Acoustic analysis of typically developing elementary school-aged (prepubertal) children’s speech has been primarily performed on cross-sectional data in the past. Few studies have examined longitudinal data in this age group. For this presentation, we analyze the developmental changes in the acoustic properties of children’s speech using data collected longitudinally over four years (from first grade to fourth grade). Four male and four female children participated in this study. Data were collected once every year for each child. Using these data, we measured the four-year development of subglottal acoustics (first two subglottal resonances) and vowel acoustics (first four formants and fundamental frequency). Subglottal acoustic measurements are relatively independent of context, and average values were obtained for each child in each year. Vowel acoustics measurements were made for seven vowels (i, ɪ, ɛ, æ, ʌ, ɑ, u), each occurring in two different words in the stressed syllable. We investigated the correlations between the children’s subglottal acoustics, vowel acoustics, and growth-related variables such as standing height, sitting height, and chronological age. Gender-, vowel-, and child-specific analyses were carried out in order to shed light on how typically developing speech acoustics depend on such variables. [Work supported, in part, by the NSF.]
more »
« less
This content will become publicly available on August 20, 2024
Developmental Articulatory and Acoustic Features for Six to Ten Year Old Children
In this paper, we study speech development in children using longitudinal acoustic and articulatory data. Data were collected yearly from grade 1 to grade 4 from four female and four male children. We analyze acoustic and articulatory properties of four corner vowels: /æ/, /i/, /u/, and /A/, each occurring in two different words (different surrounding contexts). Acoustic features include formant frequencies and subglottal resonances (SGRs). Articulatory features include tongue curvature degree (TCD) and tongue curvature position (TCP). Based on the analyses, we observe the emergence of sex-based differences starting from grade 2. Similar to adults, the SGRs divide the vowel space into high, low, front, and back regions at least as early as grade 2. On average, TCD is correlated with vowel height and TCP with vowel frontness. Children in our study used varied articulatory configurations to achieve similar acoustic targets.
more »
« less
- Award ID(s):
- 2006979
- NSF-PAR ID:
- 10470344
- Publisher / Repository:
- ISCA
- Date Published:
- Page Range / eLocation ID:
- 4598 to 4602
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
A system for the lateral transfer of information from end-to-end neural networks recognizing articulatory feature classes to similarly structured networks recognizing phone tokens is here proposed. The system connects recurrent layers of feature detectors pre-trained on a base language to recurrent layers of a phone recognizer for a different target language, this inspired primarily by the progressive neural network scheme. Initial experiments used detectors trained on Bengali speech for four articulatory feature classes—consonant place, consonant manner, vowel height, and vowel backness—attached to phone recognizers for four other Asian languages (Javanese, Nepali, Sinhalese, and Sundanese). While these do not currently suggest consistent performance improvements across different low-resource settings for target languages, irrespective of their genealogic or phonological relatedness to Bengali, they do suggest the need for further trials with different language sets, altered data sources and data configurations, and slightly altered network setups.more » « less
-
more » « less
Abstract Music and language are two fundamental forms of human communication. Many studies examine the development of music‐ and language‐specific knowledge, but few studies compare how listeners know they are listening to music or language. Although we readily differentiate these domains, how we distinguish music and language—and especially speech and song— is not obvious. In two studies, we asked how listeners categorize speech and song. Study 1 used online survey data to illustrate that 4‐ to 17‐year‐olds and adults have verbalizable distinctions for speech and song. At all ages, listeners described speech and song differences based on acoustic features, but compared with older children, 4‐ to 7‐year‐olds more often used volume to describe differences, suggesting that they are still learning to identify the features most useful for differentiating speech from song. Study 2 used a perceptual categorization task to demonstrate that 4–8‐year‐olds and adults readily categorize speech and song, but this ability improves with age especially for identifying song. Despite generally rating song as more speech‐like, 4‐ and 6‐year‐olds rated ambiguous speech–song stimuli as more song‐like than 8‐year‐olds and adults. Four acoustic features predicted song ratings: F0 instability, utterance duration, harmonicity, and spectral flux. However, 4‐ and 6‐year‐olds’ song ratings were better predicted by F0 instability than by harmonicity and utterance duration. These studies characterize how children develop conceptual and perceptual understandings of speech and song and suggest that children under age 8 are still learning what features are important for categorizing utterances as speech or song.
