Infant Detection Merrill-Palmer Quarterly, Vol. 41, No.1 1984; Jacobson et Symmes, 1990). Several explan behavior. Most of infant emits some ( such exaggerated behavior. Indeed, poiting this notion The question r entially to AI spee (such as its expan( innately reinforcin this were so, one I highly effective in they would be eff( It is on this secon Fernald and Kuhi stimuli were presf quency of the AI 5 have found such In the first 0 month-olds' preff parameters of int< nald and Simon (' between AI and of preferring an a monotonic pu salience of variol and Aslin (1990) olds; however, il found that these the fundamental were low-pass f (1993) also foun, elicitation of inf the importance have been repol Although th, ties of AI stimuli removing or cor results in a redu AA stimuli. Thi Infants' Detection of Analogs of !'Motherese" in Noise John Colombo, Janet E. Frick, and Jennifer S. Ryther University of Kansas Jeffrey T. Coldren Youngstown StateUniversity D. Wayne Mitchell SouthwestMissouri State University Adult-to-infant (AI) speech has been found to be especially salient to young infants, but the mechanism underlying this salience is not well underStood. The present study is a test of the possibility that stimuli with the properties of AI speech are more detectable in a noisy ambient environment. To test this hypothesis, 4-month-olds were habituated to white noise; following habituation, one of three different signal stimuli were added to the white noise. One signal stimulus was a monotonic pure tone, another was a frequency-modulated sweep corresponding to the intonation parameters of adult-to-adult (AA) speech, and the third was a frequency modulated sweep corresponding to the intonation parameters of AI speech. Infants showed evidence of detecting only the signal modeled after AI speech. The pattern of adult speech to infants and children is markedly differ- ent from normal conversational speech between two adults. In compari- son to adult-to-adult speech (AA), adult-to-infant speech (AI, sometimes called "motherese"; see Cooper, 1993; Fernald, 1985; Fernald & Kuhl, 1987) is characterized by a slower rate of utterances (Stern, Speiker, & MacKain, 1983; see also Masataka, 1992), and higher overall pitch (Jacobson, Boersma, Fields, & Olson, 1983). Perhaps the most prominent feature of AI speech, however, is the exaggeration of vocal intonation, relative to that of AA speech (Fernald & Mazzie, 1991; Fernald & Simon, This work was supported by grants MH42346, HD18290, and HD29960. Michelle Knoll and Laura Maag assisted in subject recruitment, and Carrie Winterowd and Shannon Rash assisted in data collection. We thank the University of Kansas Regents Center staff for their cooperation, and are most grateful for the participation of the families of Johnson County, Kansas. Correspondence may be sent to John Colombo, Department of Human Development, University of Kansas, Lawrence, KS 66045-2133 (email: colombo@kuhub.cc.ukans.edu). Merrill-Palmer Quarterly, January 1995, Vol. 41, No.1, pp. 104-113. Copyright @ 1995 by Wayne State University Press, Detroit, MI 48201 104