How does the brain recall memories?

[platosuniverse]

I'm working on a project and I've been doing research about memory recall and I haven't found a lot of info on it. I've found info on how memories are formed. A study talks about how memories form in different parts of the brain and play back like a streaming video.

In a new study from University College London (UCL), neuroscientists discovered that when someone tries to remember a singular aspect of an event from his or her past—such as a recent birthday party—that a complete representation of the entire scene is reactivated in the brain like pieces of a jigsaw puzzle coming together to create a vivid recollection.

The July 2015 study, “Evidence for Holistic Episodic Recollection via Hippocampal Pattern Completion(link is external),” was published in NatureCommunications. This research is the first to provide evidence for a pattern completion process in the human hippocampus, as it relates to the everyday experience of recalling previous life events and old memories.

In the new study, researchers were able to show how the hippocampus binds together the diverse elements from an event to form a singular and holistic memory. During memory recall, the brain recalls an old memory by piecing together various compenents via a pattern that forms a cohesive remembrance of things past.

https://www.psychologytoday.com/us/...01507/the-neuroscience-recalling-old-memories

They had people think about Obama in the kitchen with a hammer and found out each of these things are represented in the brain in 3 different areas. An area for Obama, an area for the hammer and an area for the kitchen.

What I can't find information on is how does the recall process get started. For instance, if I want to recall a memory from the Army, how does the brain know which memory I want to recall or does it know?

How does it differentiate between a memory from boot camp, a memory of repelling from helicopters or AIT graduation?

 

[Laroxe]

I suppose the real answer is that no one is sure, but there are some findings worth considering. The first and perhaps most important is that recall can't really be thought of as a replay of an experience, it doesn't work like that. What we remember is the features of an experience we attend to and the value we give to these features, if it is to be stored long term we tend to summarize these experiences even further. Physically we know that memories are stored across the whole brain and at least some of them are associated with other information from the same sensory source. This means that visual elements are stored with other visual memories. There appear to be brain structures that are responsible for organizing this storage, this involves hierarchies and networks of associations. While the structures can generally be located, the hippocampus being the most well known but other areas like the amygdala are also involved but its difficult to link most specific memories to anatomy. We also know that memory storage seems to have considerable redundancy. Recall involves activating one of these networks which then creates links to the things associated, we then create the memory out of the parts actually remembered with the gaps in in with more general expectations.
To remember something we often specifically use associations like picturing a person to remember their number and memory improvement techniques use these principles.
Generally every time we recall something we recreate it and our emotional state both at the time of the memory and the point of recall is important in what is recreated.
This is a set of general issues in memory which are not particularly controversial but there is new stuff coming out all the time. While psychologists don't really understand it they are quite good at messing with it, it appears pretty easy to get people to remember events that didn't happen or change specific detail of events.

I think the study you talk about is pretty typical of much of neuroscience in that it demonstrates a total lack of awareness of the research in other areas, no, 3 individual items will not be stored in 3 different areas, 30 might be a better guess but even this would need qualifying. Despite using huge multi million pound machines there are huge problems in trying to understand function, locating where a specific memory is stored is phrenology, the tech, while impressive is not fast enough or detailed enough to accurately reflect function and data analysis is problematic. You will find psychology research far more useful.

 

[Fervent Freyja]

For your project, the work and concepts relating to memory recall are scattered throughout numerous subfields in psychology and neuroscience. Hundreds, many outdated. I have found most of the work has been compartmentalized (hard to find) and the ones I came across lacked satisfying descriptions on memory recall. I advise you to try and stay away from explanations using a psychology as it's primary standpoint.

This project will be tedious, you will have to hunt down the different explanations independently. Start with your newer textbooks/ignore the many outdated concepts, then check recently published research to get an idea of how we may be able to one day explain memory recall in one concept alone.

 

[atyy]

https://www.psychologytoday.com/us/...01507/the-neuroscience-recalling-old-memories

They had people think about Obama in the kitchen with a hammer and found out each of these things are represented in the brain in 3 different areas. An area for Obama, an area for the hammer and an area for the kitchen.

What I can't find information on is how does the recall process get started. For instance, if I want to recall a memory from the Army, how does the brain know which memory I want to recall or does it know?

How does it differentiate between a memory from boot camp, a memory of repelling from helicopters or AIT graduation?

