Primates (2007) 48:1–12 DOI 10.1007/s10329-006-0011-4 O R I G I N A L A RT I C L E Visual search for orientation of faces by a chimpanzee (Pan troglodytes): face-specific upright superiority and the role of facial configural properties Masaki Tomonaga Received: 3 February 2006 / Accepted: 4 July 2006 / Published online: 13 September 2006 Japan Monkey Centre and Springer-Verlag 2006 Abstract A previous experiment showed that a chimpanzee performed better in searching for a target human face that differed in orientation from distractors when the target had an upright orientation than when targets had inverted or horizontal orientation [Tomonaga (1999a) Primate Res 15:215–229]. This upright superiority effect was also seen when using chimpanzee faces as targets but not when using photographs of a house. The present study sought to extend these results and explore factors affecting the face-specific upright superiority effect. Upright superiority was shown in a visual search for orientation when caricaturized human faces and dog faces were used as stimuli for the chimpanzee but not when shapes of a hand and chairs were presented. Thus, the configural properties of facial features, which cause an inversion effect in face recognition in humans and chimpanzees, were thought to be a source of the upright superiority effect in the visual search process. To examine this possibility, various stimuli manipulations were introduced in subsequent experiments. The results clearly show that the configuration of facial features plays a critical role in the upright superiority effect, and strongly suggest similarity in face processing in humans and chimpanzees. M. Tomonaga (&) Language and Intelligence Section, Primate Research Institute, Kyoto University, Inuyama, Kanrin 484-8506, Japan e-mail: [email protected] Introduction Faces are quite important stimuli for primates, including humans. We obtain a variety of information from the face, including identity, expression, sex, age, and attention. In humans, it is well known that face processing is quite different from the processing of other visual stimuli. One of the most striking phenomena in face perception is deteriorated recognition when faces are presented in an inverted orientation (Valentine 1988; Yin 1969). The inversion effect is generally limited to facial stimuli (cf. Diamond and Carey 1986) and is considered specific to faces because we process the face as a whole (Diamond and Carey 1986; Farah et al. 1995; Tanaka and Farah 1991). Faces consist of several features, such as the eyes, nose, and mouth. We recognize facial identity not on the basis of each feature separately but on the configural properties of the features. Diamond and Carey (1986) also pointed out that the configural properties of objects can be classified into two types: first-order relational information, which consists of the spatial relations of the parts of an object with respect to one another, and second-order relational information, which consists of the spatial relations of the parts relative to the prototypical arrangement of the parts (see also Farah et al. 1995). Humans form face prototypes during prolonged experience of a variety of faces, and as a result, they can encode second-order relational information in faces (Diamond and Carey 1986). Numerous past studies have explored face processing in nonhuman primates (Bruce 1982; Dittrich 1990; Hamilton and Vermeire 1983, 1988; Kanazawa 1996; Keating and Keating 1993; Martin-Malivel and Fagot 2001; Matsuzawa 1990; Overman and Doty 1982; Parr 123 2 Primates (2007) 48:1–12 et al. 1998, 1999; Perrett and Mistlin 1990; Phelps and Roberts 1994; Rosenfeld and Van Hoesen 1979; Tomonaga 1994, 1999a, b; Tomonaga et al. 1993; Wright and Roberts 1996). Several decades ago, face inversion effect studies were inconclusive in nonhuman primates (Bruce 1982; Dittrich 1990; Rosenfeld and Van Hoesen 1979), but recent progress in this area has confirmed that nonhuman primates perceive faces much as humans do (Hamilton and Vermeire 1988; Martin-Malivel and Fagot 2001; Overman and Doty 1982; Parr et al. 1998; Phelps and Roberts 1994; Tomonaga 1994, 1999b; Wright and Roberts 1996). For example, Tomonaga (1999b), using a delayed matching paradigm, reported that chimpanzees (Pan troglodytes) clearly showed the inversion effect when human faces were used as stimuli, but not when houses were used. Parr et al. (1998) obtained similar results in chimpanzees when using chimpanzee, capuchin, and human faces as stimuli. Martin-Malivel and Fagot (2001) reported the inversion effect in baboons (Papio papio) using the priming paradigm. Several studies have also reported the inversion effect in face perception by macaques (e.g., Hamilton and Vermeire 1983, 1988; Overman and Doty 1982; Phelps and Roberts 1994; Wright and Roberts 1996). Simple discrimination and matching tasks have been commonly used to examine the inversion effect (Bruce 1982; Parr et al. 1998; Tomonaga 1999b; Yin 1969). However, visual search tasks have also been increasingly employed to study face perception and recognition (Hansen and Hansen 1988; Levin 1996; Lewis and Edmonds 2005; Montoute and Tiberghien 2001; Nothdurft 1993; Öhman et al. 2001; Senju et al. 2005; Suzuki and Cavanagh 1995; Tong and Nakayama 1999; von Grünau and Anston 1995; Williams et al. 2005). In a typical visual-search task, participants are required to report the presence or absence of a predefined target among distractors (Treisman and Gelade 1980). When using animal subjects, the tasks are slightly modified, and subjects are required to detect and respond to 123 TGT: Inv Response Time (s) Fig. 1 Upright superiority effect for orientation during visual search of faces (Tomonaga 1999a). A chimpanzee subject showed a more efficient search for the upright human face target among disoriented