A few words from many years ago

I have been intrigued by a comment from my Mom’s colleague when I was in China last month, after about 5 years since I saw her last time. She can still remember a joke that I said over 10 years ago. Although she tried to make fun of me by quoting that phrase, I started contemplating about a question arisen by the instance: why human being can recall a tiny piece of information about a person after such a long time.

After reading the capacity paper (Just et al., 1992), I am aware of its theory about the activation on information processing and storage, but when it comes to retrieve information from long term memory, despite the low-span or large-span people’s capacity differences in terms of working memory, does element/item in long term memory carry activation level? If so, will the activation also mediate the process, thus affect the outcomes?

I have come across three models pertaining to information retrieval from long-term memory , and it seems that “matching” is a key word for all of them. For example, imaging the information seeking process resembles the scenario of searching for your bags at luggage claim, you are going to examine each matching item until the first piece is found, then the entire process starts over again. There is also another one called Resonance Retrieval Theory that treats information as a vector with elements representing different conceptual subjects.

Let me propose this: people may tend to organize or categorize long term chunks according to a specific object such as a pet, a friend, etc.  Each item appears like a vector with an equal number of attributes as long as that person put them into the identical category.

Element CategoryHuman_Male_001111 = { A[0], A[1], A[2] …. A[n-3], A[n-1]), A[n-1] }

The attribute can be an individual element such as hair color, relationship status, education degree, etc. However, they are assigned with different weights due to the product of external stimulus and the person’s strategy to encode information. There is an pointer that always waits at the first attribute- A[0], while the first attribute is reserved to any elements bearing the most substantial weight. That being said, they are changeable, because any attributes can be strengthened or weakened by various events, and this process may be done by an unknown internal mechanism. Thus, when the working memory calls for information activated by means of retrieving and decoding items/elements from long term memory, the weight of the candidate may affect the efficiency of the process, and I would like to name the weight as activation level for each information unit in long term memory. 

The above model/process is just my "random" thought, I am looking forward to reading more literature to see it is "correct" or the opposite. :)

Reference:
Wickens, C. (1980). The structure of attentional resource. In R. Nickerson (ed.), Attention and Performance VIII, 239-257, Hillsdale, NJ: Lawrence Erlbaum.

(© 2012 Miaoqi Zhu)