Strange sounds and the cadences of repetitive speech filled the dementia day unit. In the noisy chaos, some residents slowly limped around the room or along the halls; others sat, heads down, silent, seemingly unaware of their surroundings. Here were all the human losses that we associate with dementia, stroke, and late stages of neurologic diseases. Visitors to nursing homes know them well, and most assume, as I did, that the lost functions are gone forever. Indeed, I asked myself that day, what could music possibly do for men and women so afflicted? Still, I had come to begin my work as a music therapist, so I sat down at the piano and started to play "Let Me Call You Sweetheart."

At first, I could barely hear myself play. But after a few minutes, the sound of singing began to rise above the noise, then dominate it. As I watched, even the silent patients turned their gaze to me. It was too remarkable a change to assign only to the allure of an old familiar song. People who had seemed unable to focus became attentive. Residents whom I knew to have limited cognitive skills had recognized the melody; their voices found the right words. Some with seemingly uncontrollable repetitive movements now kept steady time with their hands and feet.

I wondered: Could our processing of music be so different, or so basic, that abilities relating to it remained accessible in people so limited in function? In 1978, little was known about music and brain function. Today, as a result of exponentially increasing research, particularly over the past five years, we can venture some initial answers to my question.

Music's Forgotten Secrets

Music predates recorded history, but its roots may lie in early human communication and rituals for healing. In traditional African cultures and rain forest cultures in other parts of the world, for example, music is connected with many of life's vital patterns and occasions. In Western culture, however, as music became increasingly accepted as an art form, its therapeutic properties were mostly forgotten—rediscovered only when music therapy became an organized field in the early 1950s. Since then, a torrent of peer-reviewed clinical and scientific studies have focused on music's therapeutic value in areas from reducing pain, to improving memory and cognition, to helping motor function. But even though we know how effective music therapy can be, the investigation of its effects on recovery of function in people with neurologic impairment is new and exceedingly challenging.

Music is a complex stimulus, involving everything from pitch to rhythm, melody to volume. Consequently, it is not processed in a single area of the brain. We can see this in what is called "amusia," in which a single musical skill is lost when a specific area of the brain is damaged—for example, loss of pitch perception resulting from lesions to the right temporal lobe. But while a component of music, such as pitch, may be processed in a specific region of the brain, the overall experience of music is a gestalt of perceptual and psychological processes occurring in synchrony and involving a spectrum of neurologic activity and brain regions.

We now know from clinical case studies that music can affect—in very specific ways—human neurological, psychological, and physical functioning in areas such as learning, processing language, expressing emotion, memory, and physiological and motor responses. How your brain perceives and processes music also differs depending on whether or not you are a musician. The effects of music raise intriguing questions about both early brain development and brain plasticity later in life.

Sam: Getting The Beat

Sam, a man in his late 60s, was recovering from a stroke. His physical therapist rated him a "guarded walker"—able to shuffle along with a quad cane, but not steady enough to walk outdoors, where he might have difficulty negotiating the uneven pavement. Because his left side was weak, his left foot dragged along the floor, causing him to take faltering steps. Each step was slow and hesitant, as Sam focused intensely on the process of walking. After he had been in traditional physical therapy for two months, and was showing little further improvement, he was referred to music therapy in the hope that he could improve his sense of his body's position and his balance.

The physical therapist tested Sam's gait, and I found music with a tempo that matched the pace of his stride. He knew the music and was comfortable walking to it. In fact, he told me how, as a teenager, he used to go dancing every week at the gym. As he walked, he became more confident of his movements. Amazingly, he began to add dance steps, sliding his feet or clicking his heels. He said he couldn't help it; it just happened. He wasn't "thinking about walking," he said, he was "thinking about dancing."

Could there be a separate motor template for these dance movements, so different from walking? Or was it lack of conscious motor planning on Sam's part that freed up his motor cortex to send the necessary signals to his legs? As the sessions went on, he became more inventive in his movements. After several weeks of twice-weekly meetings, he began lifting his left foot off the floor. Now his steps were in perfect time to the rhythm of the music. He was not consciously aware of this, but he said that he could feel the tempo in his leg and thought that he was able to feel the floor with his left foot. This suggested that he was regaining sensation and control in that side of his body. But when the music stopped, Sam would again shuffle and drag the affected leg. We worked together for two months, twice a week, and his physical therapist also had Sam sing the song to himself as he walked in the rehabilitation gym. The music's rhythm was an external cue that organized Sam's walking without conscious effort.

