Alright, friends, as I concluded our previous blog-isode last Saturday, I mentioned that we would be exploring the work of two awesome psychologists: Collins and Quillian..
Let's delve in:
The Collins and Quillian model that was originally proposed was on a hierarchical basis.
In such a semantic hierarchy, a specific word for example, "canary", is stored under the category it belongs to for instance, "bird", which will be stored under another category e.g., "animal", and so on.
In this way, some characteristics that characterize each of these words would then be laid under the most appropriate node .
For example , for "canary" you could possibly store features like: "is yellow" and "can sing".
Source
On the other hand , for "bird," you might store things like : "has wings" and "has feathers." The same holds true for bigger categories such as animal or living thing.
Collins and Quillian's idea about how our brain organizes information was very neat, but reality isn't always as tidy as theories suggest.
For instance, not all things neatly fit into clear categories. When people were asked to rate different birds based on how typical they are, robins were seen as most typical, chickens less so, and penguins were barely considered birds at all.
Source
These differences in typicality affect how quickly we recognize things in our memory. People respond faster when identifying typical birds like canaries compared to less typical ones like penguins or ostriches.
These findings suggest that the way information is stored in our memory isn't as clean-cut as the hierarchical model suggested.
The difficulties of the hierarchical model led to the formulation of several alternatives.
One was the spreading activation model developed by Collins and Loftus, in which semantic relationships were built directly into the network .
In this network, shorter arrows between two nodes showed a closer connection in meaning. This could be because of their position in a hierarchy (like canary and sing), their similarity in meaning (like apple and orange), or well-known associations (like Pepsi and Coka Cola).
Source
Collins and Loftus suggested that nodes can be turned on, kind of like what happens in pattern recognition. When a node gets turned on, it spreads activation to other nodes. This happens faster and stronger for nearby nodes compared to ones farther away because the activation weakens as it spreads.
So, How can this spreading activation hypothesis be tested?
Evidence from a study showed that when people were asked to think of words in certain categories, it affected how quickly they responded. For example, if they were asked for a fruit that starts with the letter G, the time it took for them to say their answer, like "grapes," was measured from when they heard the category word "fruit."
After a short while, the person is tested again. Sometimes, they're asked for a word in the same category but with a different starting letter, like "S - Fruit." This time, it takes much less time to come up with a word, like "strawberries," compared to the first time. What seems to have happened is that the category gets turned on and stays on for a bit, similar to leaving the lights on in a section of a library. Now, it's easier to find words in that part of the library.
More proof comes from various experiments on semantic priming. In one well-known study, subjects were shown two sets of letters, one above the other.
Here are some of examples they used:
NARDE NURSE NURSE
DOCTOR DOCTOR BUTTER
The participants had to press a "yes" button if both sets of letters formed real words (like NURSE - BUTTER and NURSE - DOCTOR), and a "no" button if one or both of the sets were not real words (like NARDE - DOCTOR).
The researchers were only interested in the two pairs where the answer was "yes." (The "no" items are just there to make sure the person follows the task.) The findings revealed that people reacted faster when the two words were related in meaning (like NURSE - DOCTOR) compared to when they weren't related (like BREAD - DOCTOR).
Interesting Right?
It seems that seeing the word NURSE made it easier for the brain to think of related words like DOCTOR, but it didn't have the same effect on unrelated words like BUTTER.
This is in line with the Collins and Loftus model : The node corresponding to NURSE was activated , and this activation spread to nearby nodes ( e.g. , DOCTOR ) on the network .
In simple terms, it means that words are stored and organized based on how they're connected in meaning(semantic relatedness).
The Bus Stops Here for today:
Thank you, friends, for staying with me through these blogisodes. Your thoughts and opinions are always welcome and appreciated. I'd be happy to hear them. We will build on this in tomorrow's blogisode. Until then, stay safe, friends.
References and Links:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2699208/
https://lauraamayo.wordpress.com/2014/11/10/collins-quillian-the-hierarchical-network-model-of-semantic-memory/
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5413589/
https://study.com/learn/lesson/semantic-network-model-overview-examples.html
https://www.sciencedirect.com/topics/engineering/hierarchical-network
https://www.researchgate.net/figure/An-illustration-of-Collins-and-Quillian-1969-hierarchical-network-model_fig7_315776430
https://psychologicalresources.blogspot.com/2014/12/hierarchical-network-model-of-semantic.html?m=1
Posted Using InLeo Alpha