faces. From Visual search for orientation of faces by a chimpanzee (Pan troglodytes) by Tomonaga (1999a) (peck, point, or touch) a target (Blough 1979; Tomonaga 1993a, 2001a). Using this visual-search task, I previously investigated the inversion effect with a chimpanzee as the subject (Tomonaga 1999a). An adult chimpanzee was trained for a visual-search task for photograph orientation. The subject was required to detect and touch the target differing in orientation from the distractors. Using human faces in three different orientations (upright, horizontal, and inverted), the chimpanzee showed better performance when the upright face appeared as the target against horizontal or inverted distracters than vice versa (Fig. 1; left and center panels).In contrast, when horizontal and inverted faces were paired, such visual search asymmetry was not obtained (Fig. 1; right panel). Subsequent experiments in which photographs of chimpanzee faces and houses were used in addition to human faces further confirmed and extended this upright superiority effect in visual search. The chimpanzee showed the upright superiority effect for chimpanzee faces but not for houses. These results indicate that the upright superiority effect, defined as a more efficient search performance for upright stimuli than for other orientations, is specific to facial stimuli, and that this effect seems to be closely related to the inversion effect in face recognition by humans and chimpanzees (Parr et al. 1998; Tomonaga 1999b; Valentine 1988; Yin 1969). Tomonaga (2001a) replicated this upright superiority effect in visual search using another chimpanzee as a subject. This chimpanzee found the human face more quickly when the target was upright than when the target was horizontal or inverted. The present study aimed to extend the results of the previous study (Tomonaga 1999a). In Experiments 1 and 4, the specificity of the upright superiority effect to facial stimuli was further tested using various types of facial stimuli (human caricaturized faces, dog faces, and schematic faces) and non-face stimuli (human hand shapes, chairs, and schematic faces with scram- 1.6 Up vs. Inv Up vs. Hor Hor vs. Inv 1.4 TGT: Hor TGT: Inv 1.2 TGT: Hor 1.0 TGT: Up 0.8 TGT: Up 0.6 0.4 1 3 6 12 1 3 6 12 Number of Stimuli 1 3 6 12 Tomonaga (1999a) Primates (2007) 48:1–12 bled features). To test the possibility that the facespecific upright superiority effect is caused by the configural processing of faces by the chimpanzee, each facial part (eyes, nose, and mouth) was presented alone or in combination with the spatial configuration (Experiments 2–4). General methods Subject An adult female chimpanzee (Pan troglodytes), Chloe, participated in the present experiments (see Fig. 2). At the onset of the experiments, she was 17 years old. She has experienced various types of computer-controlled perceptual–cognitive tasks (Tomonaga 1993b, 1999c; Tomonaga and Matsuzawa 1992), including visual search tasks (Fagot and Tomonaga 1999; Tomonaga 1993a, 1995, 1997, 2001a, 2001b). She has also experienced face recognition tasks using a matching paradigm (Tomonaga 1999b). Immediately before the onset of the present experiments, Chloe had been tested in visual search for orientation of photographs including human and chimpanzee faces and had exhibited the face-specific upright superiority effect, as mentioned in the introduction (Tomonaga 1999a). Chloe lives in a social group of 11 individuals in an environmentally enriched outdoor compound (770 m2; Ochiai and Matsuzawa 1998) connected to the experimental room by a tunnel. No special food or water deprivation was conducted during the present study. Care and use of the chimpanzees adhered to the 2002 version of the Guide for Care and Use of Laboratory Primates by the 3 Primate Research Institute, Kyoto University, Japan. The research design was approved by the Animal Welfare and Animal Care Committee of the Primate Research Institute. Apparatus Sessions were conducted inside an experimental booth for chimpanzees (1.8·2.15·1.75 m). A 21-inch color CRT monitor (NEC PC-KH2021) with a capacitive touchscreen device (Microtouch SM-T2) was installed 15 cm from the floor on one side of the booth. Touching the monitor surface with a finger was defined as a response. The screen was protected from deterioration by a transparent Plexiglas panel, fitted with an armhole (10·47 cm) that allowed hand contact with the CRT. The resolution of the monitor was 640·400 pixels. One hundred pixels corresponded to 55 mm. The subject sat approximately 40 cm away from the monitor surface, so 100 pixels corresponded to 8 of the visual angle, and the length of 1 corresponded to approximately 7 mm (13 pixels). A food tray was installed below the CRT. A universal feeder (Biomedica BUF-310) delivered pieces of food (apples or raisins) to this tray. The equipment was connected to a personal computer (NEC PC-9821 Xn) that controlled the stimulus display, detected touches on the CRT, delivered rewards, and collected data. Procedure Figure 2 shows a typical visual search task trial. Each trial began with presentation of the start key (blue circle, 30 pixels in diameter) at the bottom center of Fig. 2 A chimpanzee, Chloe, performing a visual search for orientation of faces 123 4 the CRT monitor and an accompanying 0.5-s beep. Immediately after the subject responded, the start key disappeared, and the search display appeared on the screen. The search display consisted of a 4·3 predefined stimulus presentation area and contained one target stimulus and several uniform distractors different from the target but identical to each other (see Fig. 2). Configurations of the search display randomly changed from trial to trial. The target position was counterbalanced. The task for the subject was to detect the target and touch it on