We do not know that brains remember anything; we simply know that people remember things. The credibility of claims that memory recollections come from brains is inversely proportional to the speed and capacity and reliability at which things can be learned and recalled, and the length of time things can be remembered. There has never been found in the brain any component known to be capable of a fast storage or retrieval of learned information, or the storage or retrieval of learned information at any speed. The protein molecules that brains are built from have a high rate of molecular turnover, and have an average lifetime of less than two weeks. There are numerous signal slowing factors in the brain, such as the relatively slow speed of dendrites, and the cumulative effect of synaptic delays in which signals have to travel over relatively slow chemical synapses (by far the most common type of synapse in the brain). As explained in my post here, such physical factors should cause brain signals to move at a typical speed very many times slower than the often cited figure of 100 meters per second: a sluggish "snail's pace" speed of only about a centimeter per second (about half an inch per second).
Ordinary everyday evidence of very fast learning and instant recall is therefore evidence against claims that learning, memorization and memory recall occurs because of brain activity, particularly because the brain is totally lacking in the things humans add to constructed objects to allow fast recall (things such as sorting and addressing and indexes). Attached to unstable dendritic spines that do not last for years, chemical synapses in the brain do not even reliably transmit signals. Scientific papers say that each time a signal is transmitted across a chemical synapse, it is transmitted with a reliability of 50% or less. (A paper states, "Several recent studies have documented the unreliability of central nervous system synapses: typically, a postsynaptic response is produced less than half of the time when a presynaptic nerve impulse arrives at a synapse." Another scientific paper says, "In the cortex, individual synapses seem to be extremely unreliable: the probability of transmitter release in response to a single action potential can be as low as 0.1 or lower.") So the more evidence we have of very fast and very accurate and very capacious recall (what a computer expert might call high-speed high-throughput retrieval), and the more evidence we have of very fast and very accurate and very voluminous learning and memorization, the stronger is the evidence against the claim that memory recall and memory formation occur from brain activity.
It is therefore very important to collect and very carefully study and ponder all cases of exceptional human memory performance. The more such cases we find, and the more dramatic such cases are, the stronger is the case against the claim that memory is a neural phenomenon. Or to put it another way, the credibility of claims that memory is a brain phenomenon is inversely proportional to the speed and reliability and throughput of the best cases of human memory performance. The more cases that can be found of humans that seem to recall too quickly for a noisy address-free brain to do ever do, and the more cases of humans that seem to recall too well for a noisy, index-free, signal-mangling brain to ever do, and the more cases of humans that seem to learn at a speed, reliability and capacity too fast and accurate for a noisy, relatively slow and transmission-unreliable brain to do, the stronger is the case that memory is not a neural phenomenon but instead a spiritual or psychic or metaphysical phenomenon.
A case of exceptional memory performance is well documented in the book "The mind of a mnemonist : a little book about a vast memory" by Alexander Romanovitch Luria. You can read the book on www.archive.org using the link here. The book describes a subject it calls S, a subject now known to be Solomon Shereshevsky. A scientific paper says this about the book:
"According to Luria, Shereshevsky could 'easily remember any number of words and digits' and 'equally easily he memorizes whole pages from books on any subject and in any language.' He could accurately quote information from a decade earlier, including tables of numbers and strings of nonsense words....What Luria learned was that Shereshevsky’s memory differed from that of the vast majority of individuals; time did not erode his memories. Neither did a new stimulus affect his memory of an earlier one."
On page 17 of Luria' s book we are told that when asked to memorize the table of random numbers below, Shereshevsky reproduced the table perfectly, in 40 seconds, after only three minutes of studying it.
On page 20 we are told that after Shereshevsky viewed a stimulus a certain length of time, he could continue to see the thing vividly in his mind's eye. We read, "He told us that he continued to see the table which had been written on a blackboard or a sheet of paper, that he merely had to 'read it off,' successfully enumerating the numbers or letters it contained." Around page 23 we are told that Shereshevsky had to some degree synesthesia. When tested with electronic sounds, he reported that they would produce some type of visual stimulus. On page 24 we are told, "Every sound he heard immediately produced an experience of light and color and, as we shall see later in this account, a sense of taste and touch as well."
On page 51 we are told that in 1936 Shereshevsky was asked to memorize a hard-to-remember table that started out like this, and had at least two additional rows like the rows below.
- "The best competitors manage to memorize a one-hundred-digit number in under 15 seconds."
- "The record for memorizing the sequence of cards in shuffled deck of 52 playing cards was two minutes in 1993 and is now 12.74 seconds by Shijir-Erdene Bat-Enkh set in 2018."
- "Nowadays, tens of thousands of digits are being recalled in π competitions. Modern records include memorizing 35 packs of cards or an over 3000-digit decimal number in one hour, a sequence of 145 words in 5 min, the names of a class of 30 students in under 30 s, a shuffled deck of 52 playing cards in around 12 s or a 50-digit binary number in one second."
- "But didn’t Daniel Tammet memorize 22514 digits of π? Recently, Susanne Hippauf, a policewoman, broke the German record with 18026 digits."
- "For example, in auditive digits, the objective is to memorize a sequence of decimal digits, read aloud at one digit/s [1 digit per second]. The score is the number of digits counted until the first error. The current world record of Lance Tschirhart is 456, over seven minutes of perfect memorization."
- "We tried to estimate the memorization time of the π-record using the function R(t). The π world record of Suresh Kamar Sharma stands at an impressive 70030 digits, a multiple of around 20.5 of the hour world record of 3412."
- "The speed of historic dates [is] the fastest among the five-minute disciplines (Figure 1). The objective is to memorize the year between 1000 and 2100 of fictional events (Appendix B). After memorization, the events are scrambled and competitors recall the respective years of the events. The current world record is 148 by Prateek Yadav." This presumably involved the correct memorization of the years of 148 fictional events, all within a five minute period.
