Memory and the Brain

Although not much is understood about how the brain encodes memories, it is possible that the brain is a network of information with different degrees of connectedness between different nodes in the network. Every time that information it strengthens the connections and forms new connections. It appears that these connections need to be refreshed from time to time or they become completely lost.

If this model of memory is correct for the human brain, then it appears to explain why we remember and forget certain pieces of data. If we review a piece of information frequently and under many different contexts then this creates many more connections to that information. Furthermore, when we attempt to recall the information our brain's context may be similar to when we memorized it. Repetition over a long period of time would appear to be one of the best ways to create more connections. You could also conjecture that the connections are weakened or lost altogether in the passage of time. Also, retrieving appears to strengthen the connections. Finally, it seems reasonable to believe that a particular piece of data could be duplicated in multiple places in the brain.

Intuitively, our ability to remember, or retrieve data, from our brain is determined by:

1) The strength of each connection
2) The number of connections
3) The duplication or redundancy of the data

It would appear that if data is connected in a network then these strategies of memorization should be very good at maximizing long-term retrieval:

1) Repetition
2) Review
3) Learning at a deeper level of detail

The first and second strategy are widely used by virtually everyone. The third strategy is to memorize at a level of detail beyond what one wants to retain. This would intuitively connect the brain in those areas more by allowing the data to become relied upon for other data. I was not exposed to this idea in school although it appears to happen naturally when you study history books and learn all sorts of minutia about George Washington, you tend to not forget who George Washington is.

It is very popular to emphasize mnemonic techniques in a discussion about memory, but I'll have to admit, they have not been very useful to me. For short-term cramming on exams, mnemonics seem to work great. However, despite the hundreds of times I have applied the techniques to convert words into pictures and pictures into stories, this has never resulted in enhancing my long-term memory. In general, mnemonics do not work for long-term memory. At least not in my experience. This isn't to discount certain mnemonics such as Every Good Boy Does Fine and PEMDAS, but for large volumes of information that we need to memorize, rote repetition, disciplined review, learning at a higher level of detail appear to give better retention in your long-term memory and also appear to have a higher rate of retrieval.

Two questions that I'm interested in hearing feedback on:

1) Is this model of memory in the brain a scientific model? Are there inaccuracies?
2) Has anyone had a different experience with mnemonics? Have they worked for long-term (say 5+ years length) memory?