (1987), Planalp (1998), and Anderson, leading the latter to suggest that “Emotions can be reliably identified from postures…” (Anderson, 2008, p. 147), although, of course, counter-examples can be found, particularly in reference to cross-cultural differences (Matsumoto & Kudoh, 1987, Matsumoto 2006, pp. 220–221). Components of postures that have been studied include head, arms, legs, body lean and standing/sitting position (Pitterman & Nowicki, 2004), sphere of movement, expressed by variables such as limb-to-torso distances and overall expansion of the body in frontal, lateral, and vertical directions (De Silva & Bianchi-Berthouze, 2004), and—of direct relevance to the present article—Coulson's (2004) work based on bones, joints, and rotations of bones around the joints. Variables used to describe human postures were body center of mass (leaning forward, backwards, or upright), head bend, chest bend, abdomen twist, shoulder swing, shoulder ad/abduct, and elbow bend. The dependent variable was the emotion which the participants thought best described a posture (anger, fear, happiness, sadness, surprise, disgust). The experiment was made possible by using a computer program (Curious Labs; Poser, 2002) for which each anatomical variable could be changed independently. People were modeled as the wood mannequins typically used by artists to study figure drawing, so possible confounding of effects due to posture or to facial expression were controlled. Altogether there were 176 mannequins. Each was imaged in three views (front, side, and rear). Sixty-one participants classified each stimulus into one of the six emotions, a multidimensional logistic regression model was used to predict the classifications in terms of the independent variables, and the percentage of correct classifications was used to evaluate how well the model worked. Overall, 48.5% of the predicted classifications were correct. For anger, 34.9% of the predicted classifications were correct (Coulson, 2004, p. 130).

Distance.—Hediger worked with animals in venues such as the plains of the Serengeti. Other researchers have applied the same concepts to people in more contemporary venues such as cities. For instance, Hall based his work on the work of Hediger (Hall, 1966, p. 8). Hall focused on distance. The general zones of distance in Hall's presentation are intimate (less than 0.5 m), personal (0.5 to 1.2 m), social (1.2–3.0 m), and public (beyond 3.0 m). Each zone is further divided into close and not close categories. Personal contact is possible at the closer distances with a maximum range of about 2.5 m for two people to reach out and touch each other (Hall, 1966, p. 126). Human flight distance was estimated at 3.6 m (Hall, 1966, p. 119). Feedback from the eyes becomes difficult to detect at ranges of 3.6 to 7.6 m, and at about 4 m, a whole body can be observed with a 1.05 radian field of view (Hall, 1966, p. 124). For effects distances longer than 7.5 m, Hall notes that people look like ants, details of facial expression and movement are lost, and nonverbal communication shifts to gestures and body stance (Hall, 1966, p. 125).

Other researchers have also investigated proxemics, where “proxemics” is taken to mean, “The study of communication through interpersonal space and distance…” (Anderson, 2008. p. 44). Relevant references were found in two abstract searches in the Science Citation Index on the topics of “proxemics” and “interpersonal distances.” The search on the topic of proxemics produced 54 hits, of which one reported findings for distance, and that was a survey of robotic vision (Bethel & Murphy, 2008). The search on the topic of “interpersonal distance” generated 345 hits, of which six appeared to test for effects of physical distance. Five of the articles investigated distances in the range of 0–3 m (Lombard, 1995; Kaitz, Bar-Haim, Lehrer, & Grossman, 2004; Ducourant, Vielledent, Kerlirzin, & Berthoz, 2005; Li & Li, 2007; Ozdemir, 2008); the sixth article investigated records of social interactions rather than effects of nonverbal cues (Mok, Wellman, & Basu, 2007). Similar results were obtained by other reviews of the literature, e.g., Anderson's (1978. pp. 44–48) review of proxemics covered venues such as desks, parking spaces, library tables, and rooms with maximum dimensions ranging from 1.2 to 6.4 m. Sommer (2002, pp. 647–660) provided a more recent review of over a hundred articles on personal space, with emphasis on short distances in venues such as elevators, bars, ATM's, telephone booths, pre-schools, dentists' waiting rooms, water fountains, and busses. Other examples include Burgoon's (1991) study of posture, gender, status, and distances in the range of 0.3 to 2 m on seven affective responses, Thompson, Aiello, and Epstein's (1979) study of effects of seating distance in the range of 0–3 m, and studies by Aiello (1976, 1977) on effects of distance in the range of 0.76 to 4.4 m on looking time. One study was found that did address the issues of interest at the desired range: Walk and Walters' (1988) study of how faces and postures influenced perceptions of emotions over the range of 3 to 40 m. However, the published report consisted only of an abstract of a conference paper and so there was not enough information to evaluate the quality of the work.