Rhythm is, in fact, the primary property of music and is critical to human life in other ways. Plato defined rhythm as "the order in movement," and the temporal structure of music (its movement) has suggestive parallels in human motor development. At five months of age, when a fetus's neural circuits and auditory memory are forming, it experiences rhythm through the mother's heartbeat and respiration. Immediately after birth, basic motor patterns begin to develop. While eating, crawling, and walking, each child finds a cadence, particular motor rhythms that will remain fairly consistent throughout life. Our natural and spontaneous body movements may be outward representations of inner timing mechanisms. Leon Glass, Ph.D., at McGill University, and other scientists are investigating the complex mathematics of physiological rhythms and how they interact to maintain our health. We know that an alteration in internal rhythm—cardiac arrhythmia, for example— can be the harbinger of ill health or death.

Some internal rhythms can come to match external rhythms. In effect, a rhythm in the external world is heard and internalized, evoking an answering rhythm within us. When we understand how and when external auditory rhythms, or cues, influence various internal timing mechanisms, rhythm can become a powerful therapeutic tool.

The effect of external rhythmic cues on motor function, as we saw with Sam, is a prime example of how this influence occurs. Brain-imaging studies show that an area in the prefrontal motor cortex will start to become active at precise intervals in anticipation of a sequence of motor activity, such as finger tapping at one-second intervals. The resiliency of this motor-timing mechanism is strikingly apparent in people whose motor control, or motor initiation, has been lost as a result of a stroke or Parkinson's disease, but whose brains still respond to a rhythmic stimulus.

In neuromuscular diseases affecting the ability to initiate and control movement, external rhythm seems to supply the timing information that makes movement possible. For Sam, even singing the song to himself provided the required neurologic benefit, the external cue. Writer and neurologist Oliver Sacks, M.D., author of The Man Who Mistook His Wife for a Hat, eloquently describes a similar response to music in one of his post-encephalitic patients, who had great difficulty walking alone but walked perfectly if someone walked with her—or could time her steps to music. She said: "Whether it is others, in their own natural movement, or the movement of music itself, the feeling of movement, of living movement, is communicated to me. And not just movement, but existence itself." Sacks studied this phenomenon in the EEGs of some of these patients when they merely imagined a specific piece of music. Although their regular EEGs were very abnormal—the brain was slow on one side while convulsive on the other, for example— when they played the piano or simply imagined a piece of music, their EEGs became more normal.

Why Movement Responds To Rhythm

Michael Thaut, Ph.D., and his colleagues at Colorado State University suggest that the sensitivity of our motor systems to influences from sounds may have developed during human evolution so we could use the way we process what we hear to enhance our ability to organize and control our movements. Our basic auditory-arousal mechanisms (for example, our movements in reaction to a sudden loud noise) operate primarily through the amygdala in the brain's limbic system and may have originated in adaptive evolutionary processes, namely, the fight-or-flight response. In any case, the auditory system has connections to the brain stem, midbrain, and higher cortical structures, and normal motor function requires that these subcortical and cortical regions work in concert with each other.

The basal ganglia, a brain region affected in Parkinson's disease, provides a link to still other areas of the brain that connect mental processes and the initiation of movement. While the thought or wish to move depends on higher cortical processing, the actual ability to move depends on lower brain regions. If the higher cognitive processes that can initiate movement are damaged in traumatic brain injury or stroke, the requisite will to move may nevertheless get a "jump-start" by stimulating motor nerves that are still functional. Does the patterned auditory cue supplied by musical rhythms excite the more primitive motor areas first, and only then recruit or drive higher cortical circuits into action?

New evidence from studies by Wen Jun Gao, Ph.D., and Sarah L. Pallas, Ph.D., at Georgia State University suggests that learning, or at least the organization and development of cortical circuits in the brain, is influenced by patterned sensory activity, such as listening to sound clicks presented at specific time intervals. If such sensory signals turn out to enhance neural development, what role does rhythm—patterned auditory stimulation—play in the restimulation of these networks once they have been laid down? In patients like Sam, regaining physical function began on a spontaneous, unconscious level, indicating that the subcortical areas of his brain were being activated before the restoring of the higher cortical areas involved with the thought and the intent to initiate movement.