the screen. When the chimpanzee touched the target, all the stimuli except the target disappeared, a chime sounded, and a food reward (small pieces of apple or raisins) was delivered to the food tray followed by the termination of the touched stimulus. When the chimpanzee made an error (i.e., touched one of the distractors), a buzzer sounded, and the identical search display was represented once again; if a novel incorrect response was given, only the target stimulus was represented on the screen in a second correction trial. This correction procedure was introduced because previous experiments have demonstrated that chimpanzees may stop working when the rate of non-reinforced trials is too high (cf. Fagot and Tomonaga 1999). Experiment 1 In the first experiment, Chloe was tested with various kinds of photographs including face and non-face stimuli to extend the previous results by showing the face-specific upright superiority effect in visual search for a stimulus by orientation. Methods In Experiment 1, five types of 256-step grayscale photographs were used as stimuli (see Fig. 3): human faces, human caricature faces, dog faces, human hand shapes, and chairs. Each stimulus was 100·100 pixels in size. For human faces, frontal views of 104 Japanese males were used. These faces were identical to those used in previous experiments (Tomonaga 1999a, b). For each of the other categories, 36 different photographs were prepared. Various types of hand signs were prepared (such as a ‘‘V’’ sign) for the hand stimulus. For each stimulus, global luminance was roughly equalized and retouched onto the gray background using Paintshop Pro 3.0. Each stimulus was prepared with three different orientations: upright (up), horizontal (hor), and inverted (inv). By combining these orientations, six target–distractor pairs of orientation were prepared. 123 Primates (2007) 48:1–12 Each session consisted of 60 trials. In each session, five types of stimuli and two numbers of stimuli (i.e., the numbers of a target and distractors in the search display: 4 and 10) appeared equally and randomly, but the target–distractor combination of orientation was fixed. The subject received two sessions in a day; in each session, the target and distractor orientations were reversed. Each type of session was presented in a random order and repeated more than 16 times. To analyze the subject’s steady-state performance, the last eight sessions for each condition were used for data analyses. Results and discussion Table 1 shows the mean percentage of errors for each condition. Chloe clearly made more errors for non-face than for face stimuli and for horizontal versus inverted (hor vs. inv) than for the other orientation combinations. A two-way analysis of variance (ANOVA) with orientation (up vs. inv, up vs. hor, and hor vs. inv, averaged across the target–distractor combination and number of stimuli) as the between-session factor and stimuli (five types) as the within-session factor, confirmed that both main effects [orientation, F(2,21)=17.43, P<0.001; stimuli; F(4,84)=133.15, P<0.001] and interaction [F(8,84)=6.56, P<0.001] were significant. Post hoc multiple comparisons using Ryan’s procedure indicated that errors for three types of facial stimuli were significantly less than for the other nonfacial stimuli (P<0.001). Among the three types of facial stimuli, Chloe made more errors for dog faces than for human and caricature faces (P<0.01). Furthermore, there were significantly more errors for the hor vs. inv condition than for the up vs. inv or up vs. hor conditions, when only the three types of facial stimuli were presented (P<0.05). Figure 3 shows the mean response times for each condition. As shown in Table 1, the subject performed very poorly under some conditions; the response times from incorrect trials (correction trials were excluded) were also used for analyses. Incorrect response times were not used if the subject showed a ‘‘speed–accuracy trade-off,’’ i.e., a faster (guessing) response in incorrect trials than in correct trials. However, Chloe showed longer response times on incorrect trials than on correct trials, so that response times averaged across ‘‘total’’ trials were a good measure in this study (cf. Tomonaga and Matsuzawa 1992). Thus, subsequent analyses used response times averaged across correct and incorrect trials. Separate two-way (target orientation · number of stimuli) ANOVAs with sessions as repeated measures were conducted for response-time Primates (2007) 48:1–12 5 Fig. 3 Mean response times for each condition in Experiment 1. Examples of each stimulus set are also shown in each panel. **P<0.01; ***P<0.001. TGT Target, DST distractor, Up upright, Hor horizontal, Inv inverted 3.5 Human Face 3.0 2.5 2.0 ** 1.5 *** TGT/DST 1.0 Up/Inv Inv/Up Response Times (s) 0.5 3.5 Human Caricature Face 3.0 2.5 *** 2.0 Up/Hor Hor/Up Hor/Inv Inv/Hor Human Hand p=0.061 *** 1.5 1.0 0.5 Dog Face 3.5 3.0 p=0.079 2.5 Chair *** 2.0 1.5 1.0 Up vs. Inv Up vs. Hor Hor vs. Inv Up vs. Inv Up vs. Hor Hor vs. Inv 0.5 4 10 4 10 4 10 4 10 4 10 4 10 Number of Stimuli Table 1 Mean percent errors for each condition in Experiment 1. Up Upright, Hor horizontal, Inv inverted Stimulus set Number of stimuli Up vs. inv Up/inv Human face Caricature face Dog face Human hand Chair a 4 10 Average 4 10 Average 4 10 Average 4 10 Average 4 10 Average 2.1 6.3 4.2 4.2 6.3 5.2 22.9 6.3 14.6 52.1 35.4 43.8 77.1 85.4 81.3 a Up vs. hor Hor vs. inv Inv/up Average Up/hor Hor/up Average Hor/inv Inv/hor Average 18.8 14.6 4.2 35.4 10.4 22.9 25.0 12.5 18.8 29.2 39.6 34.4 54.2 52.1 53.1 10.4 10.4 10.4 19.8 8.3 14.1 24.0 9.4 16.7 40.6 37.5 39.1 65.6 68.8 67.2 4.2 6.3 5.2 0.0 4.2 2.1 6.3 8.3 7.3 39.6 22.9 31.3 54.2 70.8 62.5 10.4 0.0 5.2 12.5 2.1 7.3 18.8 14.6 16.7 37.5 33.3 35.4 56.3 50.0 53.1 7.3 3.1 5.2 6.3 3.1 4.7 12.5 11.5 12.0 