Additional findings on distance effects were reported in the eight experiments described in Stamps (2011a, 2011b, 2012). The overall results suggest that posture communicated threat most strongly (r = .25), followed by distance (r = −.22), facial expression (r = .20), sex (r = .18) and eye gaze (r = .15). Posture communicated threat up to 84 m, while facial expression and direction of eye gaze stopped communicating threat at about 20 m. Sex had the same effect on communication of threat regardless of distance. Figure 1 shows the overall results of the three referenced articles. Figure 1 shows the ability to distinguish hostile from benign threats (r_contrast) for four visual cues at various distances. The utility of this type of work is indicated by the fact that, when studied with the formal protocols of experimental design, the range over which the non-verbal cue of posture communicates threat is an order of magnitude larger than the estimate generated by Hall's informal commentary based on observations. When choosing which cues to use in the detection of conspecific threat, the cue with the strongest effect over most distances will be posture.

Fig. 1.
Plots of findings from previous studies on how the ability to distinguish between hostile and benign people varies with different nonverbal cues at different distances.

Hypotheses

With the above literature in mind, the primary a priori hypotheses for this experiment were:

Hypothesis 1. Threat (hostile posture proximate, benign posture distal)-Threat (benign posture proximate, hostile posture distal) > 0.0. Or, in English, “When the stimulus contains a hostile posture in the foreground and a benign posture in the background, the perceived threat is greater than if the stimulus contains a benign posture in the foreground and a hostile posture in the background.”

Hypothesis 2. Threat (male proximate, female distal)-Threat (female proximate, male distal) > 0.0. In English, “When the stimulus contains a male in the foreground and a female in the background, the perceived threat is larger than if the stimulus contains a female in the foreground and a male in the background.”

In addition, since replication is an important component in the process of science, three other hypotheses, based on previous data, were tested:

Hypothesis 3. Males would be more threatening than females.

Hypothesis 4. Hostile postures would be more threatening than benign postures.

Hypothesis 5. Perceived threat would diminish with distance.

Method

The basic research method was to show participants static color images of environments and ask for ratings on a semantic differential scale such as (1) safe to (8) threatening. A very substantial amount of research has been done regarding the details on the validity and reproducibility of the protocols used in this work. Empirical data, covering over 33,000 participants and 7,700 stimuli in 21 countries, have been reported on methodological issues such as representing affective responses, number of responses, the recording task, simulation medium, presentation venue, demographic distinctions, and reproducibility. A review of the validation data, including estimates of reliability, is given in “Validation data for evaluating effects of environments on affective responses” (Stamps, 2013), as well as in this author's previous articles on conspecific threat, for readers who are unable to access the internet. Standards against duplicate publication do not allow repetition of all that data in this article. Instead, only the data of the topics of particular relevance to the present study are included here. The topic of particular relevance to the present study is the validity of simulations. Reservations regarding the validity of simulations are common but the actual empirical evidence indicates that these reservations have little scientific merit. The relevant data fall into three classifications: simulated environments per se, simulated distance, and simulated people.

Simulated environments per se.—Data on the validity of simulation media were reported in a literature review covering 1,215 stimuli and over 4,200 respondents (Stamps, 2000, pp. 101–113). At that time, there were not sufficient data in the public record to make claims regarding dynamic simulations, so the focus was on the efficacies of various static media. Media covered in that review included on-site, slides, black and white slides, tinted sketches, digitized slides, black and white sketches in low, medium, and high abstraction, computer sketches, verbal descriptions, black and white photographs, color photographs, models, full size mock-ups, 140° color photographs, 63° color photographs, unframed photographs, working drawings, pen and ink line drawings, and photomontages. Preferences obtained from slides or colored photographs correlated at r = .83 with preferences obtained on-site. Moreover, some media performed better than other media. In particular, it appears that color is a necessary requirement for a valid visual simulation protocol. With color, the correlation with on-site preferences was r = .83; without color, the correlation decreased to .4-.5.

