Poster #53 - Pitch perfect: Examining the role of pitch changes during word learning in speech and song

• In Event: Poster Session 13

Abstract

The role of attention on language learning is a frequent area of study, as attending to the appropriate stimulus is crucial for learning (McDowd, 2007). For infants and young children, songs have been found to sustain attention better than infant directed speech (IDS) (Corbeil et al., 2016; Nakata and Trehub, 2004; Trainor, 1996). Further, it has been proposed that songs are advantageous for language learning in young children (Lebedeva & Kuhl, 2010; Thiessen & Saffran, 2009). We will present data from two studies examining the influence of song during word learning.

In the first study, 47-50-month-old typically developing children completed a word learning task where they were taught new words (e.g., doop and fim) that corresponded to novel objects. Words were trained either in a spoken sentence produced in IDS (e.g., Look it’s a doop! Wow, it’s a doop! Do you see it? A doop!), or in a song where the melody was set to that of “Old MacDonald had a Farm.” Immediately after training, children were tested on their ability to recognize each item using the Preferential Looking Paradigm (Golinkoff et al., 1987). Videos of the sessions were recorded and coded offline. Analyses focus on comparing the proportion of time spent looking at the target object to a measure of chance performance, and the proportion of time spent looking at the correct object for words trained in the Spoken compared to the Song condition. In this study we found that Spoken condition led to better learning than the song condition with an 11% increase in learning in the spoken condition compared to the song condition. This was the case, even though song trials resulted in greater attention (i.e., longer looking to the objects) during training. Thus, it appeared that song facilitated attention but not learning.

The second study aimed to examine whether the results seen in the first study were due to differences in pitch variation. In the original study, pitch changes were seen before the presentation of the target word in the spoken (IDS) condition. In the song condition, while pitch changed as the melody progressed, a change in pitch was not present right before the target word was presented. Pitch variation has been shown to facilitate word learning (Seidl, 2007; Seidl & Cristiá, 2008). Therefore, we aimed to examine if differences in novel word learning existed between speech and song conditions when pitch variation was controlled for. To date, data from 5 children (3 male) have been coded and examined. When controlling for pitch variation, it appears that proportion of looking in both Spoken (63%) and Song (61%) conditions are comparable. Data from additional children will be available for the conference. This second study builds on previous work and suggests that perhaps the differences seen between learning in speech and song may be due to strategic pitch variations present in IDS that were not found in the song. A direct comparison across both the first and second study will be discussed at the conference

Authors

• Mackensie Blair, University of Delaware
  Presenting Author

• Lindsey Hawtof

• Giovanna Morini, University of Delaware