American Journal of Primatology 2:307-310 (1982) BRIEF REPORT Failure to Demonstrate Self-Recognition in Gorillas DAVID H. LEDBETTER Depurtment of Psychology, University of Texas a t Austin JEFFKY A. BASEN Department of Psychology, University of Houston Two zoo-reared gorillas were each given nearly 400 h of mirror exposure. Extensive mirror gazing and social behaviors were exhibited, the frequency of which decreased gradually over the study period. Neither animal demonstrated the transition from other-directed to self-directed behavior characteristic of both chimpanzees and orangutans, and no evidence of self-recognition was found using the Gallup marking paradigm. These negative findings, after extensive mirror exposure, suggest that the gorilla may be the only great ape which lacks the conceptual ability necessary for self-recognition. Key words: Gorilla, self-recognition, mirror image INTRODUCTION The capacity for self-recognition in great apes was first experimentally demonstrated by Gallup . Although naive chimpanzees socially interacted with their mirror images, animals given several days of mirror exposure stopped responding socially and began displaying self-directed behaviors (such as grooming areas of the body otherwise visually inaccessible). After 10 d of exposure to a mirror, the chimpanzees were anesthetized and marked with a red dye above one eyebrow and at the top of one ear. Upon awakening, the animals showed a startle response to the mirror image and began touching the mark, apparently trying to wipe it off. Since the chimpanzees were anesthetized when marked, this mark-directed behavior has been interpreted as indicating recognition of their own (altered)image reflected in the mirror. Using the Gallup paradigm, self-recognition has been demonstrated in the orangutan by Lethmate & Ducker . Repeated attempts to demonstrate self-recognition in lesser apes and monkeys, however, have failed [see Gallup et al, 19801. Thus, selfrecognition may depend on a conceptual ability present in man and the great apes but absent in lesser apes and monkeys. A t the time this research was initiated in 1978, the gorilla had not been systematically tested for self-recognition. One would expect to find behavior parallel to that of chimpanzees and orangutans because of the close relationship of the gorilla to these two species. Anecdotal reports by Francine Patterson describe “Koko,” a female gorilla Received October 26, 1981; accepted November 11, 1981. Address reprint requests to David IT. Ledbetter, The Kleberg Cylogenetics Laboratory, Department of Medicine, Baylor College of Medicine, Houston T X 77030. 027Ei-2565/82/0203-O307$01.500 1982 Alan R. Liss, Inc. 308 Ledbetter and Basen trained to use American Sign Language, putting make-up on in front of a mirror and photographing her mirror image, suggesting the ability for self-recognition . However, in a carefully controlled study, Suarez & Gallup have recently failed to demonstrate self-recognition in four gorillas after 80 h of mirror exposure, even though similarly treated chimpanzees and orangutans did show self-recognition. In the present study, two zoo-reared gorillas were each given nearly 400 h of mirror exposure. Data were collected to examine the expected transition from social to self-directed behavior, and an objective test of self-recognition using Gallup’s marking paradigm was also employed. METHODS AND RESULTS Two wild-caught lowland gorillas (Gorilla gorilla gorilla) reared and housed together at the Houston Zoological Gardens were the subjects of this study. The male (10 y old at the time of the study)and female (11y old) were both acquired by the zoo from a primate dealer in Holland, a t ages 20 and 11months, respectively. The zoo staff informed us that the animals had been exposed to small mirrors on several occasions in the past and it was their opinion that both animals “recognized themselves.” Mirror exposure was given in the morning before the animals were put on public display. At the outset of the study, mirror exposure was limited to the male, and lasted 30 min each day. The mirror (45 x 97 cm) was placed 0.6 m from the front of his cage (just out of reach).The mirror was angled such that the male could view only his own image, and not that of the female in the adjoining cage. The observer sat approximately 1.2 m from the front of the cage (and could be seen by the subject). The location in the cage and the