In the Twilight Zone of Sound

Gary Lawrence Murphy garym at
Thu Nov 6 16:13:03 PST 2003

Sure, audio illusions, I could name a few, oddities of phonetics like
-ough and-uff for example, but /none/ of them came anything close to
what Diana Deutsch's research is finding ...

  In the Twilight Zone of Sound

     San Diego

  A short phrase of speech in Diana Deutsch's mellifluous voice,
  with its faded English accent, loops over and over from the
  loudspeakers in her sparse, windowless sound studio.

  "Sometimes behave so strangely," comes the phrase.

  "Sometimes behave so strangely."

  And, strangely, after only a few repetitions, a striking
  effect emerges. With the rhythms and cadences of her voice,
  the sounds become undeniably musical.

  "It matches to a simple melody playing in B major," exclaims
  Ms. Deutsch, still surprised years after she first noticed the

  "It shows that speech and music connect," she says, and then
  speculates: "It makes sense to me to think that originally
  speech" -- all speech, she means -- "was pitched as tones,"
  the way Chinese and Vietnamese are today.

  The auditory illusion is one of many that Ms. Deutsch, a
  professor of psychology here at the University of California
  at San Diego, has included on two compact discs, Phantom Words
  and Other Curiosities (Philomel Records), which has just
  appeared, and Musical Illusions and Paradoxes (Philomel
  Records, 1995).

  The illusions range from the beguilingly simple to, as she
  says, "true weirdness, the twilight zone." Appropriately,
  then, her first disc's introduction has a tinge of Rod Serling
  as she intones: "In listening to this CD, you enter the
  curious and paradoxical world of illusion. The sounds as they
  appear to you are not only different from those that are
  really present, but they sometimes behave so strangely as to
  seem quite impossible."

  In her laboratory, sitting at a bank of amplifiers, speakers,
  and other sound-system components, Ms. Deutsch cues up what
  she calls the "octave illusion." Through headphones, high
  tones sound in one ear, low tones in the other. But when the
  earphones are switched, the high and low tones are heard in
  the same ears as before.

  In fact, Ms. Deutsch reveals, the same pattern of alternating
  high and low tones is sounding in each earphone. For some
  reason, the listener's ears create "the peculiar impression
  that the high tone has migrated from one earphone to the
  other," and the low tone, too, in the opposite direction. "It
  was a shock," she says. "It was unbelievable [to stumble on
  that illusion]. It wasn't known at the time. I thought, 'Am I
  crazy? Has the world gone crazy?'"

  As simple as the pattern is, it is hardly ever heard
  correctly. Ms. Deutsch has found that, because of differences
  in which hemisphere of the brain dominates, right-handers tend
  to hear the high tones on the right, the low tones on the
  left, while left-handers' experience varies.

  Another illusion, which Ms. Deutsch calls the "tritone
  paradox," is one that confuses even accomplished musicians.
  Her computer creates an organlike sound equivalent to, say,
  all six C's on the piano keyboard. That is followed by the
  sound of all six F sharps -- half of an octave away. When some
  subjects listen to that tritone, they hear an ascending
  pattern. Others, listening to the same tritone, hear a
  descending pattern.

  "There is no correct answer," says Ms. Deutsch, "the ambiguity
  is built into the tones themselves." The effect, Ms. Deutsch
  believes, stems from the way people perceive and store musical
  scales in their brains. Each note in a particular scale
  resembles its counterpart in other octaves -- all the C notes,
  for example, sound like C notes. Whether or not they have any
  musical training, Ms. Deutsch conjectures, listeners have
  their own mental image, like a clock, of which notes of any
  octave are the "highest," regardless of actual pitch. So, when
  they hear two notes that are half an octave apart, the first
  may sound higher, or lower, than the second.

  What is going on? Clearly, says Ms. Deutsch, when it comes to
  hearing, "we pick and choose how to bundle things together,
  rather than admit chaos." It would appear that some part of
  our hearing mechanism registers pitch, another where sounds
  are coming from. Sound may be thought of as a bundle of
  attributes that can fragment, she says, so that one may hear
  tones in the "wrong" ear. The human auditory system is
  apparently more loosely constructed than the visual system, so
  "in the lab you can propel it into absurdities, even though
  mostly, in everyday life, it works well."