By 2010 there were sufficient data in the public record to update the review to include dynamic simulations (Stamps, 2010). That review covered 6,323 participants and 967 environments. Responses included ratings of pleasure, naturalness, familiarity, order, inertia, arousal, threat, disliking, liking of the environment, nice area to walk through, good area to live in, appreciation of the area, visual appeal, evaluation, ambience, arousal, privacy, security, pleasant, interest, comfortable, excited, playful, water, built, water flow, sun, sound, strolling, resting, talking, observing, preference, and spaciousness. Some studies reporting findings for physical properties such as distance, height, width, and length. Other studies reported behaviors such as path choice and exploration. The overall correlation of subjective responses obtained either on-site or by viewing static color simulations was r = .86. The overall correlation of subjective responses obtained either on-site or by viewing dynamic simulations was r = .83. The overall correlation of subjective responses obtained either from static or dynamic media was r = .82. It is suggested that both types of simulation generate statistically equivalent results and so a choice of simulation media should be based on efficacy rather than validity.

Simulated distance. —Other studies compared estimates of distance obtained on-site with estimates obtained from computer models. Hayashibe (2002) obtained a correlation of r = .99 for rooms and plazas; Witmer and Kline (1998) obtained a correlation of r = .99 for a hallway, and Allen and Rashotte (2006) obtained r = .91 for distances in a parking lot.

Simulated people. —Studies that have used synthetic figures as stimuli include Baillenson, Pontikakis, Mauss, Gross, Jabon, Hutcherson, and Oliver (2008), Coulson (2004), De Silva and Bianchi-Berthouze (2004), Etcoff and Magee (1992), and Meeren, van Heijnsbergen, and de Gelder (2005). Direct tests of how well synthetic people communicate perceived threat were given in studies in this author's work on non-verbal communication. Faces that were designed to express hostility correlated at r = .94 with perceived threat, and postures that were intended to express hostility correlated at r = .70 and r = .61 with perceived threat (Stamps, 2011b). More recent work has focused on inferring emotions from specific parts of faces and computer recognition of emotions from faces in real time, e.g., Ekman's facial action unit system (FACS) maps changes in groups of facial muscles (action units or AU's) are mapped onto events such as raising eyebrows, wrinkling noses, and dropping jaws. It then becomes possible to test how well these various action units predict emotions. Typical stimuli for computer images are static faces or videos of unposed people. Bartlett, Hager, Ekman, and Sejnowski (1999) review the literature. Reported success rates have been high. Anderson and McOwan (2006, pp. 96–97) reviewed studies of automated facial recognition that reported recognition rates in the range of 74–98%. Another use of computer technology is to create synthetic stimuli. Etcoff and Magee (1992), e.g., created synthetic faces that began with photographs representing emotions based on the FACS but were altered to vary continuously in eyebrows, eyelids, and mouth. The result was a series of line drawings of faces that varied from one pure emotion to another. Even with the simplification of the images, the distinction between the emotions of anger and fear was quite strong, with 100% recognition rate for faces showing only anger (F_1,96 = 11.91, p < .001). Explicit mappings of AU's to perceived emotions are given in both research publications (Kohler, Turner, Stolar, Bilker, Brensinger, Gur, & Cur, 2004) and in computer programs that use AU's to create faces that express different emotions (Mortier, 2003, pp. 112–117). Accordingly, it seems reasonable to assume that people can infer emotions in general and anger in particular from faces and from specific parts of facts for representations ranging from static poses by live actors to video tapes of unposed interactions to simplified computer line drawings.

Stimuli and Experimental Design

The experimental method was to show participants static color images of environments and ask for affective responses in terms of semantic differential scaling. Stimuli consisted of locating synthetic people in a simulation of a Roman forum. The people were created with reproducible postures (either hostile or benign) for both men and women. Because the people were synthetic, the same postures could be used for both sexes. Images of the people and the validation data supporting their use are given in Stamps (2011b). The site plan for the Roman Forum is is shown in Stamps (2011a). Examples of the stimuli for this experiment are shown in Fig. 2.

