Summary: A University of Kansas study using fetal biomagnetometry found that fetuses late in pregnancy react differently to the rhythms of two distinct languages. Changes in fetal heart rate indicate that babies begin to attune to the rhythmic patterns of the language they will learn before birth.
Source: University of Kansas
Fetuses can distinguish between rhythmically different languages—such as English and Japanese—about one month before birth.
Researchers in the Department of Linguistics at the University of Kansas used a non-invasive magnetocardiography (MCG) device at the Hoglund Brain Imaging Center to test whether unborn babies respond to the rhythmic properties of language. Their findings, published in NeuroReport, demonstrate that fetuses show distinct heart-rate responses when exposed to speech in different rhythmic classes, supporting the idea that language-related auditory sensitivity begins in utero.
“Evidence shows that newborns are sensitive to rhythmic differences between languages,” said Utako Minai, associate professor of linguistics and lead author of the study. “Past infant studies measured behavior—like changes in pacifier-sucking rates—when speech switched from one rhythmic type to another. We wanted to find out whether this sensitivity emerges before birth.”
Minai noted that sounds reach the fetus in a muffled form through the womb, but rhythm and prosodic patterns remain largely preserved. Because the rhythmic structure of speech survives this attenuation, the prenatal auditory environment provides cues that may shape early language acquisition.
Previous fetal studies used ultrasound to infer heart-rate reactions to speech, but those experiments left open the possibility that fetuses were responding to differences between speakers rather than to language rhythm itself. To address that limitation, Minai and her colleagues had a bilingual speaker record passages in both English and Japanese so that both languages were presented in the same voice.
Twenty-four pregnant women, on average about 35.5 weeks gestation, participated in the study. The research team used one of only two dedicated fetal biomagnetometers in the United States, which fits over the maternal abdomen and detects minute magnetic fields generated by electrical activity in the maternal and fetal bodies. The device records fetal heart activity with higher sensitivity than ultrasound, allowing researchers to measure beat-to-beat changes.
“The heart itself does not hear; any heart-rate change in response to speech must be mediated by the fetal brain,” said Kathleen Gustafson, research associate professor of neurology at the Hoglund Brain Imaging Center and a co-investigator on the project. “Our sensor captures fetal cardiac responses that reflect neural processing of auditory input.”
In the experiment, all participants first heard a two-minute English passage read by the same bilingual speaker. After an 18-minute interval, half of the participants heard a different English passage, while the other half heard a Japanese passage. English and Japanese were chosen because they are classically distinguished by rhythm: English has a variable, stress-timed rhythm, while Japanese exhibits a more regular, mora-timed rhythm.
The researchers reconstructed the fetal magnetocardiogram, calculated mean interbeat intervals for baseline and for the first 30 seconds of each passage, and compared the changes. Fetuses exposed to English first and then to Japanese showed a significant heart-rate change when the second, rhythmically novel passage was played. By contrast, fetuses exposed to two English passages did not show the same change.
“The results were statistically robust,” Minai said. “They indicate that fetuses detect a change in language rhythm when the speech pattern shifts from English to Japanese. This suggests that prenatal exposure to speech provides rhythmic cues that help the developing brain begin tuning to the language it will acquire after birth.”
Gustafson added that the intrauterine environment is acoustically rich—maternal sounds, internal noises, and external inputs all contribute to auditory stimulation. “Without such exposure, the auditory cortex would receive less stimulation. These findings provide evidence that prenatal sound exposure contributes to early neural development related to language.”
Funding: The study was supported by a National Institutes of Health Clinical and Translational Science Award grant to the KU Medical Center. The research team included Utako Minai, Kathleen Gustafson, Robert Fiorentino, Allard Jongman, and Joan Sereno.
Source: Rick Hellman, University of Kansas
Fetal rhythm-based language discrimination: a biomagnetometry study
This study used fetal biomagnetometry to measure fetal heart-rate changes as an index of discrimination between two rhythmically different languages, English and Japanese. Two-minute passages in each language were recorded by the same bilingual speaker. Twenty-four mother–fetus pairs (mean gestational age 35.5 weeks) participated. Each fetus heard an initial English passage and then, after an 18-minute interval, either a second English passage (English–English condition, N=12) or a Japanese passage (English–Japanese condition, N=12). The fetal magnetocardiogram was processed using independent components analysis to isolate fetal cardiac signals. Mean interbeat intervals (IBIs) were calculated for a 30-second baseline preceding each passage and for the first 30 seconds of each passage; the baseline IBI was subtracted from the passage IBI to yield an IBI change score. Results showed a significant interaction between condition and passage: fetuses in the English–Japanese condition exhibited a larger IBI change (reflecting a faster heart rate) for the novel second passage than for the initial passage, while fetuses in the English–English condition did not. These findings provide biomagnetometry evidence that fetuses discriminate languages based on rhythmic properties, supporting the hypothesis that rhythm is a prenatally available building block of language acquisition.
Reference: Utako Minai; Kathleen Gustafson; Robert Fiorentino; Allard Jongman; Joan Sereno. Published online May 19, 2017 in NeuroReport. doi:10.1097/WNR.0000000000000794