Previous studies of intonation in Puerto Rican Spanish (PRS) have shown that adults use three distinct intonation contours, or melodies, to convey different types of information when they ask yes-no questions: one for when the speaker doesn't have any idea what the answer is, one for when the speaker thinks the information being asked about is true, and one for when the speaker is asking about something s/he did not expect to be true (counter-expectation). How is it that children learn these contours and their meanings?
Ms. Armstrong will travel to Puerto Rico for two months to carry out two experiments. The first experiment is designed to test how adults perceive the three contours mentioned above when they are heard in different types of speech contexts, allowing a better description of the adult system that children acquire. The second experiment is a task designed to assess how well 4-8 year olds are able to use intonation to understand what a speaker believes. For instance, when children hear counter-expectation intonation, are they able to attribute it to a speaker in an unexpected or surprising situation? If younger children tend to pass the task, it would refute claims that sentence level comprehension for intonation is late-acquired.
This is the first study to closely examine how children use intonation as an indicator of speakers' thoughts and beliefs and will be of interest to both linguists and developmental psychologists.
This project was concerned with the intonation of yes-no question intonation in Spanish. The first results from the dissertation looked at how adult speakers of Puerto Rican Spanish (PRS) produce and perceive question intonation. For instance, even in English we might use different intonation, or melodies, for the questions in (1) and (2): (1) John's coming to dinner? (2) John's coming to dinner!?!? In (1) a speaker might be confirming something s/he had inferred, where as in (2) s/he expresses disbelief about John coming to dinner, something she would have learned from a given situation or context. Three distinct intonation contours were identified in the speech of adult Puerto Ricans. They were then tested in a perception experiment to understand the meaning that Puerto Rican listeners perceive when they hear the contours in context. Together, results revealed that PRS has a general, default question intonation that can be used for any type of yes-no question. In the Sp_ToBI system used for labeling intonation in Spanish, the ¡H* L% label is used for this contour. There were two more contours in the data found to be used for more specific meanings. The contour labeled H+L* L% conveys a speaker's positive bias about the content of the question, as in (3): (3) Joanna sees a man at the mall and is very sure she met him at her work orientation. She says to him: Did I meet you at orientation? One way to ask the question in (3) in American English is with falling intonation, showing that the speaker is quite sure she met the man. This is not unlike the H+L* L% contour identified in Puerto Rican Spanish (PRS). Additionally, PRS uses the L*HL% contour for questions like (2), where the speaker is in a state of disbelief or incredulity about the content in the question. How do PRS children acquire these three contours? To answer this question I analyzed two longitudinal corpora of PRS child speech and caretaker speech (two PRS-acquiring toddlers and their caretakers). I analyzed each occurrence of a yes-no question in the corpora, and the pragmatic contexts they were produced in. Children and caretakers were shown to have an overwhelming preference for the default, general contour ¡H* L%. The caretakers used more specific contours H+L* L% and L*HL% substantially less. The only other contour the children produced was H+L*L%, but never produced L*HL% in an adult-like way. Since children didn't show evidence for L*HL% acquisition in production, a comprehension experiment was carried out to evaluate when children could reliably comprehend the L*HL% contour in context. Thirty-six PRS-speaking 4-, 5- and 6-year-olds partcipated in the comprehension study. While both 4- and 5-year-olds comprehended the disbelief meaning associated with L*HL% at above-chance levels, 6-year-olds were significantly better than both of the younger age groups. In order to understand disbelief meaning, children must be able to evaluate their own beliefs about ideas in the world. Another part of the grammar that children must acquire related to this type of development is that of modal verbs like must or might. Children are known to comprehend distinctions in modal verbs between ages 4 and 6 as well. Thus it may be possible that as children develop the ability to make assessments about thoughts and ideas in the world, linguistic forms associated with these abilities emerge.This should be investigated in more detail. Some researchers claim that there is a production/perception paradox for intonational development such that children produce some types of intonation before they comprehend them. This work makes clear that it is paramount to understand prosodic meaning in order to study its acquisition. Not all questions are aqcuired in the same way, due to effects of both frequency and the specific meaning of the contour. From a more applied standpoint, this research is important in that it pinpoints intonation related to belief states. The specific problems that autistic children (as well as adults) have with intonation are still unclear. Understanding how typically-developing children perform on tasks related to belief-state intonation can be useful in identifying the specific aspects of prosody that autistic children and adults might struggle with. This area is particularly attractive to explore since assessment of belief states, or Theory of Mind, is known to be compromised for many individuals with ASD. It is the researcher's hope to develop the work further so that it may benefit this population. Additionally, the researcher has developed a blog with a general description of the different contour types in PRS. It will be a useful teaching tool in second language classrooms, as well as a resource that the PRS-speaking community will now have access to. The blog is available at: http://entonacionpr.wordpress.com/ Data from the research are also available on the Atlas of Spanish Intonation: http://prosodia.upf.edu/atlasentonacion/index-english.html
