Does listening to Classical music actually make you smarter? A case study of a misconception
As my social media timelines fill up with birth announcements and pictures of new children, I keep seeing one recurring piece of advice from well-wishing relatives:
“Play some Mozart for them. It’ll make them smarter.”
As an eternal skeptic and neuroscientist, I’ve always been doubtful of the validity of this statement.
Like playing a childhood game of Telephone, each report of a scientific study in media outlets changes, oversimplifies, or misinterprets previous reports, ultimately leading to the birth of a new myth: Listening to Mozart as a baby increases intelligence.
This misconception was rampant across the 90’s, and the infant education-industrial complex was immersed in Mozart mania. Toys started regurgitating phrases from his prolific catalog of lullabies. Books started popping up, encouraging parents to subscribe to the values of the “Mozart Effect”. It even reached the level of government: In 1998, the governor of Georgia Zell Miller requested $105,000 as part of the state budget to provide parents of young children a cassette tape (remember these?) or a CD of classical music to improve intelligence 1.
Popular science myths tend to originate from legitimate scientific studies that have been misinterpreted by the press. After a bit of digging through the primary literature, I found this to be the case here as well.
This particular misconception arose from a short, two-column study published in 1993 in Nature, one of the top journals in the field of science research. 2 Here, researchers from the University of California Irvine gathered 36 undergraduate students and tested their performance on an intelligence task under three different 10-minute conditions: the experimental condition where they listened to a Mozart sonata, a control condition where they listened to a verbal relaxation tape with guided instructions on how to lower blood pressure, or their negative control condition of silence.
After each of the listening conditions, the students were tested using portions of the Stanford-Binet Intelligence Scale. Originally developed in 1916, this particular IQ test had undergone several changes over the decades, and by 1993, was in its Fourth Edition (SB-IV).3 This iteration of the test measures four major cognitive domains:
Sample abstract / visual reasoning multiple choice question
Verbal Reasoning: This domain includes questions that assess knowledge such as the meanings of words, subdivided into vocabulary and comprehension. It also assesses the person’s ability to identify the “absurdity” of a given situation, and to describe the relationship between three provided words.
Abstract / Visual Reasoning: To evaluate these reasoning skills, several different exams can be used. In one of them, a series of patterns were shown, and the patient is expected to predict the following image. In another, a folded piece of paper with cuts is presented, and the subject then has to select the image that shows what the paper will look like when unfolded.
Quantitative Reasoning: Here, the examinee may be given a series of numbers, such as [1, 3, 7, 15] then is asked to produce the following two numbers that completes the pattern [31, 63]. They can also be asked to rearrange a series of numbers (15, 3, 5) and operators (=, x) into a true mathematical equation (3 x 5 = 15).
Short-term Memory: These behaviors were assessed by presenting the patient with a sentence, a string of digits, or a slideshow of various objects, and then asked to repeat back the pattern in the correct order.
The rationale behind the SB-IV was that each domain assesses a slightly different skill set. All of these skills build upon each other to produce a value of “composite” intelligence, which is collective IQ.
In the study, only the abstract / visual reasoning domain of intelligence was tested. While listening to the music, the students answered pattern analysis questions, a multiple-choice matrices test, and paper folding / cutting questions. The researchers then scaled the performance on these tasks up to the composite intelligence, and observed an improvement of IQ by 8-9 points. However, the three other cognitive domains were untested.
Another huge logical leap over the years of misinterpretation is related to age. One of the reasoning tests used in the original Nature article was the paper folding and cutting test, an assessment that is only used for children ages 12 and up. Appropriate for the college aged undergraduates whose brain growth has already started to plateau, but says nothing about the brain prior to those years, and even less about the first couple years of newborn life, where growth happens exponentially.
Also, there was an effect of time: the reported cognitive benefits were only observed in the time window of 10-15 minutes, a relatively short window following exposure to the melodies. It did not follow any of the subjects beyond that original study, and did not address any long term measures of intelligence or development.
(As a meaningful control experiment, they assessed heart rate and found that there were no significant differences between all three conditions. This was done in order to rule out an arousal effect, the observation that physiologically “elevated” states are correlated with increased attention, therefore improving performance on a variety of tasks.)
To accurately summarize the results of the study, it would be best to report that:
College-aged students who listened to Mozart performed better on abstract / visual reasoning tasks for 15 minutes
…but that’s the kind of clunky headline that an editor might boil down to something misleading.
Similar research teams had a hard time trying to reproduce the study’s temporary boost in abstract reasoning skills. Future studies testing the validity of the Mozart Effect found similar, but weak effects. A large-scale meta-analysis of over 40 studies and 3,000 subjects revealed a small correlation (d = 0.37) between Mozart and performance on spatial tasks. 4 Of notable importance, they discovered a similar improvement (d = 0.38) when the subjects listened to non-Mozart musical stimuli compared to silence - suggesting that the improvements might as well be called the “System of a Down Effect” or the “Wu-Tang Clan Effect”.
There are so many things to stress out about as parents of newborns, and you certainly shouldn’t feel guilty if Mozart isn’t in your usual Annual Spotify Recap. In the end, as long as the volume is at a reasonable level, and it doesn’t cause visible distress, playing Classical music for the little ones certainly won’t hurt the baby’s intellectual growth. But if the little one seems happiest and bounces along when Taylor Swift comes on the radio, put that on repeat and dance along with them.