  Ms. Deutsch is tirelessly enthusiastic about hearing and
  sound. As a teenager, she hoped to become a professional
  pianist, but her parents persuaded her to pursue a
  more-certain financial future. She went to the University of
  Oxford and performed extremely well in psychology and
  philosophy, then came to California to earn a doctorate here
  at San Diego.

  She arrived just when computers were beginning to be used to
  manipulate sound, facilitating researchers' ability to analyze
  what, and how, people hear.

  The earliest illusion she wrote about, in the journal Science
  in 1970, may have implications for neuropsychology. "You'll
  hear two tones, with a series of tones, or words,
  intervening," she says, as she cues up another track of
  Phantom Words. "Try to remember whether the last tone is
  higher or lower in pitch than the first." The task proves much
  harder when other tones intervene than when words do.
  Cognitive and neuropsychologists have encouraged Ms. Deutsch
  to pursue the implications of that finding for understanding
  memory and for diagnosing the early signs of Alzheimer's

  Some of Ms. Deutsch's research reveals the links among
  hearing, music, and language acquisition. She is, for example,
  studying the incidence of perfect pitch -- the ability to
  consistently recognize and repeat sounds of a particular pitch
  -- in speakers of languages such as Chinese, in which words
  have different meanings depending on whether the syllables
  rise or fall in inflection. She and colleagues are studying
  populations of students at the University of Rochester's
  Eastman School of Music and a music school in Beijing to see
  if the Chinese students, who acquired perfect pitch for speech
  as infants, later tend to possess it for music. "It doesn't
  ring true that only some people would have perfect pitch," she
  says. It may, she says, be possible to help infants acquire
  the ability.

  The acquisition of toned speech, and of music, is strongly
  influenced by culture. A population of speakers agrees that
  certain sounds constitute certain words. One of Ms. Deutsch's
  illusions with cultural aspects always confuses and even
  bothers her test subjects, including whole classes of her
  students here. In the "phantom words" illusion, she plays
  fragments of words over and over. Different listeners hear
  different words emerge from the mesmerizing sequences of
  garbled fragments. In fact, many insist that she must have
  implanted the words they hear in the patterns. "People really
  do believe their own perceptions," says Ms. Deutsch. "That's
  how magicians manage, isn't it? But perception is not as good
  as people think it is. When people are told they've heard
  something else, some think it's incredibly funny. But
  sometimes I get anger."

  Subjects for whom English is a second, but fluent, language,
  often hear words in the language they first learned. Such is
  the power of early linguistic experience, Ms. Deutsch says.

  She also finds interesting differences between men and women.
  Many women hear words related to romance, or "words like 'no
  pie,' and 'Diet Coke,'" she says. "You can tell who's

  And the men? "They tend to hear things that make them ask me
  if it's OK to say them in polite company."

  She is testing her belief that the phantom-words illusion acts
  like a kind of auditory Rorschach test, diagnostic of people's
  states of mind. People suffering from depression tend to hear
  words like "attack," "blood," and "I'm dying," she says. "The
  data so far is sufficient to convince me without any doubt
  that this is going on."

  For another truly bizarre auditory illusion, or hallucination,
  she has no firm hypothesis, but is gathering data. Some
  people, usually elderly, constantly hear music -- not, say,
  the ringing of tinnitus, but actual tunes. In fact, of the
  dozens of people in several countries who have reported their
  experiences of this as-yet-unnamed syndrome to her, almost all
  hear male choirs singing patriotic songs, or a single male,
  baritone voice. Talk about a twilight zone.

  Those with the syndrome show no signs of psychiatric illnesses
  or brain tumors; they simply hear singing. "It's externalized
  music; it's not tunes running around in your head," says Ms.

  "Some people don't mind, and even find the music pleasant,"
  she says.

  "But with most people, it drives them nuts."


This article is available online at this address:

  From the issue dated November 7, 2003

Copyright 2003 by The Chronicle of Higher Education

More information about the FoRK mailing list