-
Asked by hyperohan to Yee Whye, Valerie, Nick, Mackenzie, Lin, Jun, Jacob, Brian, Anna on 9 Nov 2019.
Question: How Do We Store And Retrieve Memories?
- Keywords:
-
Jacob Leygonie answered on 9 Nov 2019: last edited 9 Nov 2019 00:50
This is a question I am not sure to understand fully, because “memory” could either be related to the the one stored in the brain, or to the “memory” of computers. I will assume we are discussing the first one (I apologise if I did not understand the question correctly).
I am not a neuro scientist, therefore I cannot pronounce myself too adamantly on that question.The (very sketchy) picture I have in mind goes as follows:
We have nodes in the brain called “neurons” joined by edges called “dendrites”. Each time you learn a new concept, you create a new neuron (or a certain number of dedicated neurons). You create new connections (dendrites) between this neutron and the related notions. For instance if you learn words in a new langage, your brain might create neurons associated to new words and link these neurons corresponding to words in your native langage.
If a new piece of information/ event was striking, maybe you will create strong and many dendrites. If on the other hand, if you don’t really care about such piece of information, it will create only weak connections, which will eventually vanish in a near future. It is probably important to solicit several times neurons and dendrites several times to make them resilient. This corresponds to the fact that learning a few times the same thing makes it possible to actually remember it.
A curious phenomenon I happened to realise a few times is the resurgence of lost memories. Let me explain. Sometimes, a specific event happens and raises a memory that you had completely forgotten about. For instance you listen to a specific music, that reminds you of some specific day or person far in the past. Without this event, maybe this piece of memory would never had reappeared on the surface of your conscience. I find this compelling. It might be that this event excites some neurons and dendrites that are close to neurons that are not solicited a lot, but are re activated thanks to their neighbors.
I would like to re insist on the fact that the above opinion is really not fully scientifically supported, it is only my way of understanding a few phenomena related to memory. I also believe that memory remains a big secret in neuroscience and that we are still far from understanding it completely.
-
Yee Whye Teh answered on 9 Nov 2019:
I’m afraid that’s one of the big open questions in neuroscience, and nobody really knows how we store and retrieve memories.
Neuroscientists know a little bit though. For example, that the brain stores information (and memories) using synapses, which are the interfaces between neurons (brain cells). Each synapse has a strength, and the brain stores information by varying the strengths of synapses. A memory (say of your lunch yesterday) has lots of information (what did you actually eat, how did it taste, who you ate your lunch with, what were you thinking etc etc), and these are stored in the configuration of strengths of many many synapses. What neuroscientists don’t really understand is how different memories correspond to different configurations of synaptic strengths and how these are processed by the brain.
Neuroscientists trying to figure out how memory gets represented by synaptic strengths is a bit like trying to figure out an unknown foreign language. Imagine an archaeologist finds a tomb of some queen of some unknown civilisation many years ago. On the tomb are carved little squiggles which look like words. The archeaologist doesn’t know what language the words are in but thinks that it must be about the life of the queen. So the archeaologist wants to decipher the and understand what the squiggles are saying. That’s a bit like how the neuroscientist wants to figure out how the strengths of synapses end up representing the memory of your lunch yesterday.
Comments
hyperohan commented on :
I think that a memory is stored like a jigsaw piece in a jigsaw puzzle, with each neuron holding a piece in a place in the brain called a hippocampus. So, when we try to retrieve memories, the neurons with the right pieces come together to make one memory. This also means that if one neuron holding a piece is changed, or damaged during a concussion or something, the memory can no longer be retrieved, or if it is, then you remember is differently. 🙂 Do you think this is possible answer?