Fig. 2.
Examples of factors and levels.

This work used a balanced incomplete block experimental design. The main factor was pairs of distance. The levels of this factor were 5 and 10 m, 5 and 20 m, or 10 and 20 m. The other two factors were posture [hostile (H) and benign (B)] and sex [male (M) and female (F)]. A fully crossed experimental design would require all combinations of posture and sex (HM, HF, BM, and BF) at each of the two levels of proximity (proximate, distal), for three pairs of distances. However, this would require 48 stimuli. In order to reduce the workload on the participants, a balanced incomplete block experimental design, using only 24 stimuli, was used. Table 1 lists this experimental design.

TABLE 1
Experimental Design

Stim Distances

5 m
10 m
20 m

Posture Sex Posture Sex Posture Sex

A H M H M

B H F B M

C B M H F

D B F B M

E H M H M

F H F B M

G B M H F

H B F B F

I H M H M

J H F B M

K B F H F

L B F B F

M H F H F

N H M B F

O B F H M

P B M B F

Q B F H F

R H M B F

S H F H M

T B M B M

U H F H F

V H M B F

W B M H M

X B M B M

Note B = benign; H = hostile; M = male; F = female.

Sample Size and Participants

Power analysis was used to calculate how many participants would be required. The target effect size was the average amount of variance of perceived threat attributable to posture, sex, or distance at ranges up to 20 m. Based on the eight previous experiments in this series, that value was about 9% of variance. With α = .05, β = 0.20, 24 stimuli, and 5 planned tests, only 7 participants were required. Twenty-four participants were recruited by a professional survey research firm, so statistical power was not an issue in this experiment. The sample included 12 males, 12 females, 8 liberals, 8 moderates, and 8 conservatives. The mean and standard deviation of age were 49.2 and 17.4 yr. Occupations ranged from accountant to social worker.

Task

The dependent variable was a semantic differential scale ranging from safe (1) to threatening (8). Images were shown on a laptop computer using a custom computer program. The screen measured 337 mm by 208 mm. The luminance of the screen was 150 lux. Participants sat approximately 400 mm from the screen in a room with an ambient light level of 150 lux. The computer program had two parts. The first part was a demonstration that showed the participants how to use the controls. After completing the demonstration, participants viewed the main part. In this part, the first screen stated what judgment was requested (“Please rate the following pictures on the criterion of how safe (1) or threatening (8) they appear.”), along with two images showing the extreme conditions in the experimental design. Then each stimulus was shown with a row of buttons on the bottom. The buttons were numbered from 1 to 8. When a button was pressed, an “OK” button appeared. When the “OK” button was pressed, the next stimulus was shown. Participants could change their minds and press other numbered buttons until they pressed the “OK” button. Participants were allowed to take as much time as they needed.

Results

Table 2 lists the means and standard deviations of perceived threat for each stimulus. Overall, the factor having the most influence on perceived threat was posture (14.0% of variance, F_1,529 = 155.56, α = 5 e-32), followed by distance (5.0%, F_1,529 = 55.14, α = 2 e-13), proximity of posture (2.2%, F_1,529 = 25.15, α = 3 e-7), sex (1.6%, F_1,529 = 17.43, α = 1 e-5), and proximity of sex (0.2%, F_1,529= 2.44, α = .07). The MSE in the analysis of variance was 2.45. The numerical values for the mean values of perceived threat for the levels of the stimulus factors were as follows. Posture: M_HH = 5.03, M_mixed = 3.94, M_BB = 2.73, where HH means both people in a scene were hostile, mixed means one hostile and one benign, and BB means both benign. Sex: M_MM = 4.29, M_mixed = 3.84, M_FF = 3.67, where MM means both male, mixed means one male and one female, and FF means both female. For distances, the levels were 5 and 10 m, 5 and 20 m, or 10 and 20 m, for means of M_5/10 = 4.29, M_5/20 = 4.32, and M_10/20 = 3.13. For proximity of sex, the levels were male in front or female in front, with the means of M_{male front}= 4.19 and M_{female front} = 3.76 (F_1,529=4.12, α = .03. For proximity of posture, the levels were hostile in front or benign in front, with means of M_{hostile front} = 4.37 and M_{benign front} = 3.51 (F_1,529 = 21.77, α = 1 e-6). Calculation of contrasts in terms of standardized mean differences can be made using the square root of MSE in the analysis of variance. In terms of correlations with perceived threat over n = 24 stimuli, the results were r = .17 for sex, r =.67 for posture, r = .39 for distance, r = .05 for proximity of sex, and r = .25 for proximity of posture.

