Pitch-Brightness Correspondence in Four-month-old Infants

Dublin Core


Pitch-Brightness Correspondence in Four-month-old Infants


Hannah Wilson




Adults, children, and infants as young as 10-months have been shown to appreciate a correspondence between auditory pitch and visual brightness, with most participants associating high-pitch with brighter stimuli and low-pitch with darker stimuli. Research across ages is vital for understanding more about the developmental trajectory of crossmodal correspondences. The present study used preferential looking to examine the sensitivity of 4-month-old infants to the pitch-brightness correspondence. Following Mondloch and Maurer (2004), infants were presented with a display of two balls bouncing simultaneously. One ball had a dark surface-brightness, whilst the other had a brighter surface. A single, high or low-pitch sound accompanied the bounce of both balls onto the surface. The research examined whether infants looked differentially to the ball which adults would generally classify as matching. Infants did not look significantly longer to the ball with the congruent pitch-brightness matching. Infants did however look preferentially towards the black ball across trials. It is proposed that this could be the result of the brightness-weight correspondence, whereby darker objects are thought of as heavier than brighter objects. It is therefore possible that infants look longer towards the black ball as it is the heavier ball which should produce the sound, regardless of pitch.


Animations were displayed on a 49cm x 39cm screen, surrounded by black card and screens to block-out excess light and other distractions. Initially, infants were seated (legs stretched forwards) on their caregiver’s lap for the experiment. In this position infants tended to lean forwards, looking towards the floor, meaning that the amount of codeable looking-time was very limited; this procedure was therefore abandoned after four infants had been tested and these infants were excluded from the sample. To increase looking towards the display new approaches were taken. Nine infants were seated 80cm from the screen in a supportive, from-birth highchair. Fourteen infants were seated in an alternative position on the caregiver/researcher’s lap. Infants were now held in a relatively upright position to reduce the likelihood that infants would look towards the floor or their feet. How infants were seated was decided by considering: infants’ head-support, familiarity with high-chair, and infant reactions. It was not thought that being in the high-chair or on the lap would significantly affect looking behaviour.
All animations involved variants of balls bouncing on a horizontal surface. All balls had a diameter of 4cm and were identical to one another aside from their surface brightness. The brightness of balls was measured in candela per square meter (cd/m²) using a lux-meter. A higher cd/m² reading equates to a higher degree of luminosity/brightness. The balls appeared to bounce on a medium brightness (39.52cd/m²), green surface, which had a 49cm diameter. The background of displays was a diffused cloud image which was also medium brightness (57.06cd/m²). The image was blurred slightly to reduce contrast and make the background less visually interesting. It was important that the surface and background had a medium brightness and were not distracting, so as not to alter the effect of ball brightness. Previously, white dots on a black background have been used to portray depth; alternatively this experiment used the clouds and converging-line surface to depict 3-Dimensions.
In the familiarization phase there was one ball of standard-brightness (53.86cd/m²) with a diameter of 4cm. This ball moved up and down along a 25cm vertical axis, stopping for 0.05s at the bottom and 0.1s at the top. The up-down motion of the ball gave the impression that it was bouncing on the surface. The bounce was accompanied by a sound of standard-pitch (782Hz) which lasted for 0.25s. The familiarization trial was presented to indicate to infants that a single ball produces a single noise when it hits a surface. This was important for the test trials as we did not want infants to perceive that the balls in unity were producing the sound. The standard-brightness and standard-pitch of this ball could also be used as reference-points for the brightness and pitch in test trials.
The test animations consisted of three balls, each with a diameter of 4cm. The standard-brightness ball (53.86cd/m²) from the familiarization phase remained stationary at the centre of the surface as a reference point. Alongside the standard ball there were two test balls. Both balls differed only in terms of their surface-brightness. One of the balls had a duller, black surface-brightness (6.3cd/m²) and the other had a brighter, white surface-brightness (144.25cd/m²). Two, independent, vertical trajectories of 25cm formed the path of movement for each ball. Similar to the standard animation, the balls bounced on the surface in synchrony, to either side of the stationary ball (see Figure 1); stopping for 0.05s at the bottom and 0.1s at the top. As the balls hit the surface, a higher-pitch (2096Hz) or a lower-pitch (228Hz) sound was produced which lasted for 0.25s.
Figure 1. These figures are screenshots of the animations seen by infants. The two shots display how the two balls moved in synchrony along vertical trajectories.
