Self-recognition in mirrors is considered an indicator of self-awareness, a capacity once assumed to be present only in human beings. Gordon Gallup's initial studies using the face-marking test, and the work they stimulated, have clearly demonstrated mirror self-recognition (MSR) in only three species: humans, chimpanzees, and orangutans (Anderson, 1983; Gallup, 1970; Platt & Thompson, 1985; Suarez & Gallup, 1981). Nevertheless field studies and laboratory tests, including simple discrimination learning, learning set, discrimination-reversal training, and oddity concept formation, have demonstrated that the cognitive abilities of all three genera of great apes are closely comparable. Gorillas and orangutans were ranked slightly above chimpanzees with respect to intelligence as measured by the transfer index, a refined measurement of learning-set ability in which species and individual differences in motivation and perceptuomotor skills are controlled (Rumbaugh & Gill, 1973). Piagetian studies also indicate that gorillas, chimpanzees, and orangutans undergo similar cognitive development (Chevalier-Skolnikoff, 1977, 1983; Mathieu & Bergeron, 1983; Redshaw, 1978). Nevertheless, the apparent inability of six gorilla subjects to recognize their mirrored images has led researchers to conclude that gorillas are the only great apes to lack the capacity for self-awareness (Ledbetter & Basen, 1982; Suarez & Gallup, 1981). Our current study with the female lowland gorilla Koko, which considers linguistic evidence as well as that provided by self-recognition tests, challenges this assertion.

Koko became the subject of an ongoing language study in July 1972, when she was 1 year old (Patterson, 1978b). She was taught sign language and continuously exposed to spoken English.