behavior of the gorilla were recorded for 5- sec blocks at intervals of 30 sec during 15-min observation periods in the first phase of the study. During the first few days, the male displayed aggressive behavior toward the mirror image, often charging it. Although the frequency of aggressive and other social behaviors (e.g.,charging, head bobbing, vocalizing, etc.) decreased significantly over the first 4 wkof the study(withmeanvaluesof 17.1,8.9,10.0, and4.9 socialresponsesper 15 min), they never completely habituated. No self-directed behavior was observed. Selfdirected behavior was defined as that which provided the animal visual information about themselves that would be inaccessible were the mirror not available (e.g., grooming parts of the body visually inaccessible without the mirror, making faces at the mirror, or watching themselves eat in front of the mirror). Six weeks later, a second mirror was placed 0.6 m outside the cage of the female such that she could see only her own reflection. Mirror exposure time for each animal was approximately 2 h per d during this phase of the study. The female’s social responses to the mirror consisted primarily of mirror gazing rather than the aggressive behaviors of the male, and tended to be less frequent but of longer duration than the male’s. The frequency of her social responses habituated slowly over the course of the study (with mean values of 11.7,6.7, and 7.3 social responses per 15 min the first 3 wk).No signs of self-recognition were evident after 3 wk of mirror exposure. At this point, since neither animal showed any evidence of self-recognition, a final methodological change was made. The two mirrors were moved to the junction of the two cages and arranged such that each animal could see not only his or her own image, but also that of the other animal. Gallup et al have hypothesized that if an animal recognizes a familiar conspecific’s reflected image, this might facilitate recognition of the other “unfamiliar”individual. In addition to this change, mirror exposure time was again increased in that the mirrors were available each day from 7:OO a.m. to 4:OO p.m. whenever the animals were not on public display. The amount of exposure time varied from day to day, but averaged 3 h Self-Recognition in Gorillas 309 per d. Informal observations and notetaking 2 to 3 times per wk replaced the daily timesampling procedure for the duration of the study. Each of the two animals quickly appeared to recognize their partner’s mirror image, and began using the mirror to watch the movements of the other. Still neither animal displayed any signs of self-recognition. Even though no subjective evidence for self-recognition had been obtained, a marking test similar to that of Gallup was employed to objectively assess self-recognition. The test was conducted in conjunction with the annual physical examination and TB testing of the gorillas. The two animals were tested 10 wk apart. The male was tested after 390 h of mirror exposure, while the female had 395 h of exposure when she was tested. While each animal was anesthetized, white clown make-up was applied over one eyebrow and on the opposite ear. The male, who weighed 177 kg, was anesthetized with 175 mg Sernalyn. He appeared awake and normally active 2 h after the initial injection. The female, who weighed 120 kg, was given a total of 2600 mg ketamine over a 1- and li2-hr period. Normal activity level was again present 2 h after the initial injection. After normal activity levels returned, an additional 2-h recovery period was allowed before initiation of the marking experiment. At that time, a 15-min pretest was conducted to assess the frequency of random mark-directed behaviors. None occurred in either animal. For the 15-min test period, the mirror was placed 0.6 m outside the holding cage into which each animal had been placed after the physical examination. The mirror was uncovered, and mark-directed behavior assessed and recorded on videotape. No mark-directed responses by either animal were observed during the test period, although both animals spent time viewing the mirror (a total of 2 min, 32 s for the female, and 5 min, 22 s for the male). DISCUSSION The ability to recognize a mirror image as self has been clearly demonstrated in the chimpanzee and orangutan, while repeated attempts to demonstrate this ability in lesser apes and monkeys have failed [see Gallup et al, 19801. The gorilla, which