When iTunes first released its shuffle feature received a slew of angry emails. The complaints came from customers who claimed the feature was broken because, when they clicked the “shuffle” button on their *NSYNC album, the tracks sometimes played in order. People felt cheated. How could a random algorithm produce three songs in the same order as they appear on the actual album? Random should mean 7, 11, 3, not 1, 2, 3, right? Except that when song selection is truly random, each song has the same probability of playing each time the current song ends. Sometimes that means 7, 11, 13, but sometimes that means 1, 2, 3 instead.
In response, iTunes changed their randomness algorithm to avoid such sequential ordering. The new algorithm feels more random to us humans despite it being objectively less random.
The most striking examples of the persistence of musical memory come from observation of patients suffering from dementia. Late-stage Alzheimer’s patients who have difficulty recognizing family members and familiar objects can still recognize familiar songs. In some instances, these patients are able to sing despite having lost the ability to speak.”
The unique strength of musical memory has puzzled researchers for years, but one possible reason for the robustness of musically encoded memory is that music is encoded by several different regions of the brain. While auditory regions are primarily involved, so are parts responsible for imagery and emotion. Because musical memories are laid down in multiple brain regions, stimulating any one of these regions may spark their retrieval. It also may be the reason musical memory persists so long in dementia patients. If one brain region becomes damaged, the other, healthy regions can pick up the slack, theoretically providing “backups.”
The taxpayers of New Jersey and Pennsylvania felt the brunt of this firsthand in a real-life experiment in 1992. Both states switched that year to a no-fault insurance regime where consumers could save money by limiting their right to sue for tort damages, but the way the option was framed differed by state: New Jersey made limited right to sue the default option, while Pennsylvanians were presumed to select full right to sue, unless they opted out. This small change in the status quo had a profound effect on their behavior: 75 percent of Pennsylvania consumers paid to retain full tort, while only 20 percent of New Jersey consumers did.
Similar effects have been found with many other kinds of behavior, including student loan repayment and even, interestingly, willingness to donate organs.