Mary: Rhythm And Melody Find A Voice

Rhythm also has a therapeutic effect for people with dysarthria, a motor/speech problem that occurs when functioning of the vocal organs is impaired. Dysarthria results in poor articulation of words; speech is slurred and, in the most severe cases, unintelligible.

Mary, a 56-year-old music therapy patient, had been in a coma for three months. It left her with severe dysarthria— a lack of vocal tone and severely distorted articulation. Spasmodic tremors contributed to the severity of her symptoms, and she had an open tracheotomy that made vocal sound production even more difficult.

Because weak muscles made her breath control poor, she also had difficulty sustaining any sounds she did make. Mary's overall comprehension of language, however, was intact. She was getting speech therapy to help her produce adequate yes or no responses, develop techniques for functional communication, and maximize existing "mouthing" skills, including motor gestures like chewing and yawning.

We knew that Mary had sung in her church choir and was familiar with many old hymns. In fact, even with her inability to sustain any intelligible sounds, she participated in weekly music therapy sessions on her hospital unit, silently smiling at the old tunes. With encouragement, she would attempt to sing along. I could see that her problem resulted in part from lack of coordination between her breathing and her attempts to form a sound, so I asked her to tap her finger as she tried to make a sound. Just that rhythm imparted enough coordination to gain some success, and soon she could sustain the tone for longer.

Once Mary became aware of her increasing ability to alternate breathing and making sounds, in a pattern cued by her tapping finger, she carried this ability over to pacing syllables and short phrases in speech. Before she started music therapy, she could articulate three-syllable phrases with the help of some cueing to breathe at the initiation of the phrase. She also knew the skills she needed to succeed: breathe, speak slowly, exaggerate articulation, and make a syllable-by-syllable attack. She could repeat single words and phrases, albeit with many attempts at self-correction.

In her music therapy sessions, we asked Mary to sing short phrases—five to six words—with the melodic line matching the natural contour of the spoken phrase. The rhythm provided a natural timing mechanism for her breathing, and the melody enabled her to lend a more natural sound to the phrase. In a relatively short time, Mary was applying these techniques outside of therapy and speaking longer, clearer phrases and even sentences.

Words Spoken And Sung

There are several cases in which a patient has recovered speech through the systematic use of rhythmic patterning, leading first to recovery of familiar lyrics and words embedded in songs, then to self-initiation of normal, fluent speech.

Because music has parallels to spoken language, much research on music and the brain has zeroed in on the similarities and differences between them. The similarities could be clues to more successful methods of using musical cueing to stimulate similar language responses in people with brain injuries. One remarkable example of the functional difference between music and language, however, occurs in people who have suffered a left-side stroke, resulting in a type of aphasia where verbal comprehension still exists but the ability to speak or find the right words is lost. In these cases, the brain lesion is often located in what is called Broca's area; speech is slow, not fluent, and hesitant, with great difficulties in articulation. Yet, despite the loss of speech, many people with this type of aphasia can sing complete lyrics to familiar songs. This has usually been attributed to the separation of function of the left and right hemispheres of the brain, speech being dominant on the left and singing on the right.

Because many clinicians assume a complete separation of function between singing and speaking, they give little attention to the potential for using music to aid speech. But there are several cases in which a patient has recovered speech through the systematic use of rhythmic patterning, leading first to recovery of familiar lyrics and words embedded in songs, then to self-initiation of normal, fluent speech. In each case, however, this remarkable change had been attributed not to the music but to spontaneous recovery during the early months after the stroke.

A similarity shared by music and speech is what we call "prosody," which includes the elements of stress, pitch direction, pitch height, and intonation contour, or inflection. People with nonfluent aphasia can perform a type of prosodic speech that includes the inflection and contour of previously known phrases. This speech differs, however, from propositional speech (which includes verbal expression of new thoughts and ideas) in its rate, discrete pitch, and increased predictability. Aniruddh D. Patel, Ph.D., a scientist at the Neurosciences Institute in California, theorizes that rhythm and song, which are inherently predictable, may create a "supra-linguistic" structure that helps cue what is coming next in an utterance.