TABLE 2
Means (M) and Standard Deviations (SD) for Perceived Threat for Each Stimulus (Stim)

Stim M SD Stim M SD

A 5.83 1.97 M 5.38 2.00

B 5.17 1.97 N 4.83 1.71

C 3.25 1.98 O 4.42 1.86

D 2.75 1.26 P 2.71 1.57

E 6.25 2.13 Q 4.42 2.02

F 3.54 1.82 R 6.25 1.94

G 4.46 2.02 S 3.58 2.00

H 2.67 1.81 T 3.38 1.81

I 4.79 1.98 U 3.50 2.02

J 3.13 1.70 V 3.33 1.69

K 2.46 1.96 W 2.92 1.89

L 2.42 1.82 X 2.50 1.22

For readers who do not need the accuracy of the preceding section, Fig. 3 shows the plots of the levels of each a priori factor. For Hypothesis 3 (males are perceived as more threatening than females), the previous findings were clearly replicated. Scenes with two hostile postures were perceived as being more threatening than scenes with one hostile posture and one benign posture, regardless of distance or whether the people were male or female. For sex, regardless of distance or proximity, scenes with both females or one female and one male were about equally threatening. It was only when both people were male that there was a detectable uptick in perceived threat. This finding replicates previous evidence for the male/female distinction on perceived threat, but adds the new assertion that in mixed company, threat was about the same as all-female. For distance, it made no detectable difference in perceived threat if the proximate distance were constant and the distal distance was increased, but it did make a detectable difference in perceived threat if the longer distance were constant and the proximate distance was increased. This finding implied that proximity matters. The next two findings (proximity of sex and proximity of posture) provided more direct evidence for the primacy of proximity in the perception of threat. For sex, a scene with a hostile person (male) in front and a less hostile person (female) in back was more threatening than the converse, regardless of distance or posture. For posture, a hostile posture in front and a benign posture in back was more threatening than the converse, regardless of distance or sex.

Fig. 3.
Mean values for levels of factors. Solid lines indicate contrasts that were significant at the individual p < .05 level. Dotted lines indicate contrasts that were not significant at the individual p < .05 level.

Discussion

The article reports findings on how the factors of posture, distance, and proximity mitigate perceived threat. Many other factors could have been investigated, e.g., cross-cultural differences were not explored because the present laboratory does not have access to cross-cultural samples. For readers interested in this topic, reviews have been provided by Matsumoto (2006) and Anderson (2008, pp. 77–106). The choice of limiting the number of factors studied was based on the requirements of formal experiment design, e.g., it is quite possible that other factors, such as relationships among interactants, the reward valence of the interactants, the personality of the observer, the situation, the culture from which the interactants are, prior relationship, and psychological state, might mitigate perceived threat. However, when added to the factors already included in this study, the number of stimuli increases to the point of being infeasible, e.g., with only two levels for each of the seven factors listed above, there would be 128 cells within the 24 cells already in the study, for a total of 3,072 cells. Possible sources of ideas for further inquiry can be found in, among other sources, Petronio (2002), Burgoon, Stern, and Dillman (1995), or Knapp, Hall, and Horgan (2014).

Permeability theory predicts that, when assessing threat from other people, distance to the source of threat should be important. If there are multiple sources of threat in an environment, priority should be given to the closest source. This implies that, when there are multiple threats, there should be a form of perceptual triage in action. There is a substantial amount of prior experimental work on how humans communicate threat at different distances, but that work was obtained from environments containing only one person. This work extends the experimental literature on conspecific danger to include the factor of proximity. Overall, it seems that the current experiment adds the following information to the collective record: For Hypotheses 3 (sex has an effect on perceived threat), 4 (posture has an effect on perceived threat), and 5 (distance has an on perceived threat), previous findings were clearly replicated. The new finding is Hypothesis 1: proximity has an effect on perceived threat when threat was communicated by the non-verbal cues of posture and sex over distances from 5 to 20 m.

