before asking how information is stored, I need to understand

  1. how does episode start and end are determined by the brain ?
  2. how do i remember a movie ?

can you suggest a link to article addressing this issue


  • $\begingroup$ We don't know; it also isn't necessarily true that it is necessary for episodic memories or experiences to actually have any distinct start and end. $\endgroup$ – Bryan Krause Dec 1 '16 at 16:32
  • $\begingroup$ thanks. how can the brain recall an episode if it's not bound by some way like start/end "flag" or by special "sub circuit" (hard wired) $\endgroup$ – BNR Dec 1 '16 at 16:54
  • $\begingroup$ If it’s associative memory the chances of ‘multiple recalls’ is high without some boundaries that reduce its ‘fuzzy level’ $\endgroup$ – BNR Dec 1 '16 at 17:18
  • $\begingroup$ Two answers to that comment: 1) Think a bit about your own memory. Are you really able to remember your whole life in sequence without interruptions or getting distracted? 2) The 'boundaries' you are talking about can be fairly subtle: just an increase in connectivity that causes an increase in probability of states progressing in the order of the original memory. There is actually a lot of research on just how poor our memories are, as well as how poor our estimates of how good our memories are (we think we remember well when we don't). $\endgroup$ – Bryan Krause Dec 1 '16 at 17:23

We don't have a full understanding of how episodic recall really happens in the brain, but there are theoretical approaches to understanding memory that are supported (but not shown definitively - it is really difficult to design appropriate experiments that are not beyond current technological capabilities) by experimental evidence. These approaches don't require any sort of hard "start" or "end" signal for episodic memory to work. In these approaches, an episodic memory is thought of as a sequence of brain states that correspond to previously experienced states.

The sequence occurs because connectivity promotes the network to progress from State A to State B to State C, etc. Additionally, all of the states consisting of that one "connected" memory are also associated, so that they prime one another.

To recall, you don't need to start at State A - you can start anywhere in the chain, and just recalling State C will prime you to also recall State A and B, so you can mentally "rewind" the memory a bit (we have no idea how exactly this type of conscious control works), even though it progresses most clearly in the A->B->C sequence.

The point of entry that allows you to initially get in to this chain might not be part of that episodic memory, either. For example, maybe you see a cat, and this leads to to an episodic memory of one event you shared with your own cat 5 years ago.

Brains are very dynamic, they do not behave like a feed-forward computation you might implement in a simple computer program. It is possible that there are "start" or "end" signals for episodic memories that we haven't found yet, but they may not be necessary at all.

For some further reading you could look for references on attractor networks, pattern separation and pattern completion, and theories of memory formation.

The book by Dayan and Abbott on theoretical neuroscience is also a good approach to some of the more basic background and includes an introduction to biologically-inspired models of learning and memory.

  • $\begingroup$ thank you Bryan. let us assume i have 5 cat's episodes. i might be confused if all are recalled. may be i have 1000 episodes with my family - all are recalled ...i'll be lost :) $\endgroup$ – BNR Dec 1 '16 at 17:27
  • $\begingroup$ You might be interested in reading about "attractor networks" and how they work - TL;DR version is that a network is structured so that there are stable states that are associated with particular memories or concepts. Thermodynamic analogies are appropriate to attractor networks, so the way you would think about it is that the 5 cat episodes are each stable attractors, with a relatively low energy barrier to passing between them. $\endgroup$ – Bryan Krause Dec 1 '16 at 17:30
  • 1
    $\begingroup$ Could you add some references to your answer? I'm not well enough versed in the area to judge whether or not it's a good answer, so some basic background references, links to wiki pages, good reviews in the literature, etc. would be much appreciated. $\endgroup$ – MattDMo Dec 1 '16 at 17:59
  • $\begingroup$ @MattDMo I'll try to come up with some basic references at least for further reading...this is the sort of thing people write books about that contain a fair bit of speculation and hand-waving, and far from settled knowledge, so it's a bit difficult to point to a particular authoritative source. $\endgroup$ – Bryan Krause Dec 1 '16 at 18:02
  • 1
    $\begingroup$ @MattDMo The first sentence of my answer was "We don't have a full understanding of episodic recall, but let me try to answer from just a theoretical perspective" - by this I meant to convey that we don't have answers as to how episodic recall really happens but there has been a decent amount of educated speculation. I'll try to clarify a bit. $\endgroup$ – Bryan Krause Dec 1 '16 at 18:10

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.