38.5 28.1 33.3 55.2 60.4 57.8 37.5 16.7 27.1 20.8 4.2 12.5 31.3 29.2 30.2 27.1 16.7 21.9 52.1 66.7 59.4 25.0 16.7 27.1 45.8 27.1 36.5 50.0 43.8 46.9 33.3 31.3 32.3 58.3 75.0 66.7 31.3 16.7 24.0 33.3 15.6 24.5 40.6 36.5 38.5 30.2 24.0 27.1 55.2 70.8 63.0 Target/distractor data of each of the trials. Table 2 summarizes the results of these ANOVAs. As shown in Fig. 3 and Table 2, the subject showed longer response times when the number of stimuli was larger. Furthermore, the subject’s behavior was significantly affected by the target orientation when the facial stimuli were used, but this effect was seen only when the upright orientation and another orientation were paired. For all 123 6 Table 2 Summary of ANOVAs (F-statistics) conducted on the responsetime data from Experiment 1. TGT target orientation, N number of stimuli Primates (2007) 48:1–12 Stimulus set Factor Up vs. inv Up vs. hor Hor vs. inv Human face TGT N TGT TGT N TGT TGT N TGT TGT N TGT TGT N TGT 13.21**a 21.91** 3.48 30.09*** 20.12** 1.49 4.21P = 0.079 94.21*** 0.04 3.91 22.83** 0.47 0.65 23.48** 0.11 74.66*** 37.37*** 10.61* 81.16*** 21.90** 4.88 P = 0.063 6.47* 18.41** 0.28 3.12 54.23*** 0.20 3.69 59.63*** 0.01 0.03 14.11** 1.59 4.96P = 0.061 53.78*** 3.14 0.00 28.26** 2.00 0.74 19.77** 2.31 2.42 13.21** 2.56 Caricature face Dog face Human hand a Degree of freedom for each F-value was 1,7 *P<0.05, **P<0.01, ***P<0.001 Chair facial stimuli, Chloe showed faster response times for upright targets than for other target orientations. The results of Experiment 1 clearly replicated and extended the results of the previous study (Tomonaga 1999a). The same subject as in Tomonaga (1999a) demonstrated the upright superiority effect for orientation during visual search only when facial stimuli were used. These results provide further evidence that chimpanzees process facial stimuli using configural properties of facial features. Furthermore, the ability to recognize a stimulus as a ‘‘face’’ is not limited to own-species or highly familiar types of stimuli (e.g., human and chimpanzee faces) but applicable broadly to stimuli that contain facial configurations (e.g., caricature and dog faces). If this is correct, chimpanzees should show similar results when the much simpler ‘‘facial’’ patterns that are frequently used in studies with human adults and infants (e.g., Johnson and Morton 1991; Nothdurft 1993; Öhman et al. 2001) are used. Experiment 4 examined this possibility. The subject’s more efficient detection of various types of upright face in the present experiment is inconsistent with some previous experiments. Parr et al. (1998) reported that chimpanzees showed a clear inversion effect for human and chimpanzee faces but not for capuchin faces when using the rotational matching paradigm. Wright and Roberts (1996) found that rhesus macaques showed the inversion effect for human faces but not for monkey faces. Phelps and Roberts (1994) also reported that squirrel monkeys showed the inversion effect only for human faces but not for monkey faces. There was, however, quite a large difference in the task demand between the present study and these previous experiments. The previous experiments required subjects to perform individual recognition (or picture matching), while the present experiment required the chimpanzee to 123 ·N ·N ·N ·N ·N discriminate the orientation of stimuli (cf. Ballaz et al. 2005; Jitsumori and Matsuzawa 1991). This critical difference in task demand may have been responsible for the different outcomes. Experiment 2 Experiment 1 further indicated that the upright superiority in visual search for orientation in the chimpanzee was face-specific. As noted in the introduction, this specificity may have reflected configural processing of faces by the chimpanzee subject. In Experiment 2 the facial stimuli were manipulated to examine this possibility directly. Methods In this experiment, facial features (eyes, nose, and mouth) were presented alone or in combination while maintaining the spatial configurations (see Fig. 4). Sixty new Japanese male faces were prepared and manipulated using Paintshop Pro 3.0. Among six types of manipulated stimuli (test stimuli) presented on the gray background, only single features were presented for the three types in which each feature was mounted at the center; two types of features were paired in the other three types of stimuli. These stimuli were arranged in two orientations, upright and inverted. Each session consisted of 72 trials. In half of the trials, original (intact) faces were presented, while in the other half of the trial’s test stimuli were presented. Original and test stimuli appeared alternately in a session. The order of original and test stimuli was randomly counterbalanced. In this experiment, only one number of stimuli (10) was used. As in Experiment 1, the chimpanzee received two successive sessions in a Primates (2007) 48:1–12 4.0 * Response Times (s) Fig. 4 Mean response times for each condition in Experiment 2. *P<0.05; **P<0.01; ***P<0.001 7 3.5 *** 3.0 2.0 TGT/DST * 2.5 Up/Inv Inv/Up ** 1.5 1.0 0.5 Original Nose Eyes (with Eyebrows) day. The target orientation was reversed between each session. If the subject made an incorrect response, in the next correction trial only the target was presented to minimize the possibility of learning during sessions. Chloe received eight sessions for each target-orientation condition. Results and discussion On average, the subject correctly detected a target in 96.2, 56.3, 18.8, 11.5, 86.5, 89.6, and 38.5% of original, eyes, nose, mouth, eyes + nose, eyes + mouth, and nose + mouth trials, respectively. The results of oneway ANOVA showed a significant main effect of stimulus type [F(5,35) =75.39, P<0.001]. Post hoc multiple comparisons (Ryan’s procedure) clearly showed no significant difference between