Child Development, May/June 2001, Volume 72, Number 3, Pages 718–735 Evidence for Referential Understanding in the Emotions Domain at Twelve and Eighteen Months Louis J. Moses, Dare A. Baldwin, Julie G. Rosicky, and Glynnis Tidball Infants as young as 12 months readily modulate their behavior toward novel, ambiguous objects based on emotional responses that others display. Such social-referencing skill offers powerful benefits to infants’ knowledge acquisition, but the magnitude of these benefits depends on whether they appreciate the referential quality of others’ emotional messages, and are skilled at using cues to reference (e.g., gaze direction, body pos- ture) to guide their interpretation of such messages. Two studies demonstrated referential understanding in 12- and 18-month-olds’ responses to another’s emotional outburst. Infants relied on the presence versus ab- sence of referential cues to determine whether an emotional message should be linked with a salient, novel ob- ject in the first study ( N = 48), and they actively consulted referential cues to disambiguate the intended target of an affective display in the second study ( N = 32). These findings provide the first experimental evidence of such sophisticated referential abilities in 12-month-olds, as well as the first evidence that infant social referenc- ing at any age actually trades on referential understanding. INTRODUCTION Human infants emerge into a social-filled world: a di- verse and ever-changing flow of social and emotional information emanates from others’ faces and voices, from the quality of their touch, the posture and move- ments of their bodies, and even from the way they smell. This rich social and emotional environment has tremendous potential to speed infants’ learning about the world, if only they can take advantage of it. If in- fants possess some basic interpretive abilities, such as the ability to decode socioemotional messages and re- late them to the appropriate objects and events in the world, then they have access to a vastly expanded body of information about both the social and nonso- cial world. Infants wielding these basic interpretive abilities could analyze the way in which objects and actions impact others’ emotions, and in turn could observe how these emotions give rise to subsequent action. This information could then be used to guide infants’ own dealings with the world. The time sav- ings with regard to learning would be nearly inesti- mable. With these abilities, infants could successfully acquire knowledge about the world purely through observation of others’ responses. Without these abili- ties, social information might simply fly past infants as scattered and relatively uninterpreted change in the sensory array, and infants would only acquire in- formation through laborious, direct, personal experi- ence with each and every new object or person. The present research focused on whether knowledge ac- quisition in infants of 12 and 18 months indeed reaps the benefits of such interpretive abilities. Research over the past 20 years has provided abun- dant evidence that infants as young as 12 months of age are very skilled at decoding emotional informa- tion (see Baldwin & Moses, 1996, and Mumme, 1993, for reviews). They can discriminate among a variety of different facial expressions (e.g., Kuchuk, Vibbert, & Bornstein, 1986; Ludemann & Nelson, 1988; Soken & Pick, 1999), as well as different affect-laden intona- tion patterns (e.g., Fernald, 1993; Soken & Pick, 1999). In addition, they are guided by adults’ emotional sig- nals in the ways in which they approach new things, an ability termed “social referencing” (e.g., Campos & Stenberg, 1981; Feinman, 1982). For example, if a parent on the opposite side of a visual cliff displays fear, infants grow wary and do not cross, whereas if a parent smiles, they approach and readily cross (Sorce, Emde, Campos, & Klinnert, 1985). None of this early research clarified one of the most important issues from the point of view of knowledge acquisition, however: whether infants possess skills for appropriately linking adults’ emotional signals with the things to which adults are actually referring. Consider the visual cliff, for example: Recognizing that their parents are fearful will not stop infants from falling over a cliff if they fail to realize that the cliff— or perhaps, more accurately, their approach to the cliff—is what renders their parents fearful. In partic- ular, infants need to appreciate that (1) another’s so- cial message relates to a specific thing in the world, (2) it is the emoter who determines which thing is the ref- erent, and (3) the emoter’s referential cues can be used to discover that thing in a given case. In short, they must recognize that emotional outbursts are referen- © 2001 by the Society for Research in Child Development, Inc. All rights reserved. 0009-3920/2001/7203-0008