LING 316 - Learning Sound Patterns

What makes each language unique?

Each language has its own set of sounds (phonemes) and rule for combining them

• Like Lego sets → each language has different shaped blocks and different building instruction

What is the “melodic system” of a language?

The tune of speech → tone, stress, length, etc.

• Think of English stressing syllables like a drum beat vs. Chinese using tones like musical notes

What sounds does a fetus hear in the womb?

Mostly muffled background noises → mother’s heartbeat, blood flow, digestion, and some external sounds

• Like hearing a concert through a wall → muffled, but still rhythmic

How well can a fetus recognize speech sounds?

About 30% of speech sounds are recognizable in the womb

What did Minai et al. (2017) test?

Whether fetuses in the third trimester can tell languages apart based on rhythm

What is the Rhythm-Based Language Discrimination Hypothesis?

Babies can distinguish language if their rhythms are very different

• Like telling rap from opera → you don’t need to understand the words, the beat gives it away

How did they measure fetal processing (Minai 2017)?

• Faster hear = more processing/attention

• Heart speed measured by “inter-beat interval (IBI)”

What languages did they use and why (Minai 2017)?

English (stress-timed) vs. Japanese (mora-timed)

• English has strong beats like drumming → Japanese has a steady and even rhythm like a ticking clock

How did they control for differences in speaker voice (Minai 2017)?

One bilingual speaker read both English and Japanese passages

How were the sounds played to the fetus (Minai 2017)?

Through a plastic cone and tubing placed 1 cm above the mother’s abdomen near the baby’s head

Mothers wore headphone so they didn’t react

What was the procedure (Minai 2017)?

• Familiarization phase → fetus hears English

• Novelty phase → fetus hears either English again or Japanese (2 conditions)

• Researchers compared IBI changes

What did they find (Minai 2017)?

Fetuses reacted when the language changed from English to Japanese

• Shows fetuses can discriminate based on rhythm → not just sound content

What does this suggest about language learning (Minai 2017)?

Babies’ brains are already tuning into their native language properties before birth

• Like downloading the “demo version” of your first language in the womb

Who were the babies tested in Mehler et al. (1988)?

4-day-old infants with French-speaking mothers

What method did they use (Mehler 1988)?

High-Amplitude Sucking Procedure (HAS) → pacifier test

• When babies are interested → they suck harder/faster on the pacifier

How does HAS work?

• Baby has a pacifier with a sensor

• Sucking rate shows attention/interest

• High Sucking = high interest

What are the phases of HAS?

• Conditioning → baby learns that sucking makes a sound play

• Habituation → baby hears the same sound repeatedly → gets bored → sucking slows

• Testing → New sound is played

  • If sucking rate goes back up → baby noticed the change

  • If it stays low → baby didn’t notice a difference

Like Netflix auto play → if a new show comes on and you perk up, you noticed the change

What was Experiment 1 (Mehler 1988)?

French and Russian spoken by a bilingual speaker

4 Conditions:

• French → French

• French → Russian

• Russian → Russian

• Russian → French

What were the results of Experiment 1 (Mehler 1988)?

Babies noticed when the language switched (French to Russian - vice versa)

How was Experiment 7 different (Mehler 1988)?

Same as Exp 1 but speech was filtered at 400 Hz → mimics muffled womb hearing

What did the results of Exp 7 shows (Mehler 1988)?

Same outcome → babies still noticed the language switch

• Even through the “womb filter” newborns could tell French and Russian apart