The idea of "pattern completion" is that given a fragment of the thing to remembered, the brain is able to recall the entire thing to be remembered. The most common idea for how this is implemented is that there should be something like attractors in the brain, with each attractor corresponding to a different memory. Starting from different points (different fragments that initiate recall), the state trajectory of the system is pulled towards a different attractor. The points which lead to each attractor is called the attractor's basin of attraction. A toy model of this process is given by the Hopfield network.

Why one fragment does not accidentally activate the wrong memory is the problem of "pattern separation".

https://en.wikipedia.org/wiki/Hopfield_network
http://www.scholarpedia.org/article/Hopfield_network
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3812781/

 

[aheight]

The idea of "pattern completion" is that given a fragment of the thing to remembered, the brain is able to recall the entire thing to be remembered. The most common idea for how this is implemented is that there should be something like attractors in the brain, with each attractor corresponding to a different memory. Starting from different points (different fragments that initiate recall), the state trajectory of the system is pulled towards a different attractor. The points which lead to each attractor is called the attractor's basin of attraction. A toy model of this process is given by the Hopfield network.

Neural assemblies in the brain are of course non-linear and recurrent so one would expect to find the trappings of non-linear dynamics including strange attractors, their basins, sensitive dependence and catastrophe in the brain's dynamics. A simple google search of strange attractors in the brain gives for example: https://link.springer.com/chapter/10.1007/978-3-642-70795-7_8

A nice practical hands-on approach to understanding attractors and their basins would be the Lorenz Attractor, that owl-eyed icon of Chaos Theory. It is unfortunately, a coupled system of non-linear differential equations but one can numerically solve them easily in Mathematica. Now, start with a grid of points surrounding the attractor and using each of these points as initial conditions, solve the Lorenz equations. Some points will cause the solution to rapidly traject into the attractor and never leave. But some initial conditions will cause the solution to fly off into infinity. Starting points which cause the solution to traject into the attractor lie in the basin of attraction. It is an interesting exercise to actually draw this basin surrounding the attractor. I seem to recall it is a fractal but not entirely sure of this.

 

[Stephen Tashi]

I'm working on a project

What kind of project? Are you writing a term paper? - for a class?. What grade level?

 

[platosuniverse]

What kind of project? Are you writing a term paper? - for a class?. What grade level?

I'm in a local debate group here and one of the members is into the Deepak Chopra and Biocentrism and he said it's impossible for the material brain to tell the material brain which memories it wants to recall.

He says their must be an immaterial observer. I just though there would be some mechanism that I could point to when we meet up on Thursday. Sadly, I haven't found a good explanation.

 

[Drakkith]

'm in a local debate group here and one of the members is into the Deepak Chopra and Biocentrism and he said it's impossible for the material brain to tell the material brain which memories it wants to recall.

Whoever is telling you this has absolutely no idea what they're talking about. Their opinion is not based on factual evidence from observing the real world, but on some abstract philosophical idea that simply sounded good to them.

He says their must be an immaterial observer. I just though there would be some mechanism that I could point to when we meet up on Thursday. Sadly, I haven't found a good explanation.

Unfortunately we just don't know enough about the brain at this time. But there is no reason to believe in some immaterial observer. There's no evidence for such a thing and many similar ideas have been shown to be false in the past as our knowledge of the world advanced.

 

[sysprog]

I'm in a local debate group here and one of the members is into the Deepak Chopra and Biocentrism and he said it's impossible for the material brain to tell the material brain which memories it wants to recall.

He says their must be an immaterial observer. I just though there would be some mechanism that I could point to when we meet up on Thursday. Sadly, I haven't found a good explanation.

Instead of setting about the daunting task of outlining the whole process of memory recollection, you might look at some neural events that can break the chain. For example, some stroke victims suffer expressive aphasia (also called anaphasia), which is correlated to region-specific drastic reduction in cortical neural activity.

Just as a paralytic can't keep the efferent and afferent source-sink motor-neural message throughput working, so the expressive aphasic can't remember all the formerly at-the-ready word-to-concept lexical correlations, For example, he/she might have to resort in mid-sentence to a circumlocution such as "thing that you write with" for pen, and can say "yes, that's it" when someone suggests "... pen?".

The overall system embodies a multitude of necessary conditions for accomplishing the realization of sufficiency of conditions for the feat of memory recollection.