6
$\begingroup$

In computers, finding a single word is realized through serial attempts across all available connections to find a specified target.

How does the brain solve this? How does the whole process, from the command to recall some memory to actually "play" its algorithm look like?

By "recall" I mean literally, attempting to bring back an image, a word, a melody or any remembrance from memory.

$\endgroup$
  • 2
    $\begingroup$ "In computer, for finding a single word there has to be a movable part that tries all the connections to get what you're finding." - there are literally no moving parts in modern computers with solid state drives and fanless processing units. Moving things (e.g. magnetic hard drives) have nothing to do with computer's ability to find information. It's all about file system abstraction and hierarchy. $\endgroup$ – Eli Korvigo Dec 1 '15 at 17:52
  • $\begingroup$ @EliKorvigo aha...IT teacher told me...anyway, I still can't imagine the mechanism neither for the brain and computer $\endgroup$ – Probably Dec 1 '15 at 19:05
3
+50
$\begingroup$

This is a great quesion! There is lot of work still being done on this. Unfortunately I don't think I know enough about CS or computer engineering to make analogies, so I'll give a quick crash course and some examples instead and hope you find it helpful. Long story short though is that we are just now making really exciting headway into how memory might actually be stored and recalled and its usually a very dynamic, active, and possibly even stochastic process, not really like a search algorithm in a computer at all.

The brain works differently than a computer in that it encodes all of its information in frequency of signal firing and the patterns of firing between circuits with varying stregnths of connections based on experiences. Frequency is dependent on the neuron itself and patterns of firing between circuits are dependent on where the neurons project to, and how strongly they interact with the next neuron. At this very basic level memory is stored and recalled by placing proteins in the membranes of neurons that are strongly activated to make them easier activated in the future. These are sort of the first steps toward memory. This is called long term potentiation and just means that this neuron is primed to fire more readily and stronger in response to similar input in the future.

To build on this, there is a classic example of this occuring and causing sensitization ( an exaggerated response to the same stimulus) in conditioning experiments in the sea slug Aplysia. Aplysia has a reflex where it will withdraw its gills slightly if you touch its siphon. You can pair a shock to the tail (unconditioned stimulus) which elicits a large withdrawal of the gills, with a conditioned stimulus (touching its syphon to elicit the withdrawal reflex) to get the sea slug to associate the two. The end result is that the sea slug will learn/"remember" to withdraw its gills fully in the presence of the syphon touch. No learning cannot happen without recall of some sort, and in this example the slug was associating and learning from two "circuits" to produce a different behavior.

In something a little more complex than a slug, memory and recall still functions much the same way, its just the circuits that are interacting are entire brain regions. Most people view the hippocampus as acting sort of like a relay center and boot loader for different memories by taking external input and sorting them to the right parts of the brain. A good example of this would be the difference between dorsal and ventral streams of information in the brain. For a really basic breakdown of this phenomenon there are two areas of your brain that I will focus on - the parietal lobe and the temporal lobe. The parietal lobe is known to be involved in spatial processing, and the temporal lobes are known to be involved in semantic and auditory processing. When information enters the brain (eyes, ears, nose etc.) it will be processed by sensory structures, sent to the thalamus and hippocampus to be "relayed" to the right areas of the brain. Relayed in this case is going to be dictated by whatever connections and how strong they are in the brain. In this case from proteins expressed in the cellular membrane that make some neurons in the hippocampus more responsive than others, so they are more likely to fire. Then, it depends on where they are projecting to that determines where they information is sent.

It just so happens the brain is arranged in such a way that neurons projecting dorsally, or to the parietal lobe I mentioned, will tell you "where" the stimulus is, and those projecting ventral to the temporal lobes will tell you "what" it is. This is achieved because these lobes are highly interconnected to even more regions in the brain and send their processed "where" and "what" info as input to other regions. These regions will all fire a certain way in response to this information and that pattern of firing can be considered a recollection or memory. Sorry that this is dense and so crash course-y but I hope it is helpful in clarifying how the brain recalls some information. Of course this is a developing field and we are learning more about how it functions every day. These are just some classic examples.

$\endgroup$
  • $\begingroup$ The ventral and dorsal streams in the visual system are the what and where streams in visual processing from the bottom up (optic-nerve - thalamus - V1 - etc). OP asks for recall of visual images, i.e. top-down. This is not an answer to the question. $\endgroup$ – AliceD Dec 3 '15 at 12:41
  • 1
    $\begingroup$ @Christiaan, he answered the question. I think you are getting confused by his reference to the processing of sensory information. What he is saying is that the signals produced by sensory structures trigger a chain reaction (i.e. action potential) that is transmitted from one neuron to the next in a very specific network of neuronal connections to store an equally specific memory associated with whatever sensory signal was sent. Correct me if I misunderstood you drtran, and +1 for a well put and accurate answer. $\endgroup$ – CDB Dec 5 '15 at 0:40
  • 2
    $\begingroup$ @ CDB I did mean that sensory input will trigger memory recall by activating specific networks. I should have gone further and said that there can be other types of input into these networks to elicit memory recall other than external stimulus. For example, top down processing from the cortex can be considered an external stimulus to the hippocampus if the hippocampus is your reference frame. I just thought external stimuli would be easier to explain. $\endgroup$ – drtran Dec 5 '15 at 0:53
  • $\begingroup$ @Christiaan, splendid, thank you! I suspected that it was merely a misunderstanding. I'm glad that I could clear it up. $\endgroup$ – CDB Dec 5 '15 at 21:49

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.