Stim	Distances
A	H	M	H	M
B	H	F	B	M
C	B	M	H	F
D	B	F	B	M
E	H	M			H	M
F	H	F			B	M
G	B	M			H	F
H	B	F			B	F
I			H	M	H	M
J			H	F	B	M
K			B	F	H	F
L			B	F	B	F
M	H	F	H	F
N	H	M	B	F
O	B	F	H	M
P	B	M	B	F
Q	B	F			H	F
R	H	M			B	F
S	H	F			H	M
T	B	M			B	M
U			H	F	H	F
V			H	M	B	F
W			B	M	H	M
X			B	M	B	M

Stim	M	SD	Stim	M	SD
A	5.83	1.97	M	5.38	2.00
B	5.17	1.97	N	4.83	1.71
C	3.25	1.98	O	4.42	1.86
D	2.75	1.26	P	2.71	1.57
E	6.25	2.13	Q	4.42	2.02
F	3.54	1.82	R	6.25	1.94
G	4.46	2.02	S	3.58	2.00
H	2.67	1.81	T	3.38	1.81
I	4.79	1.98	U	3.50	2.02
J	3.13	1.70	V	3.33	1.69
K	2.46	1.96	W	2.92	1.89
L	2.42	1.82	X	2.50	1.22

References

Aiello

J. R.

(1976) Visual interaction at extended distances. Personality Social Psychology Bulletin, 3, 83–86.

Aiello

J. R.

(1977) A further look at equilibrim theory: Visual interaction as a function of interpersonal distance. Environmental Psychology and Nonverbal Behavior, 1(2), 122–140.

Allen

G. L.

, & Rashotte

M. A.

(2006) Training metric accuracy in distance estimation skill: Pictures vs words. Applied Cognitive Psychology, 20, 171–186.

Anderson

(1978) Visual resource assessment: Local perceptions of familiar natural environments. In Kaplan

& Kaplan

(Eds.), The experience of nature: A psychological perspective (1989 ed.,). Cambridge, England: Cambridge Univer. Press. Pp. 217–220.

Anderson

, & McOwnan

P. W.

(2006) A real-time automated system for the recognition of human facial expressions. IEEE Transactions on Systems, Man, and Cybernetics-Part B Cybernetics, 36 (1), 96–105.

Anderson

P. A.

(2008) Nonverbal communication: Forms and functions. Long Grove, IL: Waveland Press.

Argyle

(1988) Bodily communication. New York: International Universities Press.

Bailenson

J. N.

Pontikakis

E. D.

Mauss

I. B.

Gross

J. J.

Jabon

M. E.

Hutcherson

A. C.

Oliver

(2008) Real-time classifiction of evoked emotions using facial tracking and physiological responses. International Journal Human-Computer Studies, 66, 303–317.

Bartlett

M. S.

Hager

J. C.

Ekman

, & Sejnowski

T. J.

(1999) Measuring facial expressions by computer image analysis. Psychophysiology, 36, 253–265.

10.

Bethel

C. L.

, & Murphy

R. P.

(2008) Survey of non-facial/nonverbal affective expressions for appearance-constrained robots. IEEE Transactions on Systems, Man and Cybernetics Part C Applications and Reviews, 38, 83–92.

11.

Bull

P. E.

(1987) Posture and gesture. Oxford: Pergamon.

12.

Burgoon

J. K.

(1991) Relational message interpretations of touch, conversational distance, and posture. Journal of Nonverbal Behavior, 15 (4), 233–259.

13.

Burgoon

J. K.

Stern

L. A.

, & Dillman

(1995) Interpersonal Adaptation: Dyadic interaction patterns. Cambridge, England: Cambridge Univer. Press.

14.

Coulson

(2004) Attributing emotion to static body posture: Recognition accuracy, confusions, and viewpoint dependence. Journal of Nonverbal Behavior, 28 (2), 117–139.

15.

Curious Labs. (2002) Poser 5. Santa Cruz, CA: Author.

16.

Darwin

(1872) The expression of the emotions in man and animals. Chicago: Univer. of Chicago Press.

17.

De Silva

P. R.

, & Bianchi-Berthouze . (2004) Modeling human affective postures: An information theoretic characterization of posture features. Computer Animation and Virtual Worlds, 15, 269–276.