Research Highlights Children and adults conceptually and perceptually categorize speech and song from age 4.
Listeners use F0 instability, harmonicity, spectral flux, and utterance duration to determine whether vocal stimuli sound like song.
Acoustic cue weighting changes with age, becoming adult‐like at age 8 for perceptual categorization and at age 12 for conceptual differentiation.
Young children are still learning to categorize speech and song, which leaves open the possibility that music‐ and language‐specific skills are not so domain‐specific.
-
International Ocean Discovery Program Expedition 397T sought to address the shortage of drilling time caused by COVID-19 mitigation during Expedition 391 (Walvis Ridge Hotspot) by drilling at two sites omitted from the earlier cruise. A week of coring time was added to a transit of JOIDES Resolution from Cape Town to Lisbon, which would cross Walvis Ridge on its way north. These two sites were located on two of the three seamount trails that emerge from the split in Walvis Ridge morphology into several seamount chains at 2°E. Site U1584 (proposed Site GT-6A) sampled the Gough track on the southeast side of the hotspot track, and Site U1585 (proposed Site TT-4A) sampled the Tristan track on the northwest side. Together with Site U1578, drilled on the Center track during Expedition 391, they form a transect across the northern Walvis Ridge Guyot Province. The goal was to core seamount basalts and associated volcanic material for geochemical and isotopic, geochronologic, paleomagnetic, and volcanological study. Scientifically, one emphasis was to better understand the split in isotopic signatures that occurs at the morphologic split. Geochronology would add to the established age progression but also give another dimension to understanding Walvis Ridge seamount formation by giving multiple ages at the same sites. The paleomagnetic study seeks to establish paleolatitudes for Walvis Ridge sites for comparison with those published from hotspot seamount chains in the Pacific, in particular to test whether a component of true polar wander affects hotspot paleolatitude. Hole U1584A cored a 66.4 m thick sedimentary and volcaniclastic section with two lithostratigraphic units. Unit I is a 23 m thick sequence of bioturbated clay and nannofossil chalk with increasing volcaniclastic content downhole. Unit II is a >43 m thick sequence of lapillistone with basalt fragments. Because the seismic section crossing the site shows no evidence as to the depth of the volcaniclastic cover, coring was terminated early. Because there were no other shallow sites nearby with different characteristics on existing seismic lines, the unused operations time from Site U1584 was shifted to the next site. The seismic reflector interpreted as the top of igneous rock at Site U1585 once again resulted from volcaniclastic deposits. Hole U1585A coring began at 144.1 mbsf and penetrated a 273.5 m thick sedimentary and volcaniclastic section atop a 81.2 m thick series of massive basalt flows. The hole was terminated at 498.8 mbsf because allotted operational time expired. The sedimentary section contains four main lithostratigraphic units. Unit I (144.1–157.02 mbsf) is a bioturbated nannofossil chalk with foraminifera, similar to the shallowest sediments recovered at Site U1584. Unit II (157.02–249.20 mbsf), which is divided into two subunits, is a 92.2 m thick succession of massive and bedded pumice and scoria lapillistone with increased reworking, clast alteration, and tuffaceous chalk intercalations downhole. Unit III (249.20–397.76 mbsf) is 148.6 m thick and consists of a complex succession of pink to greenish gray tuffaceous chalk containing multiple thin, graded ash turbidites and tuffaceous ash layers; intercalated tuffaceous chalk slumps; and several thick coarse lapilli and block-dominated volcaniclastic layers. Befitting its complexity, this unit is divided into eight subunits (IIIA–IIIH). Three of these subunits (IIIA, IIID, and IIIG) are mainly basalt breccias. Unit IV (397.76–417.60 mbsf) is a volcanic breccia, 19.8 m thick, containing mostly juvenile volcaniclasts. The igneous section, Unit V (417.60–498.80 mbsf) is composed of a small