Essentials: The Science of Learning & Speaking Languages | Dr. Eddie Chang

What It Covers

Neuroscientist Dr. Eddie Chang explains the distinction between speech and language, the mechanics of vocal production involving larynx vibrations at 100–200 Hz, breakthroughs in brain-machine interface technology that restored communication to paralyzed patients, and the neuroscience behind stuttering, including auditory feedback loops and coordination breakdowns.

Key Questions Answered

• •Speech vs. Language Architecture: Speech refers specifically to the physical audio signal produced by the vocal tract, while language encompasses semantics, syntax, and pragmatics. Understanding this distinction matters clinically — patients with brainstem injuries can lose speech entirely while retaining full language comprehension, meaning communication tools must target the correct neural layer.
• •Vocal Mechanics — Larynx Frequency: The larynx generates voice by vibrating vocal folds at approximately 100 Hz in men and 200 Hz in women during exhalation. Everything above the larynx — tongue, lips, jaw, pharynx — then shapes that raw sound into recognizable consonants and vowels. Optimizing vocal performance starts with breath control and exhalation management.
• •BRAVO Trial — BCI Speech Decoding: Dr. Chang's BRAVO clinical trial implanted electrode arrays over speech-motor cortex in a patient paralyzed for 15 years. An AI algorithm trained over weeks decoded brain activity patterns into a 50-word vocabulary with autocorrect assistance — the first demonstrated decoding of intended speech directly from cortical signals in a paralyzed person.
• •Stuttering — Coordination Breakdown and Auditory Feedback: Stuttering is a speech-motor coordination failure, not a language or purely anxiety-driven disorder. Anxiety triggers episodes but does not cause the condition. Critically, altering auditory self-feedback — what a person hears themselves say in real time — can measurably reduce stuttering frequency, pointing toward feedback-based therapeutic interventions as a practical treatment pathway.
• •Avatar-Based Neuroprosthetics: The next phase of speech BCI involves decoding facial muscle movements and expressions alongside vocal signals to animate a personalized avatar in real time. This embodied feedback approach accelerates learning for neuroprosthetic users faster than text-on-screen methods, and positions paralyzed individuals to participate in increasingly virtual social environments.