Misconceptions about Brains, Part 2: why super-villains waste a lot of money on mind control

A common belief about the human brain is that memories are stored perfectly. People expect each memory to be encoded in neurons as if they were recorded with a camera. In fact, human memory is unreliable.

Humans have three types of memory: short-term, long-term, and functional. Functional memory includes skills, but does not help remember the act of learning the skills. Short-term memory, which lasts around 15 minutes at most, has the most detail. Most information is filtered out and discarded during the transfer to long-term memory.

Since long-term memory lacks so many details, it is mutable. Every time a memory is accessed, it can be altered by filling in missing details incorrectly. A person may fill in gaps by guessing what probably happened. If they are telling a story they may adjust details, not to lie, but to emphasize something other than what they were focusing on at the time.

An example of how memory can be completely incorrect can be found by asking people if they saw clowns one day. Assuming that they did not see clowns, they will answer “no”. If they are asked again a few days later, they will remember something about clowns, but forget the negative (that they did NOT see clowns), and will be tricked into remembering clowns.

Psychologist Elizabeth Loftus called this the misinformation effect. In her research into eyewitness testimony, she discovered that leading questions can trick witnesses into remembering false information. Even open-ended questions which are not usually called leading can change a person’s memory by asking them about details they did not originally notice. Since the human brain excels at pattern-matching, people are more likely to fill in the blanks to finish the pattern than admit, even to themselves, that they didn’t notice something.

Further research by Loftus discovered that repressed memories are not real. People who claim to have such memories of childhood trauma did not uncover the memories with the help of a psychologist, but instead the psychologist planted them by asking leading questions. The psychologists in these cases believe that they are following a lead into the repressed memories, but in fact there is no evidence that the brain has a mechanism to repress memories at all.

Admittedly memory storage is poorly understood, but all evidence so far suggests that most information from human sensory organs are discarded; either it is never stored or it degrades with time. Reversible amnesia occurs when the retrieval mechanism is damaged, but there is no plausible way this can be caused by the trauma associated with the memory (unless it was the memory of receiving a traumatic brain injury). Once a memory itself is gone, it cannot be un-repressed. It is gone.

Another misconception revolving around memory is that each memory is stored in a specific neuron, perhaps in a specific molecule. It is possible that memories are stored in molecules, but neurobiologists currently know more about how memories are not stored than how they are.

First, although storing memories is localized to the hippocampus, after they are stored memories are “located” throughout the brain. This was discovered when epilepsy patients had their hippocampuses removed, resulting in a lack of new memories being stored but not in a loss of existing memories. (It also reduced their seizures, which is why they had had the surgery; technically the operation was successful.) Further studies found other places where memories were stored, but these were many and spread apart.

Second, although the exact medium for storing memories is unknown, the idea of one-to-one correspondence between memory and events has been refuted. As part of the brain’s habit of pattern-matching to fill in blanks, similar memories are frequently combined, resulting in two unrelated events being remembered as a single, complex event, or sometimes in a detail from one event being associated with another.

Third, memories are never stored in a single neuron. This was confirmed by a rather simple discovery: neurons die and regrow regularly, albeit at a much slower rate than other cells (the hippocampus produces 700 new cells per day while stomach cells are all replaced every few days). If memories were contained in individual neurons, long-term memory would last a few years.

In popular culture and science fiction, villains commonly use dangerous surgery or advanced technology to remove incriminating memories in their adversaries. With today’s technology, this would probably kill the patient. Slightly better technology would cause brain damage and general amnesia, since localizing memories is implausible. Doing this successfully would be absurdly difficult and expensive.

On the other hand, anyone can change someone’s memory by asking leading questions.