number of massive basaltic lava flows. It is divided into three igneous lithologic units, with Unit 2 represented by a single 3 cm piece of quenched basalt with olivine phenocrysts in a microcrystalline groundmass. This piece may represent a poorly recovered set of pillow lavas. Unit 1 is sparsely to highly olivine-clinopyroxene ± plagioclase phyric massive basalt and is divided into Subunits 1a and 1b based on textural and mineralogical differences, which suggests that they are two different flows. Unit 3 also consists of two massive lava flows with no clear boundary features. Subunit 3a is a 10.3 m thick highly clinopyroxene-plagioclase phyric massive basalt flow with a fine-grained groundmass. Subunit 3b is a featureless massive basalt flow that is moderately to highly clinopyroxene-olivine-plagioclase phyric and >43.7 m thick. Alteration of the lava flows is patchy and moderate to low in grade, with two stages, one at a higher temperature and one at a low temperature, both focused around fractures. The Site U1585 chronological succession from basalt flows to pelagic sediment indicates volcanic construction and subsidence. Lava eruptions were followed by inundation and shallow-water volcaniclastic sediment deposition, which deepened over time to deepwater conditions. Although the massive flows were probably erupted in a short time and have little variability, volcaniclasts in the sediments may provide geochemical and geochronologic data from a range of time and sources. Chemical analyses indicate that Site U1585 basalt samples are mostly alkalic basalt, with a few trachybasalt flow and clast samples and one basaltic trachyandesite clast. Ti/V values lie mostly within the oceanic island basalt (OIB) field but overlap the mid-ocean-ridge basalt (MORB) field. Only a handful of clasts from Site U1584 were analyzed, but geochemical data are similar. Paleomagnetic data from Site U1585 indicate that the sediments and basalt units are strongly magnetic and mostly give coherent inclination data, which indicates that the basaltic section and ~133 m of overlying volcaniclastic sediment is reversely polarized and that this reversal is preserved in a core. Above this, the rest of the sediment section records two normal and two reversed zones. Although there are not enough basalt flows to give a reliable paleolatitude, it may be possible to attain such a result from the sediments.more » « less
-
International Ocean Discovery Program Expedition 397T sought to address the shortage of drilling time caused by COVID-19 mitigation during Expedition 391 (Walvis Ridge Hotspot) by drilling at two sites omitted from the earlier cruise. A week of coring time was added to a transit of JOIDES Resolution from Cape Town to Lisbon, which would cross Walvis Ridge on its way north. These two sites were located on two of the three seamount trails that emerge from the split in Walvis Ridge morphology into several seamount chains at 2°E. Site U1584 (proposed Site GT-6A) sampled the Gough track on the east, and Site U1585 (proposed Site TT-4A) sampled the Tristan track on the west. Together with Site U1578, drilled on the Center track during Expedition 391, they form a transect across the northern Walvis Ridge Guyot Province. The goal was to core seamount basalts and associated volcanic material for geochemical and isotopic, geochronologic, paleomagnetic, and volcanologic study. Scientifically, one emphasis was to better understand the split in geochemical and isotopic signatures that occurs at the morphologic split. Geochronology would add to the established age progression but also give another dimension to understanding Walvis Ridge seamount formation by giving multiple ages at the same sites. The paleomagnetic study seeks to establish paleolatitudes for Walvis Ridge sites for comparison with those published from hotspot seamount chains in the Pacific, in particular to test whether a component of true polar wander affects hotspot paleolatitude. Hole U1584A cored a 66.4 m thick sedimentary and volcaniclastic section with two lithostratigraphic units. Unit I is a 23 m thick sequence of bioturbated clay and nannofossil chalk with increasing volcaniclastic content downhole. Unit II is