18.

Ducourant

Vielledent

Kerlirzin

, & Berthoz

(2005) Timing and distance characteristics of interpersonal coordination during locomotion. Neuroscience Letters, 389 (1), 6–11.

19.

Etcoff

N. L.

, & Magee

J. J.

(1992) Categorical perception of facial expressions. Cognition, 44, 227–240.

20.

Floyd

(2006) An evolutionary approach to understanding nonverbal communication. In Manusov

& Patterson

M. L.

(Eds.), The Sage Handbook of nonverbal communication. Thousand Oaks, CA: Sage. Pp. 139–157.

21.

Goffman

(1971) Relations in public. New York: Basic Books.

22.

Greene

M. R.

, & Oliva

(2009) The briefest of glances: The time course of natural scene understanding. Psychological Science, 20 (4), 464–472.

23.

Hall

E. T.

(1966) The hidden dimension. New York: Doubleday.

24.

Hansen

C. H.

, & Hansen

R. D.

(1988) Finding the face in the crowd: An anger superiority effect. Journal of Personality and Social Psychology, 54 (6), 917–924.

25.

Hayashibe

(2002) Apparent distance in actual, three-dimensional video-recorded, and virtual reality. Perceptual and Motor Skills, 95, 573–582.

26.

Hediger

(1955) The psychology and behaviour of animals in zoos and circuses ( Sircom

, Transl. 1968 ed.). New York: Dover.

27.

Kaitz

Bar-Haim

Lehrer

, & Grossman

(2004) Adult attachment style and interpersonal distance, Attachment and Human Development, 6(3), 285–304.

28.

Knapp

M. L.

Hall

J. A.

, & Horgan

T. G.

(Eds.). (2014) Nonverbal communication in human interaction. Boston, MA: Wadsworth.

29.

Kohler

C. G.

Turner

Stolar

N. M.

Bilker

W. B.

Brensinger

C. M.

Gur

R. E.

, & Cur

R. C.

(2004) Differences in facial expressions of four universal emotions. Psychiatry Research, 128, 235–244.

30.

LeDoux

(1996) The emotional brain: The mysterious underpinnings of emotional life. New York: Simon & Schuster.

31.

LeDoux

(1998) Fear and the brain: Where have we been, and where are we going? Biological Psychiatry, 44, 1229–1238.

32.

LeDoux

(2003) The emotional brain, fear, and the amygdala. Celluar and Molecular Neurobiology, 23(4/5), 727–738.

33.

, & Li

(2007) How far is far enough? Environment & Behavior, 39 (3), 317–331.

34.

Lombard

(1995) Direct responses to people on the screen-television and personal space. Communication Research, 22 (3), 288–324.

35.

Lundqvist

, & Öhman

(2005) Emotion regulates attention: The relation between facial configurations, facial emotion, and visual attention. Visual Cognition, 12, 51–84.

36.

Matsumoto

(2006) Culture and nonverbal communication. In Manusov

& Patterson

M. L.

(Eds.), The Sage handbook of non-verbal communication. Thousand Oaks, CA: Sage. Pp. 219–235.

37.

Matsumoto

, & Kudoh . (1987) Cultural similarities and differences in the semantic dimensionbs of body postures. Journal of Nonverbal Behavior, 11(3), 166–179.

38.

Meeren

K. M.

van Heijnsbergen

C. C.

, & de Gelder

(2005) Rapid perceptual integration of facial expression and emotional body language. Proceedings of the National Academy of Science, 102 (45), 16518–16523.

39.

Mok

Wellman

, & Basu

(2007) Did distance matter before the internet? Interpersonal contact and support in the 1970s. Social Networks, 29(3), 430–461.

40.

Montepare

J. M.

(2003) Introduction evolution and nonverbal behavior: Adaptive social perceptions. Journal of Nonverbal Behavior, 27 (2), 61–64.

41.

Mortier

R. S.

(2003) The Poser 5 Handbook. Hingham, MA: Charles River Media.

42.

Öhman

(1986) Face the beast and fear the face: Animal and social fears as prototypes for evolutionary analysis of emotion. Psychophysiology, 23(2), 123–145.

43.