Brain-imaging studies by Dr. Pascal Berlin, of the Service Hospitalier Frederic Joliot in France, and more recently by Dr. Burkhard Maess at the Max Planck Institute of Cognitive Neuroscience, used PET and MEG scans to determine that areas peripheral to the left language regions of the brain are involved in processing the singing of single words. Additional imaging studies suggest that some aspects of music and language are processed in both the right and left sides of the brain. In many patients who are able to carry over speech techniques from music, success seems to come from their increased ability to attend to sounds and to initiate them, perhaps because parallel mechanisms for these functions have been called into play by music and singing.

Sally: Out Of Silence, A Remembered Song

Just as rhythm can affect motor function and the initiation of movement, a familiar tune or melody can reawaken in persons with dementia, or with traumatic brain injury, seemingly lost memories and feelings. We are so much the sum of our experiences and memories that we cannot help associating each new experience with something that came before it. Imagine how the world must seem to someone with no memory link from past to present. But sometimes music can provide a bridge.

Sally had been diagnosed with leucoencephalopathy. She was mute; apart from crying, she made no vocal sounds. She spent her days pacing the long nursing home corridor and crying. Although she seemed to have lost the ability to recognize objects, she navigated well. If she walked into something, including a person, she would touch it and immediately seem to identify its purpose. One day, as I played some tunes to other residents, I was surprised to hear a beautiful voice singing the complete lyrics to the song I was playing. I turned to the door to see Sally dancing and singing her way into the room.

Later, I telephoned her sister and learned that Sally had played the piano; she had loved to entertain at parties, singing many of the songs I had been playing for the residents. Nevertheless, Sally's sister was astonished at what I reported, because Sally had fallen mute long before her illness was fully diagnosed. The nursing home staff began singing to Sally every day; she sang back in a kind of chanting tone. Her crying stopped, as did her restless wandering of the halls. Soon she began speaking and became more integrated into the world of the nursing home.

We do not know specifically how music affects memory, but most of us experience that effect every time we hear a favorite song. Indeed, music is capable of arousing in us deep and significant emotions. Memories of music can be so well preserved that the merest fragment of a melody stimulates recall of the song's title or lyrics. Emotionally charged responses to familiar music are probably the result of connections from the auditory nerve to key limbic structures in the brain. The limbic area, which is associated with emotion, includes the olfactory cortex, amygdala, and hippocampus. The amygdala gets its input from our senses and directly affects our autonomic responses; it is also involved with our moods through interconnections with the frontal cortex and thalamus. The hippocampus plays a significant role in storage of factual information, including conscious (declarative) memory.

For people with neurologic impairments or diseases, music therapy can be an essential first step in recovering functions such as speech or the ability to experience emotion. Scientists are beginning to discover how the elements of music may aid in this process. It is rare for someone to lose all capacity to experience rhythm, harmony, pitch, melody, or other aspects of music.

Because memories persist when they have personal significance, the emotional content of music seems to be processed immediately, even by people with severe dementia. Is this a possible pathway we can use to reach their sense of self? Ernest G. Schachtel said in 1947 that memory, as a function of the living personality, can be understood only as the capacity to organize and reconstruct past experiences and impressions in the service of present needs, fears, and interests. Just as there is no such thing as impersonal perception and impersonal experience, there is no impersonal memory. Thus, familiar songs may serve as cues to recall memories. People with dementia, who may have lost the capacity to process many types of information, including the ability to identify a song, may still respond to that song spontaneously and emotionally. In "Music and the Brain," Oliver Sacks writes that "it is the inner life of music which can still make contact with their inner lives which can awaken the hidden, seemingly extinguished soul; and evoke a wholly personal response of memory, associations, feelings, images, a return of thought and sensibility, an answering identity."

Observing how people with dementia respond to music gives us an inkling of how remarkable and instantaneous some of these subcortical processes are. But if, as pointed out earlier, the brain's processing of music is complex, involving many areas, what specific component of music does a person perceive and process to allow for these immediate responses?

