Hacker News Reader

Comment by shrimp_emoji | Hacker News Reader

>Sweller's Cognitive Load Theory theorizes that we have an unlimited memory capacity

How is a memory stored on a neuron or series of neurons?

And, given that, why would the amount of memory not be capped by number of neurons and synapses?

My understanding is there is a "cap", but the brain basically has the equivalent of "garbage collection" for memory objects that are no longer being used.

https://www.sciencemag.org/news/2014/05/how-brain-deletes-ol...

CLT is a theoretical model indicating how we process information, not how it is stored. There is long term memory, which stores it schemata (knowledge), and working memory, where schemata are retrieved from long term memory and processed. Working memory, however, has a finite capacity, resulting in mental effort (cognitive load) being required for processing and maintaining schema. The latency/delay in retrieving and processing schemata (or lack thereof) is labeled as "expertise". CLT is primarily used to study the mental effort in processing things like worked examples and problem solving. This is where the idea of having random pictures on PowerPoint slides is seen as bad.

However, all of our current theoretical models on memory hold the premise that our capacity is infinite, as we are able to recall memories and information from decades prior (think about minuscule knowledges like walking). There is a "forgetting curve" model as well that supports the idea to spaced repetition as a method of reducing the latency of retrieving schemata from long term memory. Further, we identify diseases like Alzheimer's, dementia, etc as diseases as one of their trademark symptoms is the loss of mental ability. Is this forgetfulness due to losing scheme from long term memory or an inability to appropriately retrieve/process schemata?

Regardless, as I mentioned in my previous comment, if our capacity is infinite, then our only limitations are time and motivation. Adults, unlike children, need to be motivated beyond "because mommy and daddy said so" in order to allocate time to practice.

I think this is why "full immersion training" when learning a language is so powerful - regardless of motivation, you have allocated MUCH more time to "practice". Adults that move to another country and are willing to learn will have an easier time learning than someone that doesn't move but wants to learn or someone that does move but doesn't try to learn.

The last bit would be people willing to "teach" novice adults. This is another issue all together, but could still be framed as a time and motivation problem.

My guess is you are actually storing pointers into off-brain storage such as books, using keywords.

So perhaps there is in fact a limit, but you get a lot further with pointers than actually saving the periodic table, Hamlet, and a French dictionary.

Maybe the amount of neurons grows as their "usage percentage" reaches a certain threshold, making the memory capacity unlimited.

Surely we've done enough brain experimentation to know that "people who do more learning have more neurons" is false?

Well, haven't there been findings that London cab drivers have higher-volume hippocampuses because they use them more in orientation[0] and that meditation increases white/gray matter[1]?

0: https://www.wired.com/2011/12/london-taxi-driver-memory/

1: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3004979/

Maybe it is limited but you can't hit the limit in a lifetime.

> How is a memory stored on a neuron or series of neurons?

We don't exactly know. Sweller's Cognitive Load "Theory". There's a reason it's a theory.

I would advise against mixing theory's scientific meaning with the more common "guess" meaning. The heliocentric theory and cell theory would like a word.

Reply on news.ycombinator.com