Michael & Ellen

Music is a significant facet of human culture, yet little is known about its neuronal evolutionary origins. Rodent models are heavily employed in translational neuroscience research including cognitive and auditory disorders. However, it is unknown whether mice possess any neuronal auditory precursors to human music preference. For our course project, we investigated music preference in mice by exposing them to a range of artificial, natural, and musical sound clips. All mice showed evidence they were aware that their location controlled the sound output of the speakers. Individual choices varied widely, but some mice showed modest preferences for particular songs. Our future directions include observing mouse music choice in their home cage over longer periods of time.

We used a high-speed camera in conjunction with Bonsai software to monitor the position of individual mice in a large arena, which in turn controlled the sound output in a closed-loop circuit. Our experiments evolved over 4 testing days. The arena was placed on a white opaque base and was bottom lit with an IR lamp. The camera and speaker was mounted 60 cm above the cage. All experiments were performed in a reverberant white painted room with bright fluorescent illumination. We used four wild type C57BL/6 mice housed as sibling pairs in clear plastic shoebox-style cages with ad libitum water and standard kibble. Females: 60 days old, 19 g, 4 recording days; males: 70 days old, 23 g, 2 recording days.

Day 1) Two female sisters were individually placed into a large rectangular cage (standard rat cage, 34 x 38 cm) and allowed to acclimate to the arena for 5 minutes. They were then presented with a range of pure tones continuously played in two sessions (15 and 30 min) . The centroid of the mouse was continuously monitored with the x-coordinate controlling tone frequency (0.5 to 18 kHz) and the y-coordinate controlling amplitude (30 to 80 dB). Due to the variability in the precise location of the centroid, the sound had a warbling melodious quality. One mouse appeared to prefer the quiet low-frequency corner, but the other mouse showed high activity with no preference, running continuously stopping mainly in the corners.

Day 2) Two female sisters were individually placed into the same rectangular cage as Day 1. They both acclimated to arena for 12 minutes and were then presented with the same sounds as Day 1 for 35-37 minutes. Our results show that they appeared to be stressed and lacked any preference for the pure tones and preferred the quiet low frequency corner if at all.

Following our mid-point presentation we divided the arena into five discrete sound regions. We played music along with some artificial sounds as well as silence. Based on their location, the mice controlled whether they heard one of three genera specific songs (e.g., classical, electronic, or pop music), an artificial sound (e.g., beeps, buzzing, or pure tones), or silence. Thus, they could choose to listen to a choice of three songs, an artificial sound, or sit in silence.

Day 3) Two male brothers and two sisters were individually placed into the same rectangular cage as Day 1. We recorded two different sessions of music choices, each split into two counter-balanced runs. The mice were acclimated to the arena for 3 minutes with no sound and then sounds were presented for 10-15 minutes. To control for location preference we physically rotated the cage 180 degrees and rearranged the music location. Thus, we recorded the mice in four 10-15 minute runs with 20-30 minutes total in two different soundscapes.

Day 4) Two male brothers and two sisters listened to six songs overnight that were to be used in the experiments on the following day. We reasoned that familiarity of the songs might encourage them to choose what to listen to. During recording sessions, they were individually placed into a circular arena 32 cm in diameter. We repeated the recording strategy as on Day 3. The lack of corners removed corner preferences and the mice appeared to lose physical place preference. However, we still counter-balanced the physical location of the round arena by rotating it 180 degrees between runs and re-arranged the sound locations. We found considerable individual variation in sound choice, but observed some general trends. First, we found that the mice spent more time in the center of the round arena when they were in control of the sound as compared to when there was no sound feedback, indicating that they had successfully learned to override their natural edge preferences in order to control the sounds being played. Shown here is a representative heatmap and histogram displaying one animal’s time spent in each region of the arena. Analysis is still ongoing, but so far we have noted a slight preference for the electronic music (Lemonade, by Boys Noize), possibly due to its high-frequency components.


We investigated whether mice will select sounds based on their location in an arena. We found considerable variation of choices between mice, but some showed a clear preference for electronic and rock music. This could be due to the greater representation of high frequency sounds than in chanting. Also, while we attempted to equalize the songs for overall volume, some were qualitatively louder than others. Perhaps they chose the electronic and rock music because they could hear it much easier than the classical and chanting music. Our analysis continues and we will update the results when we can.