145th ASA Meeting, Nashville, TN

Infants in Cocktail Parties

Rochelle S. Newman - rnewman@hesp.umd.edu
Department of Hearing and Speech Sciences and
Program in Neurosciences and Cognitive Sciences
University of Maryland
College Park, MD 20742

Popular version of paper 3aSC2
Presented Wednesday Morning, April 30, 2003
145th ASA Meeting, Nashville, TN

Cars screeching across the TV screen; siblings bickering in another room; the blender whirrs in the kitchen; and somehow, in the midst of all the noise, an infant sifts the sound of her own name out of the tumult and responds. This environment, typical of modern households, could not be more different than the quiet laboratory settings in which most research on how infants learn language has taken place. The present research seeks to bridge that gap and explores how well young infants can understand speech spoken to them in the noisy environment of daily life. The study examined how long a 4.5-month-old child spent listening when a female voice said either the baby’s name or another infant’s name while other voices spoke in the background.

Names were chosen because prior research in quiet labs has shown that at 4.5 months, babies will listen longer to their own name than to another name (Mandel, Jusczyk & Pisoni, 1995). We examined if this would hold true in a noisy environment. If infants are not able to recognize speech in these settings, it would suggest that infants who spend a significant amount of time in noisy settings such as daycare centers, or in the presence of siblings, would be at a distinct disadvantage for learning language.

Testing Infants sat on their caregiver’s lap in a three-sided booth. A light on one of the side panels attracted the infants’attention; once the infant was looking in that direction, the woman’s voice began calling a name. On some trials, the name was that of the infant being tested; on other trials, it was the name of a different child. The woman continued to repeat the name until the baby turned his or her head away, and we measured the amount of time each infant spent listening to the different names. Results In the first version of the experiment, the woman’s voice was 10 decibels louder than the sounds of the other people talking; this is roughly equivalent to the noise level you might experience while having a one-on-one conversation at a reasonably quiet restaurant (one filled with other patrons, but in which people are speaking softly). Examples of these sounds are presented below. In this case, infants listened longer to their names than to the names of other children. Alexander      Michelle In the second version of the experiment, the woman’s voice was only 5 decibels louder than the sounds of the background talker. (Examples of these sounds are presented below.) Here, infants found the task much harder; they did not appear to be able to distinguish their own name from other names in this situation. Allison      Jeremiah These results suggest that infants are much more sensitive to noise than are adult listeners. Adults can understand speech when the voice they are listening to is 8 dB softer than the background noise (Hawkins & Stevens, 1950); yet infants were having difficulty when the target was still louder than the background. Thus, infants are likely to have difficulty listening in levels of noise that adults wouldn’t even think twice about. This is particularly worrisome since daycare centers can often have noise levels that even adult listeners find daunting.1 These results imply that infants in day care settings are likely unable to distinguish much of the speech spoken to them, and thus have fewer opportunities to learn their native language.

Main references Hawkins, J. E., Jr., & Stevens, S. S. (1950) The masking of pure tones and of speech by white noise. Journal of the Acoustical Society of America, 22(1) 6-13. Mandel, D. R., Jusczyk, P. W., Pisoni, D. B. (1995). Infants’ recognition of the sound patterns of their own names. Psychological Science, 6(5), 314-317. Picard, M. & Bradley, J. S. (2001) Revisiting speech interference in classrooms. Audiology, 40, 221-244. Endnotes 1 Picard and Bradley (2001) report that signal to noise ratios in schools average somewhere between the two signal-to-noise ratios used in our study, and are sometimes much worse. Although similar measurements have not been done for day care centers, day care settings tend to be noisier than schools, suggesting that the signal-to-noise ratios would also be worse. This material is based upon work supported by the National Science Foundation under Grant No. 0196498 and by grant number HD37822 from the National Institute of Child Health and Human Development. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author and do not necessarily reflect the official views of the National Science Foundation or the National Institutes of Health.