performances for nose and mouth and between eyes + nose and eyes + mouth. Figure 4 shows the mean response times (averaged for correct and incorrect trials) for each condition. As shown in this figure, the upright superiority effect was seen only when the eye region was presented alone or combined with other features. Independent paired t-tests verified this inspection [original, t(7)=5.166, P<0.01; eyes, t(7)=3.492, P<0.05; eyes + nose, t(7)=7.851, P<0.001; eyes + mouth, t(7)=3.063, P<0.05]. The results of Experiment 2 indicated that the upright superiority effect was stronger when the facial features were presented with their spatial configuration intact than when the features were presented alone. These results support the hypothesis that the upright superiority effect in the chimpanzee is based on the configural processing of faces. Furthermore, the eye region was necessary for the upright superiority effect to occur. The Mouth Eyes + Nose Eyes + Mouth Nose + Mouth latter finding is quite interesting given that primates, including humans, generally pay attention to the eyes when viewing the face (Gothard et al. 2004; Hunnius and Geuze 2004; Keating and Keating 1993; Luria et al. 1964; Nahm et al. 1998; Sato and Nakamura 1999; Yarbus 1967). It is plausible that chimpanzees may also pay attention predominantly to the eye region of the facial image given the finding that chimpanzee infants prefer direct-gaze human faces to averted-gaze faces (MyowaYamakoshi et al. 2003). Chloe showed upright superiority when the eye region was presented alone. However, we should interpret these results with caution because the eyes stimuli used in the present experiment combined eyes and eyebrows. The spatial configuration of the eyes and eyebrows may be sufficient to cause the upright superiority effect. This possibility was further tested in Experiment 4. Experiment 3 Experiment 3 tested the effects of outer facial contours. In the previous experiments, the facial stimuli included hair, ears, and facial contours, unlike those commonly used for human experiments, in which these features are removed by trimming with an elliptic window (e.g., Senju et al. 2005; Yamaguchi et al. 1995). Human adults recognize familiar faces better by their internal parts than their external parts (Ellis et al. 1979), but younger children show the opposite tendency (Campbell et al. 1995). To rule out the possibility that the chimpanzee predominantly used external parts during the visual-search task, external and internal parts were presented separately in Experiment 3. 123 8 Primates (2007) 48:1–12 Methods Experiment 4 The 60 original faces were the same as those used in Experiment 2. In the external parts stimuli, the inner facial features were replaced by a gray ellipse. In the internal parts stimuli, only the inner facial features were presented in an elliptic contour on the gray background (see Fig. 5). Each session consisted of 72 trials, of which half were the original face trials and the other half were the test (external and internal stimuli, six trials each) and catch trials (in which no target was presented). The procedure was the same as in Experiment 2. In the final experiment, two sets of stimuli were used for additional testing. The first set was of schematic faces (and those with features in scrambled patterns) to extend the results of Experiment 1. The second set was eyes + eyebrows (the same as in the eyes stimuli in Experiment 2) and eyes-only stimuli, in which the eyebrows had been removed from the eyes + eyebrows stimuli. These were used to confirm the results of Experiment 3. Results and discussion The original faces were the same as those in Experiments 2 and 3. For schematic faces, three different eye parts and two different mouthparts were combined to yield six different schematic faces (see Fig. 6). Each facial feature was randomly arranged to create six corresponding scrambled patterns. The eyes + eyebrows stimuli were the same as the eyes stimuli used in Experiment 2. For the eyesonly stimuli, the eyebrow region was removed from each of the eyes + eyebrows stimuli (see Fig. 6). Each session consisted of 48 trials. In half of the trials, original faces were presented and in the other half, test stimuli were presented. The procedure was the same as in Experiments 2 and 3. Chloe correctly detected a target in 96.5, 52.1, and 88.5% of trials on average for the original, external, and internal stimuli, respectively. She showed significantly better accuracy for internal than for external stimuli [t(7)=6.999, P<0.001]. Figure 5 shows the mean response times averaged across correct and incorrect trials for each condition. The upright superiority effect was significant only for internal stimuli [t(7) = 6.685, P < 0.001], suggesting that the source of the upright superiority effect was not the external but the internal parts of the face. Methods Results and discussion 4.0 TGT/DST Up/Inv Inv/Up Response Times (s) 3.5 *** 3.0 2.5 2.0 ** 1.5 1.0 0.5 Original External Parts Internal Parts Fig. 5 Mean response times for each condition in Experiment 3. **P<0.01; ***P<0.001 123 Chloe correctly performed in 94.8, 63.5, 36.5, 68.8, and 31.3% of trials on average across target orientations for the original, schematic, schematic scrambled, eyes + eyebrows, and eyes only stimuli, respectively. She showed higher accuracy for the schematic face than for the schematic scrambled stimuli [t(7)=4.202, P<0.01] and higher accuracy for the eyes + eyebrows than for the eyes only stimuli [t(7)=7.180, P<0.001]. Figure 6 shows the mean response times averaged across the correct and incorrect trials for each condition. As in Experiment 1, the upright superiority effect was also observed for the schematic face [t(7)=3.974, P<0.01) but not for the scrambled patterns. These results further support the face