a >43 m thick sequence of lapillistone with basalt fragments. Because the seismic section crossing the site shows no evidence as to the depth of the volcaniclastic cover, coring was terminated early. Because there were no other shallow nearby sites with different character on existing seismic lines, the unused operations time from Site U1584 was shifted to the next site. The seismic reflector interpreted as the top of igneous rock at Site U1585 once again resulted from volcaniclastic deposits. Hole U1585A coring began at 144.1 mbsf and penetrated a 273.5 m thick sedimentary and volcaniclastic section atop a 81.2 m thick series of massive basalt flows. The hole was terminated at 498.8 mbsf because allotted operational time expired. The sedimentary section contains four main units. Unit I (144.1–157.02 mbsf) is a bioturbated nannofossil chalk with foraminifera, similar to the shallowest sediments recovered at Site U1584. Unit II (157.02–249.20 mbsf), which is divided into two subunits, is a 92.2 m thick succession of massive and bedded pumice and scoria lapillistone with increased reworking, clast alteration, and tuffaceous chalk intercalations downhole. Unit III (249.20–397.76 mbsf) is 148.6 m thick and consists of a complex succession of pink to greenish gray tuffaceous chalk containing multiple thin, graded ash turbidites and tuffaceous ash layers; intercalated tuffaceous chalk slumps; and several thick coarse lapilli and block-dominated volcaniclastic layers. Befitting the complexity, it is divided into eight subunits (IIIA–IIIH). Three of these subunits (IIIA, IIID, and IIIG) are mainly basalt breccias. Unit IV (397.76–417.60 mbsf) is a volcanic breccia, 19.8 m thick, containing mostly juvenile volcaniclasts. The igneous section, Unit V (417.60–498.80 mbsf) is composed of a small number of massive basaltic lava flows. It is divided into three lithologic units, with Unit 2 represented by a single 3 cm piece of quenched basalt with olivine phenocrysts in a microcrystalline groundmass. This piece may represent a poorly recovered set of pillow lavas. Unit 1 is sparsely to highly olivine-clinopyroxene ± plagioclase phyric massive basalt and is divided into Subunits 1a and 1b based on textural and mineralogical differences, which suggests that they are two different flows. Unit 3 also consists of two massive lava flows with no clear boundary features. Subunit 3a is a 10.3 m thick highly clinopyroxene-plagioclase phyric massive basalt flow with a fine-grained groundmass. Subunit 3b is a featureless massive basalt flow that is moderately to highly clinopyroxene-olivine-plagioclase phyric and >43.7 m thick. Alteration of the lava flows is patchy and moderate to low in grade, with two stages, one at a higher temperature and one at a low temperature, both focused around fractures. The Site U1585 chronologic succession from basalt flows to pelagic sediment indicates volcanic construction and subsidence. Lava eruptions were followed by inundation and shallow-water volcaniclastic sediment deposition, which deepened over time to deepwater conditions. Although the massive flows were probably erupted in a short time and have little variability, volcaniclasts in the sediments may provide geochemical and geochronologic data from a range of time and sources. Chemical analyses indicate that Site U1585 basalt samples are mostly alkalic basalt, with a few trachybasalt flow and clast samples and one basaltic trachyandesite clast. Ti/V ratios lie mostly within the oceanic island basalt (OIB) field but overlap the mid-ocean-ridge basalt (MORB) field. Only a handful of clasts from Site U1584 were analyzed, but geochemical data are similar. Paleomagnetic data from Site U1585 indicate that the sediments and basalt units are strongly magnetic and mostly give coherent inclination data, which indicates that the basaltic section and ~133 m of overlying volcaniclastic sediment is reversely polarized and that this reversal is preserved in a core. Above this, the rest of the sediment section records two normal and two reversed zones. Although there are not enough basalt flows to give a reliable paleolatitude, it may be possible to attain such a result from the sediments.more » « less