The pitch of a sound can affect its perceived loudness. Controlling for sound loudness was crucial because of the correspondence between loudness and brightness (Marks, 1989). To ensure that loudness was not responsible for the effect, the perceived loudness of sounds were equalized. dBA (A-weighted decibels) is a measure of relative intensity perceived by the human ear, weighted for frequency (Plack, 2013). To ensure that perceived loudness did not vary, it was important that sound dBA was approximately the same. A sound-level meter (placed where the infant would sit) was used to measure the dBA of each sound. When sounds were created in Audacity, they were produced with equivalent amplitude. However when the sounds were played, the medium-pitch (782Hz) sound produced a higher 77.2dBA , compared to 72dBA and 71.3dBA for the high (2096Hz) and low-pitch (228Hz) sounds respectively. This meant that the medium-pitch sound would be perceived louder. To compensate for this, a 6dB gain was added to the high and low-pitch sounds. Therefore the relative loudness of the sounds had a much smaller range of values: 76dBA (high-pitch), 76.6dBA (low-pitch), and 77.2dBA (medium-pitch). Although the dB of two sounds needed to be increased, the frequencies of all sounds are within the normal hearing range of 20Hz to 20kHz (Plack, 2013). To confirm that these tones sounded psychologically equivalent, four adults listened to the sounds and were asked ‘Does any tone sound louder than any other tone?’ All participants reported that sounds had equivalent volumes.
The dependent variables in this study were looking-time and number of fixations to the white and black balls. These variables were measured to examine whether infants look preferentially to congruent/incongruent pitch-brightness displays to determine whether they appreciate the pitch-brightness correspondence. To examine this, a 4 (trial) x 2 (pitch: low vs. high) x 2 (brightness: black vs. white) x 4 (condition) mixed ANOVA design was used with 3 within factors (trial, pitch, brightness) and 1 between factor (condition).
There were four distinct trials that were all seen twice by each infant, producing eight test trials. In one trial the black ball was on the left of the screen and a high-pitch sound was heard. In another high-pitch trial, the black ball was on the right. In one trial the black ball was on the left and a low-pitch sound was heard. In another low-pitch trial, the black ball was on the right. It was important that infants saw an equal number of trials with each ball on each side as this ensured that preference for looking towards one side did not affect the results. Each infant was randomly allocated to one of four conditions. All conditions contained the same displays, the conditions varied only in terms of order.
The parent and infant were greeted by the experimenter, told the aims of the study and given the opportunity to ask any questions. Parents were instructed not to point towards the screen to avoid influencing looking. Informed consent was then obtained from all parents. Once the introduction was complete, the infant was seated and the study began.
Firstly, infants were shown the standard animation which consisted of a single, standard-brightness ball bouncing with a standard-pitch tone. This display lasted for a maximum duration of 120s, however the trial ended once 20s of looking towards the screen had been accumulated. This time was pre-defined to ensure that all infants saw the initial display for the same duration. Once the accumulated looking-time was reached, an attention getter (auditory and visual rattle) was presented to re-direct the infants’ attention towards the screen.
The test trials begun when the observer used a computer key-press to indicate that the infants attention had been retrieved. The two balls of the test trials were accompanied by a high or low-pitch sound upon bounce. Each display was presented for a maximum duration of 60s, however the trial ended if the infant looked away from the screen for 2s or longer. After each test trial, an attention getter was presented until the infant looked again. The next test trial was presented when the observer indicated that the infant regained attention. Eight test trials were presented meaning the test trials lasted a maximum of 8 minutes.
Looking-times were coded live using the updated version of Habit2000 software (Cohen, Atkinson & Chaput, 2000). Each session was recorded on camera so that a proportion of infant data could be re-coded by a second observer. This allowed measurement of inter-rater reliability.
Once the infant had completed the experiment, the parent was thanked for their time and given a book for their infant. They were also given a debrief and reminded of their right to withdraw their infant’s data.


Lancaster University




John Towse











Gavin Bremner
Peter Walker

Project Level



Cognitive Psychology
Developmental Psychology

Sample Size

Twenty-three, 4-month-old infants (12 girls and 11 boys; mean age = 123 days, range: 109 to 142 days) comprised the final sample in this experiment. All infants were healthy when they participated in the study. An additional six infants (4 boys and 2 girls) completed the experiment but were unable to be included in the sample because of lack of interest or distraction.
Experiment 2:Ten, 4-month-old infants (6 girls and 4 boys; mean age = 121 days, range: 109 to 140 days) were included in this sample.

Statistical Analysis Type





Hannah Wilson, “Pitch-Brightness Correspondence in Four-month-old Infants,” LUSTRE, accessed April 19, 2024, https://www.johnntowse.com/LUSTRE/items/show/30.