specificity of the upright superiority effect in visual search. Chloe again showed a significant upright superiority effect for eyes + eyebrows [t(7)=2.806, P<0.05] but not for eyes only. These results were consistent with those of Experiment 2, suggesting that the upright superiority effect is due to the configural processing of facial stimuli. Primates (2007) 48:1–12 9 Fig. 6 Mean response times for each condition in Experiment 4. *P<0.05; **P<0.01 4.0 Up/Inv Inv/Up 3.5 Response Times (s) ** TGT/DST * 3.0 2.5 2.0 1.5 ** 1.0 0.5 Original General discussion The present experiments clearly replicated and extended the previous study in tests with the same subject (Tomonaga 1999a). Chloe exhibited the upright superiority effect for orientation during visual search of caricatures, dogs, and schematic faces, but not hand shapes or chairs, indicating the face specificity of this effect. Furthermore, if the normal spatial configuration of the facial image could not be used, as in presentations of a single facial feature or due to rearrangements of the spatial configuration of features, the upright superiority effect was no longer observed. These results provide strong support for the hypotheses that chimpanzees process facial images using configural properties, as proven by the inversion effect under the rotational matching paradigm (Parr et al. 1998; Tomonaga 1999b), and that the upright superiority effect is caused by the configural processing of faces by the chimpanzee (Tomonaga 1999a, 2001a). There are no other reports of the same kind of effect in primates, including humans, under identical experimental procedures. Thus, we should examine this effect both in humans and in other primates such as macaques. Humans should show similar results if this effect is closely related to configural processing of faces. Replicating the present results with macaque monkeys would also be of interest because there have been some confounding results concerning the facial inversion effect in macaques. Eyes + Eyebrows Eyes only Schematic Schematic Scrambled Although the same effect has not been reported, related phenomena have been observed in human studies. One of these phenomena is called the ‘‘face detection effect,’’ as reported by Purcell and Stewart (1986, 1988). They found that detection of an upright face in a brief time-frame is easier than detection of inverted and scrambled faces. Recently Lewis and Edmonds (2005) found a similar effect using visual search tasks. They found that detection of an upright face in scrambled scenes was easier than detection of an inverted face. The difference between the present experiments and those of Lewis and Edmonds is the demand for discrimination; the present experiments required the subject to detect a target face among distractor faces, while Lewis and Edmonds had the subject detect a target face among non-face distractors (except in their second experiment). A visual search for facial stimuli requiring discrimination among faces is more difficult than discrimination between face and non-face stimuli or among non-face stimuli (e.g., Hershler and Hochstein 2005; Lewis and Edmonds 2005; Nothdurft 1993). However, even though the task demands were different, the present results are consistent with these human studies, and indicate better detection of upright faces than inverted faces. An upright face is also easier to find because it is highly familiar to the subject compared with disoriented faces (Tomonaga 2001a). However, recent advances in the relationship between search asymmetries and familiarity have provided opposite predictions. 123 10 When using familiar and unfamiliar items, human observers have shown more efficient detection of unfamiliar targets among familiar distractors than vice versa (Malinowski and Hübner 2001; Shen and Reingold 2001; Wang et al. 1994; Wolfe 2001). Interestingly, Wolfe (2001) reported that human observers found the inverted silhouette of an elephant among upright elephants more efficiently than vice versa, which is inconsistent with the present results. However, the relationship between familiarity and search asymmetries has shown various results when using facial stimuli. Montoute and Tiberghien (2001) and Masuda (2003) reported that famous faces can be more efficiently detected than unknown faces. Tong and Nakayama (1999) found that subjects searched for their own face more efficiently than other faces. In contrast, Levin (1996) reported that Caucasian subjects searched for the faces of Africans among Caucasian faces more efficiently than vice versa, even though Caucasian faces are more familiar to Caucasian people. As Diamond and Carey (1986) have suggested, increased familiarity via prolonged exposure enhances configural processing of stimuli. Although we should consider the different levels of familiarity carefully, familiarity enhances configural processing of upright faces and their efficient detection among disoriented faces or non-face stimuli, as in the visual search for famous faces. In conclusion, the present results provide further support for the configural processing of faces in chimpanzees (Parr et al. 1998; Tomonaga 1999a, b, 2001a). In humans, the development of face recognition has been extensively investigated from very early infancy to childhood (Carey and Diamond 1977, 1994; Johnson and Morton 1991). Human children show a shift from featural (or analytical) to configural processing of faces during development. At present, comparative developmental studies on face recognition in chimpanzees are relatively few (Kuwahata 2004; Kuwahata et al. 2003; Myowa-Yamakoshi et al. 2005; Tomonaga et al. 2004). Because of the relatively low task demand in comparison to the matching paradigm (Tomonaga 1995), visual search for orientation of faces is an effective means of investigating developmental