We have been thinking about joint attention—what it is, where it comes from, and its role in human development—for almost 2 decades now. Most of the research has focused on two sets of phenomena. Ontogenetically first is the tendency of human infants, in the months immediately preceding their first birthday, to follow or to direct the visual attention of adults to outside entities (e.g., Bates, 1976; Butterworth … Cochran, 1980; Corkum … Moore, this volume; Scaife … Bruner, 1975). Ontogenetically second is the tendency of somewhat older children, in their second year of life, to participate with adults in more extended bouts of joint visual attention as they begin to acquire their first linguistic conventions (e.g., Bakeman … Adamson, 1984; Baldwin, this volume; Dunham … Dunham, this volume; Ninio … Bruner, 1978; Tomasello … Todd, 1983). More recently, however, researchers with a more strictly cognitive bent have begun to look at children's joint attentional skills as “precursors” to their representational theories of mind (e.g., Baron-Cohen, 1991, 1993; Meltzoff … Gopnik, 1993; Wellman, 1993). This new perspective has opened up a whole new set of questions about the cognitive and social-cognitive bases of joint attention.