In some instances, factual memories return. New research is shedding light on how this may happen. Ann Blood, Ph.D., Robert Zatorre, Ph.D., and their colleagues at the Montreal Neurological Institute investigated the brain mechanisms involved in emotional responses to music. They found that regions previously identified with pleasant or unpleasant emotional states (with the exception of fear) were activated in the para-limbic brain regions, rather than areas normally associated with music perception. Studies like this reinforce the concept of musical processing as a "whole brain" phenomenon. With the proper musical cue, we may gain access to another system, with enough overlap to jump-start similar areas that are now dysfunctional. That is, when higher cortical processing is compromised, there may be another way into the brain.

Harnessing Music's Power

Perhaps if we understood more about the relationship between the auditory system and other aspects of human cognitive function, we could reach more people like Sam, Mary, and Sally. For those with neurologic impairments and diseases like Parkinson's or multiple sclerosis, music therapy is only beginning to be recognized as a promising treatment. In its "Primer on Reimbursement," the American Music Therapy Association notes that music therapy is recognized as a viable treatment option, including in federal law and by accrediting agencies. It is included in the Older Americans Act Amendments of 1992 and the Individuals with Disabilities Education Act, and recognized by the Rehabilitation Accreditation Commission and the Joint Commission on the Accreditation of Health Care Organizations. Even so, the availability of music therapy for the whole range of situations where it could help is gravely limited.

Although much is being discovered about music's effects on the brain's functioning, we have no cohesive, detailed theory of how this takes place. For example, what specific element of music aids in the recovery of language in a person with aphasia? Is it the articulation and rhythmic cueing of familiar speech patterns? Or does singing the lyrics stimulate and improve word retrieval for normal speech? How, specifically, does music affect retrieval of memories? When stimulated by music, what role do lower brain areas (the cerebellum, reticular formation, and others) have in the upward activation of higher cortical mechanisms?

The great Russian neuropsychologist Alexander Luria observed that what we know of brain function is based on what has been lost and what remains following a traumatic brain injury. Music therapists who do neurologic rehabilitation know that it is almost impossible to lose all aspects of music perception. Knowing how the brain processes the elements of music—rhythm, pitch, harmony, timbre, tempo, contour, loudness, spatial location, and melody— as well as associations and memories, and where overlapping or parallel regions share this processing, could support increased use of these components of music early in treatment, the better to take advantage of brain functions that have been preserved.

With the advent of new imaging techniques, we know that the brain is a dynamic, ever-changing system of interconnecting neurons that work in concert to produce our complex, dynamic responses to the world around us. The discovery that new networks and connections may be formed in the brain every time we learn a new skill has implications not only for early childhood development, but also for potential recovery of function after injury.

I will never forget one patient, admitted for short-term rehabilitation when he was in the early stages of dementia. He no longer could dress himself. He seemed not to have the fine motor skills to button his shirt, yet he could play the opening of the "Hungarian Rhapsody" on the violin. Both skills obviously had been used almost every day throughout this man's life, yet he had lost one and not the other. How can rehabilitation take advantage of such similar but subtly different functions?

It is highly unlikely, for example, that a symphony conductor and a tennis player would have the same motor skills and memories for movement in the left and right hands and arms, yet standard physical and occupational rehabilitation practices would treat them as identical. Conductors, at least the good ones, must be able to give two simultaneous signals that may convey completely different messages—for example, cueing the violins while setting the timing patterns for the percussion section. They will tell you that they can separate the functioning of their left and right sides. In musicians with these overlearned motor skills, certain motor neural networks and overlaying motor areas in the brain may remain intact even after a stroke, and could aid in earlier recovery of function or even development of compensatory mechanisms. But to help, we simply have to know more.

Both basic research and clinical investigations on the underlying brain mechanisms stimulated by different elements of music will continue. It is fairly safe to predict that we will discover that certain elements of music are processed in "primitive" brain regions, including some that are highly resistant to the ravages of traumatic injury and disease. Then we must ask: How do these deeper regions of the silenced brain, reached by rhythms or melodies of music, in turn stimulate the brain's higher regions (or bypass them) so as to switch on motor, cognitive, or emotion-related functions that had appeared lost forever? The answers will come, though no one can predict how rapidly, and then we may see more often— even routinely—what now seems (and is) a miracle: the man struggling to walk will dance; the haunted, weeping woman who walks the halls will rejoin us, singing; and the mind drained of its memories will know the comfort of a familiar old tune

By Concetta Tomaino