Öhman

(2008) Fear and anxiety: Overlaps and dissociations. In Lewis

Haviland-Jones

J. M.

& Feldman

L. B.

(Eds.), Handbook of emotions. New York: Guilford Press. Pp. 709–729.

44.

Öhman

, & Mineka

(2001) Fears, phobias, and preparedness: Toward an evolved module of fear and fear learning. Psychological Review, 108(3), 483–522.

45.

Ozdemir

(2008) Shopping malls: Measuring interpersonal distance under changing conditions and across cultures. Field Methods, 20 (3), 226–248.

46.

Patterson

M. L.

(2003) Commentary evolution and nonverbal behavior: Functions and mediating processes. Journal of Nonverbal Behavior, 27 (3), 201–207.

47.

Petronio

(2002) Boundaries of Privacy: Dialectics of disclosure. Albany, NY: State Univer. of New York Press.

48.

Pitterman

, & Nowicki

(2004) A test of the ability to identify emotion in human and standing postures: The diagnostic analysis of nonverbal accuracy-2 Posture Test (DANAVA2-POS). Genetic, Social and General Psychology Monographs, 130(2), 146–162.

49.

Planalp

(1998) Communicating emotion in everyday life: Cues, channels, and processes. In Anderson

P. A.

& Guerrero

L. K.

(Eds.), Handbook of communication and emotion: Research, theory, applications, and contexts. San Diego, CA: Academic. Pp. 29–48.

50.

Sommer

(2002) Personal space in a digital age. In Bechtel

R. B.

& Churchman

(Eds.), Handbook of environmental psychology. New York: Wiley. Pp. 647–660.

51.

Stamps

A. E.

(2000) Psychology and the aesthetics of the built environment. Norwell, MA: Kluwer Academic.

52.

Stamps

A. E.

(2010) Use of static and dynamic media to simulate environments: A meta-analysis. Perceptual and Motor Skills, 111(2), 1–12.

53.

Stamps

A. E.

(2011a) Distance mitigates perceived threat. Perceptual and Motor Skills, 113 (3), 1–13.

54.

Stamps

A. E.

(2011b) Validation of computer simulations for effects of eye gaze, sex, facial expression, and posture on perceived threat. Perceptual and Motor Skills, 112 (3), 1–8.

55.

Stamps

A. E.

(2012) Resume, Arthur E Stamps III. Retrieved from http://home.comcast.net/∼InstituteOfEnvironmentalQuality/RESUME.pdf

56.

Stamps

A. E.

(2013) Validation data for evaluating effects of environments on affective responses. San Francisco, CA: Institute of Environmental Quality. Retrieved fromhttp://www.home.comcast.net/∼instituteofenvironmentalquality/validationData.pdf

57.

Thompson

D. E.

Aiello

J. R.

, & Epstein

Y. M.

(1979) Interpersonal distance preferences. Journal of Nonverbal Behavior, 4 (2), 113–118.

58.

Walk

R. D.

, & Walters

K. L.

(1988) Perception of the smile and other emotions of the body and face at difference distances. Bulletin of the Psychonomic Society, 26 (6), 510.

59.

Witmer

B. G.

, & Kline

P. B.

(1998) Judging perceived and traversed distance in virtual environments. Presence, 7 (2), 144–167.

Stim	Distances
	5 m		10 m		20 m
	Posture	Sex	Posture	Sex	Posture	Sex
A	H	M	H	M
B	H	F	B	M
C	B	M	H	F
D	B	F	B	M
E	H	M			H	M
F	H	F			B	M
G	B	M			H	F
H	B	F			B	F
I			H	M	H	M
J			H	F	B	M
K			B	F	H	F
L			B	F	B	F
M	H	F	H	F
N	H	M	B	F
O	B	F	H	M
P	B	M	B	F
Q	B	F			H	F
R	H	M			B	F
S	H	F			H	M
T	B	M			B	M
U			H	F	H	F
V			H	M	B	F
W			B	M	H	M
X			B	M	B	M

Mitigation of Threat by Posture,Distance,and Proximity 1

Abstract

Permeability Theory

Non-verbal Communication by Posture and Distance

Hypotheses

Method

Stimuli and Experimental Design

Sample Size and Participants

Task

Results

Discussion

References