changes in face processing in chimpanzees. Acknowledgments This study and preparation of the manuscript were financially supported by Grants-in-Aid for Scientific Research from the Japanese Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan Society for the Promotion of Science (JSPS; Grant # 04710053, 05206113, 05710050, 07102010, 09207105, 10CE2005, 12002009, 13610086, 16002001, 16300084) and MEXT Grants-in-Aid for the 21st Century COE Program (A14 and D10). The author wishes to thank Drs. Tetsuro Matsuzawa and Masayuki Tanaka for their valuable comments on this study. Thanks are also due to Sumi- 123 Primates (2007) 48:1–12 haru Nagumo for his technical support and to Kiyonori Kumazaki, Norihiko Maeda, and the staff of the Center for Human Evolution Modeling Research of the Primate Research Institute (PRI) for their care of the chimpanzees. References Ballaz C, Boutsen L, Peyrin C, Humphreys GW, Marendaz C (2005) Visual search for object orientation can be modulated by canonical orientation. J Exp Psychol Hum Percept Perform 31:20–39 Blough DS (1979) Effects of the number and form of stimuli on visual search in the pigeon. J Exp Psychol Anim Behav Process 5:211–223 Bruce C (1982) Face recognition by monkeys: absence of an inversion effect. Neurophsychologia 20:515–521 Campbell R, Walker J, Baron-Cohen S (1995) The development of differential use of inner and outer face features in familiar face identification. J Exp Child Psychol 59:196–210 Carey S, Diamond R (1977) From piecemeal to configurational representation of faces. Science 195:312–314 Carey S, Diamond R (1994) Are faces perceived as configurations more by adults than by children? Visual Cogn 1:253– 274 Diamond R, Carey S (1986) Why faces are and are not special: an effect of expertise. J Exp Psychol 115:107–117 Dittrich WH (1990) Representation of faces in longtailed macaques (Macaca fascicularis). Ethology 85:265–278 Ellis HD, Shepherd JW, Davies GM (1979) Identification of familiar and unfamiliar faces from internal and external features: some implications for theories of face recognition. Perception 8:431–439 Fagot J, Tomonaga M (1999) Global-local processing in humans (Homo sapiens) and chimpanzees (Pan troglodytes): use of a visual search task with compound stimuli. J Comp Psychol 113:3–12 Farah MJ, Tanaka JW, Drain HM (1995) What causes the face inversion effect? J Exp Psychol Hum Percept Perform 21:628–634 Gothard KM, Erickson CA, Amaral DG (2004) How do rhesus monkeys (Macaca mulatta) scan faces in a visual paired comparison task? Anim Cogn 7:25–36 von Grünau M, Anston C (1995) The detection of gaze direction: a stare-in-the-crowd effect. Perception 24:1297–1313 Hamilton CR, Vermeire BA (1983) Discrimination of monkey faces by split-brain monkeys. Behav Brain Res 9:263–275 Hamilton CR, Vermeire BA (1988) Complementary hemispheric specialization in monkeys. Science 242:1691–1694 Hansen C, Hansen R (1988) Finding the face in the crowd: an anger superiority effect. J Pers Soc Psychol 54:917–924 Hershler O, Hochstein S (2005) At first sight: a high-level pop out effect for faces. Vis Res 45:1707–1724 Hunnius S, Geuze RH (2004) Developmental changes in visual scanning of dynamic faces and abstract stimuli in infants: a longitudinal study. Infancy 6:231–255 Jitsumori M, Matsuzawa T (1991) Picture perception in monkeys and pigeons: transfer of rightside-up versus upside-down discrimination of photographic objects across conceptual categories. Primates 32:473–482 Johnson MH, Morton J (1991) Biology and cognitive development the case of face recognition. Blackwell, Cambridge Kanazawa S (1996) Recognition of facial expressions in a Japanese monkey (Macaca fuscata) and humans (Homo sapiens). Primates 37:25–38 Primates (2007) 48:1–12 Keating CF, Keating EG (1993) Monkeys and mug shots: cues used by rhesus monkeys (Macaca mulatta) to recognize a human face. J Comp Psychol 107:131–139 Kuwahata H (2004) Comparative–cognitive studies on face recognition in primates. PhD dissertation, Kyoto University, Japan Kuwahata H, Fujita K, Ishikawa S, Myowa-Yamakoshi M, Tomonaga M, Tanaka M, Matsuzawa T (2003) Recognition of facial patterns in infant chimpanzees (in Japanese). In: Tomonaga M, Tanaka M, Matsuzawa (eds) Cognitive and behavioral development in chimpanzees: a comparative approach. Kyoto University Press, Kyoto pp 89–93 Levin DT (1996) Classifying faces by race: the structure of face categories. J Exp Psychol Learn Mem Cogn 22:1364–1382 Lewis MB, Edmonds AJ (2005) Searching for faces in scrambled scenes. Visual Cogn 12:1309–1336 Luria AR, Pravdina-Vinarskaya EN, Yarbus AL (1964) Eye movement mechanisms in normal and pathological vision. Sov Psychol Psychiatry 2:28–39 Malinowski P, Hübner R (2001) The effect of familiarity on visual-search performance: evidence for learned basic features. Percept Psychophys 63:458–463 Martin-Malivel J, Fagot J (2001) Perception of pictorial human faces by baboons: effects of stimulus orientation on discrimination performance. Anim Learn Behav 29:10–20 Masuda S (2003) Effect of familiarity on visual search for faces (in Japanese). In: Proceedings of the 67th annual convention of the Japanese Psychological Association. Japanese Psychological Association, Tokyo, p 599 Matsuzawa T (1990) Form perception and visual acuity in a chimpanzee. Folia Primatol 55:24–32 Montoute T, Tiberghien G (2001) Unconscious familiarity and local context effects on low-level face processing: a reconstruction hypothesis. Conscious Cogn 10:503–523 Myowa-Yamakoshi M, Tomonaga M, Tanaka M, Matsuzawa T (2003) Preference for human direct gaze in infant chimpanzees (Pan troglodytes). Cognition 89:B53–64 Myowa-Yamakoshi M, Yamaguchi M, Tomonaga M, Tanaka M, Matsuzawa T (2005) Development of face recognition in infant chimpanzees (Pan troglodytes). Cogn Dev 20:49– 63 Nahm FKD, Perrett A, Amaral DG, Albright