Alvin Meyer Liberman, whose ideas set the agenda for fifty years of research in the psychology of speech and laid the ground for modern computer speech synthesis, died on January 13th [2000] at a hospital in Mansfield, Connecticut. Dr. Liberman, former President and Director of Research at Haskins Laboratories in New Haven, Professor Emeritus of Psychology at the University of Connecticut and of Linguistics at Yale University, was 82. He died of complications following open heart surgery.

The goal of his early work, sponsored by the Veterans' Administration after World War II, was to develop the sound output of a reading machine for the blind, a device that would scan print and produce a distinctive acoustic pattern for each letter of the alphabet. Despite years of effort, Dr. Liberman and his colleagues at Haskins Laboratories never succeeded in devising an acoustic array that listeners could follow faster than Morse code, roughly 1/10 of a normal speaking rate, and intolerably slow for extended use.

This failure raised the question to which Dr. Liberman devoted much of the rest of his research life: Why is speech so much faster and more efficient as a carrier of linguistic information than other sounds? The answer gradually emerged from dozens of experiments in the 50s and 60s.

Speech is not an arbitrary signal that just happened to be available as language evolved; rather, speech is an integral part of language. Consonants and vowels, the discrete phonemic elements essential for a sizeable lexicon, do not combine like beads on a string, but are overlapped, or encoded, into syllables; speed is thus purchased at the price of complexity. Human listeners are biologically adapted to decode the continuously variable signal of running speech and to recover its discrete phonemic components.

In the course of developing this answer, Dr. Liberman and his colleagues discovered many of the main acoustic cues to the consonants and vowels of English. These cues later served to guide the development of artificial speech synthesis, now widely used for machine to human communication.

Dr. Liberman's provocative work was largely responsible for drawing speech into the mainstream of experimental cognitive psychology, where his "nativist" views were not to everyone's liking. But he thrived on controversy, and up until the last months of his life he maintained a steady stream of ingenious and telling experiments to support what he liked to call his "unconventional view" against the "conventional view" of most other experimental psychologists.

During the 70s and 80s, Dr. Liberman increasingly collaborated with his wife, the late Dr. Isabelle Yoffe Liberman, and other Haskins scientists on reading. A central discovery of this work was that children who have difficulty in learning to read almost always lack what Isabelle Liberman termed "phoneme awareness": they cannot easily learn to break a word into its component consonants and vowels. The critical requirement of phoneme awareness for learning to read alphabetic print is now internationally recognized, in large part due to the two Libermans' passionate advocacy of the "alphabetic principle" against the "whole word," or "sight reading" method of instruction.

Alvin Liberman was a member of the National Academy of Sciences and of the American Academy of Arts and Sciences, a Fellow of the American Psychological Assocation and of the Acoustical Society of America. He also received many other awards, including the Warren Medal from the Society of Experimental Psychologists; honorary doctoral degrees from the State University of New York and from the Universite Libre de Bruxelles, and a medal from the College de France, Paris. He is survived by two sons, Mark of Ardmore, Pennsylvania, and Charles of Milton, Massachusetts; by a daughter, Sarah Ash, of Raleigh, North Carolina, and by nine grandchildren.