Referential Communication by Chimpanzees (Pan troglodytes) David A. Leavens University of Georgia, Yerkes National Primate Research Center, and University of Sussex William D. Hopkins Berry College and Yerkes National Primate Research Center Roger K. Thomas University of Georgia Two experiments were conducted to assess the referential function of chimpanzee (Pan troglodytes) gestures to obtain food. The chimpanzees received 1 trial per condition. In Experiment 1 (N = 101), in full view of the chimpanzee, a banana was placed on top of 1 of 2 inverted buckets or was hidden underneath 1 of the buckets. In Experiment 2 (N = 35), 4 conditions were presented in constant order: (a) no food, no observer; (b) no food, observer present; (c) food present, no observer; and (d) food present, observer present. Gestures and visual orienting were used socially and referentially. The capacity for nonverbal reference may predate the Hominidae–Pongidae split, and the development of nonverbal reference may be independent of human species-specific adaptations for speech. Referential signaling has been historically viewed as being uniquely human (Werner & Kaplan, 1963). Referential signaling, as defined here, means the capacity to direct the attention of an observer to a distal object or entity. In linguistic reference, a symbol represents a concept, and through shared lexicons people can direct the attention of their conversational partners to these conceptual entities. Nonverbal reference, as the term is used in comparative research (e.g., Call & Tomasello, 1994; Leavens & Hopkins, 1998) and human developmental research (e.g., Adam- son, 1996), is the capacity of a signaler to direct the attention of an observer through deictic and iconic manual gestures, the signaler’s visual orienting behavior, and attention-directing tactile or audi- tory signals. Pointing is the paradigmatic act of nonverbal reference (Bates, O’Connell, & Shore, 1987). It has been claimed (a) that apes do not point (Corballis, 1991); (b) that they do not point with the index finger unless explicitly trained to point (Povinelli & Davis, 1994); and (c) that compared with humans, when apes point, they do so with little understanding that pointing gestures influence the “internal mental states” of the observer (Povinelli, Bering, & Giambrone, 2000, pp. 527–536). It has also been said that apes do not point while interacting with conspecifics (e.g., Povinelli et al., 2000). However, pointing among apes has been reported in cap- tivity by de Waal (1982) and Savage-Rumbaugh (1986), among others, and in the wild by Inoue-Nakamura and Matsuzawa (1997) and Vea ` and Sabater-Pi (1998). Nevertheless, pointing behavior does not seem to be as common an occurrence among wild apes as it is among humans: Long-term field studies of chimpanzees in their natural habitats have not identified pointing as a frequent constituent of their communicative repertoires. However, in captive settings, pointing is frequently exhibited by apes interacting with humans (e.g., Call & Tomasello, 1994; Krause & Fouts, 1997; Leavens & Hopkins, 1998; Leavens, Hop- kins, & Bard, 1996; C. R. Menzel, 1999; Miles, 1990; Savage- Rumbaugh, 1986). Pointing is not limited to apes that have been language trained or raised as part of a human family unit (cf. Tomasello & Call, 1997), and most reports of pointing by apes in captivity describe it as emerging “spontaneously,” in the absence of explicit training (Call & Tomasello, 1994). Apes in captivity usually point with their whole hands, rather than with their index fingers (de Waal, 1982; Leavens & Hopkins, 1999), except for language-trained apes, which frequently do point with the index finger (Call & Tomasello, 1994; Krause & Fouts, 1997); although less frequently, index finger pointing has been seen in apes that have not been language trained (Leavens & Hopkins, 1998, 1999; Leavens et al., 1996). Perhaps because of this, it has been argued that apes do not point to distal objects, although they may exhibit “reaches” for unreachable food (e.g., David A. Leavens, Department of Psychology, University of Georgia; Division of Psychobiology, Yerkes National Primate Research Center, Atlanta, Georgia; and Psychology Department, School of Life Sciences, University of Sussex, East Sussex, England. William D. Hopkins, Depart- ment of Psychology, Berry College, and Division of Psychobiology, Yerkes National Primate Research Center. Roger K. Thomas, Department of Psychology, University of Georgia. This research was supported by National Institutes of Health Grants RR-00165 and NS-29574. Thanks to Dawn Pilcher for invaluable assis- tance with data collection. Thanks to J. Allen, Irwin Bernstein, Dorothy Fragaszy, Janet Frick, and B. E. Mulligan at the University of Georgia for helpful discussion of these data and for suggesting Experiment 2. Thanks to Tony Stubbens, University of Sussex, for technical assistance. Logistical support was provided by Frans de Waal and Darren Long at The Living Links Center for the Advanced Study of Ape and Human Evolution, Yerkes National Primate Research Center. Special thanks to Kim A. Bard, University of Portsmouth, for helpful commentary and critique. Thanks to Beena Khurana, University of Sussex, for helpful discussion of these findings. Correspondence concerning this article should be addressed to David A. Leavens, Psychology Department, School of Life Sciences, University of Sussex, Falmer, East Sussex BN1 9QH, England. E-mail: davidl@ sussex.ac.uk Journal of Comparative Psychology Copyright 2004 by the American Psychological Association, Inc. 2004, Vol. 118, No. 1, 48 –57 0735-7036/04/$12.00 DOI: 10.1037/0735-7036.118.1.48 48