is more closely related to the chimpanzee genetically [see Bruce & Ayala, 19791 and chromosomally [see Miller, 19771than is the orangutan, would be expected to display the ability for self-recognition. When this study was begun, the anecdotal reports of Koko’s ability for self-recognition led us to expect that an attempt to experimentally demonstrate self-recognition in gorillas would be successful. The initial failure with the male animal could have been due to unique characteristics of an individual animal, but the additional failure to observe self-directed behaviors from either the male or female after prolonged mirror exposure was more puzzling. It could be argued that the absence of self-recognitionin these two gorillas is due to the limited social experience of the animals. Gallup et al  have shown that chimpanzees raised in complete social isolation fail to display self-recognition; only after social rehabilitation did these animals show self-directed responses to their mirror image. In the present study, the two zoo-reared gorillas had had extensive social interaction with humans, but the only conspecificinteraction since an early age (11 and 20 months) had been with one another. The mirror image was the first encounter with an adult same-sex conspecific for each animal since their acquisition by the zoo. I t is possible this limited social atmosphere was not adequate for development of the self-recognition response. The possibility of our results being limited to zoo-reared gorillas is greatly lessened, however, by the results of the study by Suarez & Gallup involving four group-reared gorillas who showed no signs of self-recognition. 310 Ledbetter and Basen One might argue that perceptual or motivational factors rather than a conceptual deficit could account for the gorilla’s lack of self-directed behavior towards the mirror image. Although this possibility is not ruled out in our study, Suarez & Gallup included an additional control procedure in which marks were placed on each gorilla’s wrist as well as above one eyebrow. The subjects were aware of the marks on their wrists and repeatedly groomed the area, but were totally oblivious to similar marks over their brow. This demonstrates that the gorillas could see the marks, and showed interest in the marks on the wrist to the point of trying to rub them off. The failure of these two studies to demonstrate self-recognition in gorillas remains difficult to explain in light of the well-established close relationship of chimpanzees, gorillas, and orangutans. Although one cannot conclude from this that gorillas possess no concept self, these results raise the possibility that a less well developed selfconcept may be present in gorillas. Suarez & Gallup [19Sl] have proposed that a “subthreshold form of self-concept may exist which is not evident using current selfrecognition tests and suggested that it might be possible to design alternative testing procedures to specifically elicit self-directed behavior in monkeys. Perhaps alternative procedures will salvage the self-concept of the gorilla as well. 6.f ACKNOWLEDGMENTS The authors would like to express their sincere appreciation to the director, John Werler, and the many staff members of the Houston Zoological Gardens whose cooperation and interest made this study possible. A special thanks to Dr. Gordon G. Gallup, Jr., for his advice and encouragement throughout the project. REFERENCES Bruce, E.J.; Ayala, F.J. Phylogenetic relationships between man and the apes: Electrophoretic evidence. EVOLUTION 33(4): 1040-1056.1979. Gallup, G.G., J r . Chimpanzees: Self-recognition. SCIENCE 167:86-87, 1970. Gallup, G.G., Jr.; McClure, M.K.; Hill, S.D.; Bundy, R.A. Capacity for self-recognition in differentially reared chimpanzees. PSYCHOLOGICAL RECORD 21~69-74, 1971. Gallup, G.G., Jr.; Wallnau, L.B.; Suarez, S.D. Failure to find self-recognition in mother-infant rhesus monkey pairs. FOLIA PRIMATOLOGICA 33:210-219,1980, Lethmate, J.; Ducker, D. Untersuchungen zum sefbsterkennen im spiegel bei orang-utans und einigen anderen affernarten. ZEITSCHRIFT FUR TIERPSYCHOLGIE 33: 248-269.1973. Miller, D.A. Evolution of primate chromosomes. SCIENCE 198:1116-1124,1977. Patterson, F. Conversations with a gorilla. NATIONAL GEOGRAPHIC MAGAZINE 154: 438-465,1978. Suarez, S.D.; Gallup, G .G., J r . Self -recognition in chimpanzees and orangutans, but not gorillas. JOURNAL OF HUMAN EVOLUTION 10:175-188. 1981.