From: LAURA SLOCUM Interesting case to support cross-linguistic universals in motherese. Not necessarily surprising though. Noteworthy implications: 1) "Melodies in early motherese may function as didactic caregiving messages and facilitate the infant's processing of contextual information (e.g., form=function)."; 2) "Highly consistent idiosyncratic shapes of caregiving melodies may represent salient cues for the identification of and familiarization with individual caretakers. General questions: Are there other related studies to consider? What happens in cultures in which motherese is not emphasized? What are other potential implications of this line of research? Kathleen Currie Hall Notes on: Papoušek, Mechtild, Papoušek, Hanuš, & Symmes, David. (1991). The meanings of melodies in motherese in tone and stress languages. Infant behavior and development, 14, 415-440. Summary: In this article, the authors explore how mothers' native languages (Mandarin Chinese vs. American English) interact with the intonational contours of the utterances they produce to their 2-month-old infants. They show that while there are some differences across individuals and languages, there are also several characteristics that point to a "universal" motherese. Details: 1. Motherese in general is associated with a high F0 and an expanded pitch range, characteristics that are believed to make motherese easier for infants to process and more likely to attract and keep infants' attention. 2. 10 AE and 10 MC mother-infant dyads participated; each interacted in the lab and was audio and video-recorded. 3. Semantic "contexts" of the interactions were encoded / transcribed without reference to the audio portions, but with reference to the actual words / sounds uttered by the participants (i.e., no prosodic info. was given). These were then compared to prosodic F0 contours. 4. Despite some variation by subject and by language, a lot of cross-linguistic similarity was found. Rising intonation tended to be used to elicit a child's attention and participation, and such elicitation was almost always done with a rising tone. Soothing an infant was done with a gentle falling tone, but falling tones of other types were also used to reward a child, discourage unfavourable behaviour, and encourage imitation, play, or attention. Rewarding a child was primarily done using a bell-shaped contour, though again this contour was used for other purposes as well. 5. Chinese tended to have more falling and bell-shaped contours than English. English tended to do less encouragement of imitation and soothing but more rewarding and evaluating of state. Questions: 1. I'd be interested in hearing more about the point mentioned at the very end of the article, about Chinese mothers "outwitting" the tone rules of their language. From: Kirk Baker Looked at prosody in mothers' verbalizations for pre-linguistic infants in Chinese and English for different types of interactions. Intonation contour slightly influenced by language; largely correspond across the two languages (and German). -What are underlying tone frequencies in Chinese lexicon? What are tone freqs for child directed lexical speech? From: Fangfang Li 1. not sure how exactly they separate out the F0 for lexical tone from F0 for the pragmatic fuction for Mandarin mothers. expecially the combination of the lexical tones are very likely to result in bell shape or U shape contour even in turn opening speech act. 2. What is the cross-linguistic difference in motherese melodies? the authors admit at the end of the paper that the American mothers' exaggerated pitch contour and pitch range may be owing to American parenting style as opposed something inherent in the language. if this is the case, then are there any differences in motherese melodies that are rooted in the specificity of language? From: HELENA RIHA The study aims to analyze the functions and meanings of melodic contours in parental speech as they naturally occur and to analyze the variabilities and universalities of these contours. Several studies are reviewed on p. 416 which show some functions of prosodic adjustments in melodic contours. One type of activity that has not been studied is games that mothers play with their babies and toddlers that involve rhythm, rhyme, and repetition. One example of this type is telling infants the story of the three little pigs, the itsy bitsy spider, etc. It would be useful to know what kind of melodic contours, rhythms, and physical gestures are involved in the repeated telling of these (non-spontaneous) stories, whose rhythmic sequences and prosody are memorized by the infant. The authors note that the coding of audiotapes was done by a musically experienced observer. Why did the authors choose someone with experience in music rather than language? They don't discuss the issue, but they must be assuming some kind of relationship between language and music. (p. 419) It is interesting to see the frequency of caregiving contexts in which the maternal utterances occurred (p. 423). Encouraging an infant turn, evaluating state, and encouraging imitation were the top three. This tells us that mothers work very actively to engage babies in interaction. The earliest stages of acquisition involve not just the sounds of language itself but also the discourse functions of speech communication -- i.e., figuring out that interaction through speech is meaningful; understanding how turn-taking works; and realizing that imitation is useful and that one will be rewarded for effective communication. The authors found that falling melodies are used for soothing, rising melodies encourage an infant turn, and turn-taking interventions are marked with rising versus falling/bell-shaped patterns (p. 435). This reminds of previous articles we read which emphasized that change away from a previous pattern is an important signal in the perception of language. Mothers intuitively use two maximally different patterns to express opposing meanings, and they move between them by using the opposite pattern to signal change. The authors? findings also show that mothers adjust the acoustic properties of their melodies to the infant?s behavioral state in a counterbalancing way (p. 436 ). Mothers accommodate to their babies to facilitate interaction. It seems that accommodation from the caregiver to the child are often overlooked in language acquisition except by those who study motherese. Normally, the child?s own ?working out? of the language is focused on, but based on this study and others of mot her-infant dyads, acquisition is an interactional process. From: ejkong@ling.ohio-state.edu - The article studied the characteristics of Motherese of Mandarine Chinese and American English in their interactions with babies, analyizing caretakers' melodic contours in terms of categorical types, acousitic specificis and their relationships wit interactional contexts. - The results show that the use of contour varies both categorically and gradiently depending on interacting contexts: caretakers tend to use types of countour in context specific ways, and also the same type of contours gradiently varies according to the contexts. Questions: - Wouldn't it be necessary (or meaningful) to seperate lexicons from non-lexical utterances if we compare Chinese and English motherese in terms of the use of contour acoustics? (since two languages differ in their use of F0 contour in communiation.) Actually, I don't think I got the point why they are comparing Chinese and English in particular. - in : a. What can we get if we measure Mean Frequency? What does that value tell us about acoustic/perceptual differences of contour shape? b. what is the use of "semitone" here in addition to 'Hz'? From: nagar@ling.ohio-state.edu The research focuses on the use of melodic utterances that parents use while communicating with babies. The way these melodic utterances function in child directed speech is the research question that the authors propose. They contend that melodic counters may function as ?mediators of effective communication?. The study was conducted with the help of two ethnic groups who were recorded talking to their kids in as natural environment as the study permitted. Eight care giving contexts were identified and the utterances of the care giver were analyses in accordance with those. Rising, falling, complex, bell-shaped contours were found to be associated with different kinds of child directed speech. It was found that the context in which a melodic utterance was made affected the way in which the contour will be shaped. The authors do not clearly state the implications of their study. That context has an affect on the way speech events are conducted is an established fact. This study is another way of establishing this relationship but how does it inform language acquisition? From: Junko Davis - It does not seem that they controlled the lexical tone effects for their analysis. Or at least their methodology on this issue is not clearly stated in the article, so I am not sure how they dealt with the lexical tones. - "Chinese mothers seemed to employ a number of intuitive "strategies" which allowed them to outwit the tone rules of their language." (p.438): This sounds very interesting, but there is no further mentioning of it. Did all mothers use this strategy? What was the frequency of utilization of this strategy? Was there any variation among individual mothers? From: ASIMINA SYRIKA -I think the writers should have compared more languages (at least 2 tone and 2 stress languages) and not just Mandarin Chinese and American English. I'm not convinced that melodic contours and pitch patterns used in motherese carry the similar meanings worldwide. I think that pitch contours can be very language specific across different cultures. -Are there any studies that support cross-linguistic universals in the speech of other caregivers, like fathers or older siblings?