ABSTRACT
Some time ago, Bruner and his colleagues (1978, 1983; Scaife & Bruner, 1975) pointed out that joint attention plays a crucial role in children's acquisition of language. They argued that one of the primary tasks in language learning is for the child to link words and sentences with the correct objects, events, or properties in the world, and that mechanisms promoting joint attention between parent and child must certainly facilitate this mapping process. Clear empirical support can be found for this view. Parents readily provide attentional cues such as line-of-regard and gestures while speaking (e.g., Messer, 1983), and infants are able to make use of such cues from a relatively early age: They follow line-of-regard and pointing gestures to nearby objects by roughly 9–12 months (Butterworth, 1991; Butterworth & Jarrett, 1991; Corkum & Moore, 1992, this volume; Murphy & Messer, 1977; Scaife & Bruner, 1975). Further, when parents are more cooperative with respect to joint attention—they follow in on infants' own focus of attention when providing language—infants' vocabulary acquisition proceeds at a faster pace (e.g., Akhtar, Dunham, & Dunham, 1991; Dunham, Dunham, & Curwin, in press; Harris, Jones, Brookes, & Grant, 1986; Tomasello & Farrar, 1986). There is then every reason to agree with the essentials of this analysis: Mechanisms that serve joint attention enable parents and infants to achieve the social coordination necessary for language learning. But I would argue that something important is missing, or at least underemphasized, in this view. While it is clear that joint attention is critical to language learning, even more critical is what infants understand about the joint attention enterprise and especially, what they understand about its relevance to language. In other words, being aware that attention on some external thing is shared and understanding the significance of such intersubjectivity for communication would make all the difference for expediting language acquisition, and for that matter, any other kind of learning that is social in kind. What follows is an elaboration of this point, the result of which is twofold: a new understanding of some of the facts concerning early language learning, and a new approach to investigating the ontogeny of intersubjective understanding.
