Research projects

Learn more about participating in current research in the Auditory Development Lab

The development of sensitivity to musical structure
Body movement shapes how we hear musical rhythms
Encoding melodic information in the brain
The nature and purpose of infant-directed singing
Emotional expression in infant-directed speech
Infants' long-term memory for music
Temporal resolution in the developing auditory system
Related Links

McMaster Institute for Music and the Mind

Measuring brain development using event-related potentials
Auditory what and where pathways
Cortical plasticity as measured by the effects of musical training on the brain
Designing a hearing aid that enchances speech signals in noisy enviroment

1. The development of sensitivity to musical structure

Adults in Western society with no formal musical training have learned a lot about Western musical structure through years of hearing music. For example, they know which notes belong in a musical key and are good at identifying "wrong" notes in a melody when they go outside the key or the implied harmony of the melody. We have shown that 8-month-old infants do not yet hear according to key structures because they are equally good at telling when a melody contains a wrong note, whether it is out of the key, out of the implied harmony, or in both the key and harmony. Five-year-olds hear according to key structures, but it is not normally until about seven years of age that children become sensitive to implied harmony. Although infants are not sensitive to key structure, they are sensitive to many basic musical features. For example, we have shown that as young as 3 months of age infants prefer to listen to consonant chords and music over dissonant chords and music. Melodically, they are better able to encode and process octaves and perfect fifths over more dissonant intervals such as tritones and minor ninths. In sum, there appears to be a progression from early sensitivity to consonance, later sensitivity to key membership, and latest sensitivity to implied harmony. We are currently interested in the effects of musical experience on this developmental progression.

Hear Mozart in the original
Hear Mozart made dissonant

2. Body movement shapes how we hear musical rhythms

We created an ambiguous rhythm ( Sound File 1) that could be interpreted as a march (ONE-two-ONE-two-ONE-two!) if every second beat was accented or as a waltz (ONE-two-three-ONE-two-three!) if every third beat was accented. One group of infants was bounced in the arms of an adult on every second beat and another group on every third beat for two minutes. After this experience, infants were given a preference test to determine how they encoded the ambiguous rhythm. Infants controlled how long they listened to two auditory-alone versions of the rhythm, one with accents on every second beat and one with accents on every third beat (Sound Files 2, 3). In this experimental setup, infants turned on each rhythm by looking at a flashing light and turned it off by looking away. Infants who had been bounced on every second beat preferred to listen to the version of the rhythm that had accents on every second beat, while infants bounced on every third beat preferred to listen to the version that had accents on every third beat. Similar studies in adults have shown that their movement also affects how they encode auditory rhythm patterns. This research demonstrates that music is a multi-sensory experience and suggests that the concurrent experience of listening and moving to a musical rhythm wires the brain so that the different sense work together.

3. Encoding melodic information in the brain

Recent research has shown that from the cochlea in the inner ear through subcortical pathways and into primary auditory cortex, there is a tonotopic organization in that similar frequencies are encoded in adjacent areas. However, in most situations it is relative pitch (the distance between two tones) rather than absolute pitch (the actual frequencies of the tones) that is important. In other words, a familiar tune such as "Happy Birthday" is recognizable regardless of whether it is rendered in a high or low pitch range, as long as the pitch distances between notes are correct. We are investigating how and where in the brain relative pitch information is extracted. Using the mismatch negativity component of the event-related potential derived from EEG recordings, we have shown that both relative pitch and pitch contour (up/down) information is encoded automatically in auditory cortex. We have shown that this is true even in those without musical training, although the brains of musicians show larger responses to pitch changes. Recently, we have shown that when listening to polyphonic music with two simultaneous melody lines, separate memory traces are made in cortex for each melody.

4. The nature and purpose of infant-directed singing

Across cultures, caregivers sing to infants, and most cultures have special songs for infants. We are asking why people sing to infants and in what style(s) they sing to infants, as well as what infants perceive when they hear singing, and how they react to it. Studies to date show that caregivers sing in a higher pitch, at a slower tempo, with a more loving tone of voice, and that they put longer pauses between phrases than they do when they are not singing to infants. In turn, infants prefer infant-directed singing over non-infant-directed singing, and like the higher-pitch and loving tone of infant-directed singing. In addition, caregivers singing in two distinct styles to infants, one a lullaby style, which serves the function of calming infants and putting them to sleep, and the other a playsong style, which serves the function of rousing infants and directing their attention to interesting things in their environment. Infants show positive reactions to infant-directed singing, and parents appear to use it to help regulate their infants' state.

Hear a mother singing to her infant
Hear the same mother singing when her infant is not there

5. Emotional expression in infant-directed speech

There is much research documenting the differences between speech directed to infants and speech directed to adults. Infant-directed speech is slower, higher in pitch, and has slow, exaggerated pitch contours. What is the function of infant-directed speech? It certainly attracts infants' attention, and it highlights important words in the input when the infant is learning language. Infant-directed speech is also highly emotional compared to typical adult-directed speech. We have shown that when infant-directed speech is compared to emotional adult-directed speech, the acoustic differences between the two are minimal. In both cases, love and comfort are conveyed with slow, low-pitch and falling pitch contours; surprise is conveyed with large up/down pitch contours; and fear is fast with a steady, fairly high pitch. Thus, much of the difference between infant-directed and adult-directed speech may reflect the fact that we freely express emotion with infants whereas we are normally much more constrained in our emotional expression with other adults. In our most recent work, we are investigating whether parents' use of infant-directed speech is innate, or whether they use it because they have learned that it leads to a positive response from their infant.

