Chapter 8 - Associative Theories of Long-Term Memory (LTM)
The Network Notion
- Nodes: represent individual ideas like the knots in a fisherman’s net.
- Nodes are tied together by connections called associations or associative links.
o Like highways connecting cities
- Therefore, searching through memory is to begin at one node and travel via the associations until the target info is reached.
- These connections are established if:
o The learner pays attention
o The learner engages in active intellectual engagement
o The learner thinks in distinctive ways
- Nodes receive activation from their neighbours, and as more and more activation arrives at a particular node, the activation level
for that node increases.
- Node becomes activated only if it has received a strong enough input signal
o Like a light bulb being turned on or a neuron firing.
- Eventually the activation level reaches the node’s response threshold. Once this happens, the node fires.
- Below threshold activation levels (subthreshold activation) is still important, as it is like the node is already “warmed up.” if
another subthreshold input arrived, the inputs my add together or summate, causing the node to fire.
- Nodes themselves are NOT neurons
- The network of nodes in LTM includes the network of detectors in chapter 3.
- Spreading activation: as each node becomes activated and fires, it serves as a source for further activation, spreading through the
- However, not all pathways/connections/associations are equally effective, some are better than others.
Evidence Favouring the Network Approach
- Why do hints help us remember?
- Mention of a topic will activate all nodes representing knowledge about that topic, activation spreads outward, eventually
reaching the target information.
- However, with a hint, two topics are given, and therefore more nodes will activate.
- The target info will be activated by 2 sources instead of one.
o Ex: Capital of South Dakota? Hint: It’s a man’s name.
- Context Reinstatement:
o Ex: if you learned underwater, you will remember better underwater.
- The context serves like a hint would, triggering extra info to stimulate the nodes.
More Direct Tests of the Network Claims
- Lexical-decision task: participants are shown a sequence of letters on a screen and are asked to determine whether or not these
sequences make real words. They answer in a yes or no fashion, basing the responses on whether the word is in their “mental
dictionary” or not. This allows the speed of how quickly one can locate a word in memory to be measured.
- Meyer and Schvaneveldt (1971) expanded this by presenting participants with pairs of letter strings. Sometimes the 2 words were
related, sometimes they weren’t, and sometimes one was a real word and one wasn’t.
- Participant’s responses’ were faster if the stimulus words were related, as the first word would prime the second.
o Ex: Bread / Butter: participants see the word bread, and all nodes related to brad activate as well (including butter).
Then, when the participant tries to remember the word butter, the nodes are already warmed up due to subthreshold
- Sentence verification task: participants are shown a series of sentences, both true and false. Participants are asked to hit “true”
and “false” buttons as quickly as possible.
- Responses were faster to sentences like “a canary is a canary” than “a canary can fly” because the latter sentence requires
traversing two links, from canary to bird, and then from bird to fly. (Collins and Quillian).
- Also, more typical stimuli make faster response times (robin > peacock).
- Degree of fan: the amount of nodes fanning out from a topic.
o Ex: you probably know more about robins than aardvarks.
- So, if there are only 5 links radiating from aardvark, each link will receive 20% of the total activation. However, if there are
much more links radiating from robin, the activation spreading outward will be much more thinly divided. Therefore less
activation will be received and less activation will be carried to the neighboring nodes, making the memory search very slow.
Retrieving Info From a Network
- Associative links guide searches through the network like hyperlinks guide a search through the Internet. However instead of
typing a word into a search bar, the network has:
o Input nodes: receive most of thei