TD (1998) How do monkeys look at faces? J Cogn Neurosci 9:611–623 Nothdurft HC (1993) Faces and facial expressions do not pop out. Perception 22:1287–1298 Ochiai T, Matsuzawa T (1998) Planting trees in an outdoor compound of chimpanzees for an enriched environment. In: Hare VJ, Worley E (eds) Proceedings of the third international conference on environmental enrichment. The Shape of Enrichment, San Diego pp 355–364 Öhman A, Lundqvist D, Esteves F (2001) The face in the crowd revisited: a threat advantage with schematic stimuli. J Pers Soc Psychol 80:381–396 Overman WH, Doty RW (1982) Hemispheric specialization displayed by man but not macaques for analysis of faces. Neuropsychologia 20:113–128 Parr LA, Dove T, Hopkins WD (1998) Why faces may be special: evidence of the inversion effect in chimpanzees. J Cogn Neurosci 10:615–622 Parr LA, Winslow IT, Hopkins WD (1999) Is the inversion effect in rhesus monkey face-specific? Anim Cogn 2:123–129 Perrett DI, Mistlin AJ (1990) Perception of facial characteristics by monkeys. In: Stebbins WC, Berkley MA (eds) Comparative perception vol 2, Wiley, New York, pp 178– 215 11 Phelps MT, Roberts WA (1994) Memory for pictures of upright and inverted primate faces in humans (Homo sapiens), squirrel monkeys (Saimiri sciureus), and pigeons (Columbia livia). J Comp Psychol 108:114–125 Purcell DG, Stewart AL (1986) The face-detection effect. Bull Psychon Soc 24:118–120 Purcell DG, Stewart AL (1988) The face-detection effect: configuration enhances detection. Percept Psychophys 43:355– 366 Rosenfeld SA, Van Hoesen GW (1979) Face recognition in the rhesus monkey. Neuropsychologia 175:503–509 Sato N, Nakamura K (1999) Detection of directed gaze in rhesus monkeys (Macaca mulatta). J Comp Psychol 115:115–121 Senju A, Hasegawa T, Tojo Y (2005) Does perceived direct gaze boost detection in adults and children with and without autism? The stare-in-the-crowd effect revisited. Visual Cogn 12:1474–1496 Shen J, Reingold EM (2001) Visual search asymmetry: the influence of stimulus familiarity and low-level features. Percept Psychophys 63:464–475 Suzuki S, Cavanagh P (1995) Facial organization blocks access to low-level features: an object inferiority effect. J Exp Psychol Hum Percept Perform 21:901–913 Tanaka JW, Farah MJ (1991) Second-order relational properties and the inversion effect: testing a theory of face perception. Percept Psychophys 50:367–372 Tomonaga M (1993a) Use of multiple-alternative matching-tosample in the study of visual search in a chimpanzee (Pan troglodytes). J Comp Psychol 107:75–83 Tomonaga M (1993b) Tests for control by exclusion and negative stimulus relations of arbitrary matching to sample in a ‘‘symmetry-emergent’’ chimpanzee. J Exp Anal Behav 59:215–229 Tomonaga M (1994) How laboratory-raised Japanese monkeys (Macaca fuscata) perceive rotated photographs of monkeys: evidence for an inversion effect in face perception. Primates 35:155–165 Tomonaga M (1995) Visual search by chimpanzees (Pan): assessment of controlling relations. J Exp Anal Behav 63:175–186 Tomonaga M (1997) Precuing the target location in visual searching by a chimpanzee (Pan troglodytes): effects of precue validity. Jpn Psychol Res 39:200–211 Tomonaga M (1999a) Visual search for orientation of faces by a chimpanzee (Pan troglodytes) (in Japanese with English abstract). Primate Res 15: 215–229 Tomonaga M (1999b) Inversion effect in perception of human faces in a chimpanzee (Pan troglodytes). Primates 40:417– 438 Tomonaga M (1999c) Establishing functional classes in a chimpanzee (Pan troglodytes) with a 2-item sequential-responding procedure. J Exp Anal Behav 72:57–79 Tomonaga M (2001a) Investigating visual perception and cognition in chimpanzees (Pan troglodytes) through visual search and related tasks: from basic to complex processes. In: Matsuzawa T (eds) Primate origin of human cognition and behavior. Springer, Tokyo, pp 55–86 Tomonaga M (2001b) Visual search for biological motion patterns in chimpanzees (Pan troglodytes). Psychologia 44:46– 59 Tomonaga M, Matsuzawa T (1992) Perception of complex geometric figures in chimpanzees (Pan troglodytes) and humans (Homo sapiens): analyses of visual similarity on the basis of choice reaction time. J Comp Psychol 106:43– 52 123 12 Tomonaga M, Itakura S, Matsuzawa T (1993) Superiority of conspecific faces and reduced inversion effect in face perception by a chimpanzee. Folia Primatol 61:110–114 Tomonaga M, Myowa-Yamakoshi M, Mizuno Y, Yamaguchi MK, Kosugi D, Bard KA, Tanaka M, Matsuzawa T (2004) Development of social cognition in infant chimpanzees (Pan troglodytes): face recognition, smiling, gaze and the lack of triadic interactions. Jpn Psychol Res 46:227–235 Tong F, Nakayama K (1999) Robust representations for faces: evidence from visual search. J Exp Psychol Hum Percept Perform 25:1016–1035 Treisman A, Gelade G (1980) A feature-integration theory of attention. Cogn Psychol 12:97–136 Valentine T (1988) Upside-down faces: a review of the effects of inversion upon face recognition. Br J Psychol 79:471–491 123 Primates (2007) 48:1–12 Wang Q, Cavanagh P, Green M (1994) Familiarity and pop-out in visual search. Percept Psychophys 56:495–500 Williams M, Moss SA, Bradshaw JL, Mattingley JB (2005) Look at me, I’m smiling: visual search for threatening and nonthreatening facial expressions. Visual Cogn 12:29–50 Wolfe JM (2001) Asymmetries in visual search: an introduction. Percept Psychophys 63:381–389 Wright AA, Roberts WA (1996) Monkey and human face perception: inversion effects for human but not for monkey faces or scenes. J Cogn Neurosci 8:278–290 Yamaguchi MK, Hirukawa T, Kanazawa S (1995) Judgment of gender through facial parts. Perception 24:563–575 Yarbus AL (1967) Eye movements and vision. Plenum, New York Yin RK (1969) Looking at upside-down faces. J Exp Psychol 81:141–145
© Copyright 2026 Paperzz