Hear examples of emotional speech:

6. Infants' long-term memory for music

Infants can remember music for long periods of time. We play a piece for infants at home for 3 minutes a day for 7 days. On the 8th day, we find that they prefer to listen to a novel piece that they have never heard before in comparison to the familiar piece we played for them. However a couple of weeks later, they prefer to listen to the familiar piece in comparison to the novel piece. This pattern of preference change is similar to that of adults. Interestingly, if we change tempo (i.e., play the piece a little faster or slow than they heard at home) during the preference test, the infants behave as if they had never heard the piece before. Similarly, if we change the instrument on which the piece is played from a piano to a harp or vise versa, again the infants behave as if they had never heard the piece before. This tells us that infants are remembering the specific tempo and timbre of the piece. However, if we change the pitch range (i.e., play the piece half a octave higher or lower than they heard at home), infants still prefer to listen to the novel piece, indicating that they forget or ignore the absolute pitch, and focus on the relative pitch information.

7. Temporal resolution in the developing auditory system

Successful discrimination of speech sounds depends on good temporal resolution in the auditory system, and deficits in this area have been linked to language-learning problems. Because the brain is most plastic and able to learn and change early in life, it is important to diagnose temporal processing problems in infancy, when infants are just beginning to learn language. We are thus developing measures of temporal resolution that can be used with infants. In particular, we have shown that the mismatch negativity component of the event-related potential can be measured in 6-month-old infants, and that it can be used as an index of the ability to discriminate sounds. Using this measure, we have determined that basic temporal resolution is similar in normal 6-month-olds and adults in a gap detection task (using short Gaussian-enveloped 2000 Hz tone-pip markers to define the gap). However, infants do much worse than adults when the markers defining the to-be-detected gap differ in frequency, and are therefore processed in different frequency channels, necessitating a comparison of timing across channels at a more central stage of processing. This between-channel gap task more closely reflects the temporal processing needed to decode speech, and is therefore a good candidate for a diagnostic test.

2000 Hz tone pip with no gap
2000 Hz tone pip with 8 ms gap

8. Measuring brain development using event-related potentials

In adults, occasional changes in a repeated sound or sound pattern triggers the response of a change-detection mechanism in auditory cortex. The response is seen as the mismatch negativity component in the event-related potential. In contrast to adults, young infants mainly show slow positive brain waves. We have investigated infant brain responses to changes in pitch, location, and timing changes. In general, in response to sound change, younger infants show an increase in the slow positive wave and older infants show a faster, earlier negative increase, resembling the mismatch negativity. However, these components overlap, and the negativity emerges around 2 months for pitch changes, around 4 to 6 months for temporal changes, and after 8 months for location changes. Thus, it appears that the change-detection mechanisms in auditory cortex for each of these sound features mature at different rates.

9. Auditory what and where pathways

In the octave illusion, discovered by Diana Deutsch, two sine wave tones an octave apart alternate in an extended pattern such that the input to the two ears is out of phase (i.e., when the right ear hears the high tone, the left ear hears the low tone). When presented with this stimulus, most people report hearing alternating high and low tones, but with the high tones always coming from the right ear and the low tones always from the left ear. Thus the dominant right ear appears to determine what people hear, but where the tones appear to be localized depends on which ear is getting the higher tone at that time. We have shown, however, that the what pathway does not simple discard the information from the non-dominant ear because the octave illusion stimulus is clearly distinguishable from an "illusion-consistent" stimulus mimicking the illusory percept, in which sine wave tones simply alternate between the ears. As well, the illusion is highly unstable for many listeners, with the ear hearing the higher tones changing from right to left fairly frequently. We have also discovered that the octave illusion stimulus gives rise to an intensity illusion: when the illusion-consistent stimulus follows the illusory stimulus it sounds substantially louder than the illusory stimulus, but when the illusory stimulus follows to illusion-consistent stimulus they sound equally loud. Thus, how the incoming information from the two ears is used by the what pathway is complex.

10. Cortical plasticity as measured by the effects of musical training on the brain

The brains of adult musicians process musical sounds differently than do those of non-musicians. In particular, we have found that the P2 component of the event-related potential is larger for musical tones in musicians than it is in nonmusicians. However, it is unclear as to whether musicians inherited these brain differences and are naturally talented, or whether they are the result of musical training. To examine the effects of musical training on the brain, we are taking EEG and MEG measures of children as they listen to musical sounds. First, we have shown that auditory cortex undergoes a long maturational processing, with N1 and P2 components increasing from age 4 to 12 and diminishing thereafter, reaching adult levels by late adolescence. Second, we have shown that N1 and P2 are already larger in 4- and 5-year-old children who are studying music, and that these effects are largest for sound from the instrument of training (piano or violin). Thus, musical training is affecting brain development early in young children.

11. Designing a hearing aid that enhances speech signals in noisy environments

For many people, hearing aids do not help them understand speech in a noisy environment such as at a cocktail party, but current tests of hearing typically do not predict how well a person will do in real world environments. We are currently working on algorithms that could enhance speech signals in noisy environments. We have developed a Realistic Hearing in Noise Testing Environment (R-HINT-E) in which noisy real world environments can be simulated and hearing tested in a rigorous yet flexible manner.

McMaster University | Faculty of Science | Department of Psychology, Neuroscience & Behaviour

Dr. Laurel J. Trainor, Professor

Research projects