The semantic network model of Collins and Loftus allows for more "freedom" than that of Collins and Quillian and for that reason seems to "work" better for me. One of its biggest strengths in my opinion is that it recognizes that many connections in our mind are not logic-based at all and are instead formed through personal experiences. This is very true for me at least- if someone were to say donut I would not necessarily think of the food first but instead think of a pillow (because of my favorite donut pillow). The connection is not logical but it makes sense to me because of my personal experiences. Another strength of Collins' and Loftus' semantic network model is that allows for varying degrees of strength between connections. As pointed out in the article, although an ant and a canary are both animals it would take one longer to connect an ant to animal than it would with the canary. The article however points out that this theory's weakness is that allows for very little prediction of reaction times. Before being able to predict someone's time in a reaction test you would have to map out their thought network first.
The hierarchical network is not learned and part of a formal education but rather how we associate concepts and properties. The connections made are from the subjects, their functions and what they can be linked to. The strengths of the activation model can be classified under different categories, such as, the variety of information gathered around something when brainstorming or the relations of subjects and how they support and affect each other. From the diagram in the document, Bird and Airplane are linked together with Can fly, and Bird is also linked to Animal. The link shows that the concept and property of Airplane can be defined differently. The airplane can fly, which is true, or the airplane is a kind of bird, which is false. Those kinds of relations in a single network can be confusing at times. The weakness of the spreading activation can be observed when making a research on a computer and it gives the opposite result of what the human brain is thinking. This is because the information was automatically merged together with other concepts and properties. When we hear a word, the activation that spreads could give us irrelevant information, or the process to come to the right one could take a long time. The length of the links in the network makes the information travels faster but might not directly relate.
I agree with Collin’s and Loftus understanding of the semantic network model. I believe that it allows the individuality of the person to explain the semantic networks in their brain as opposed to general logic. I agree with Kristen’s example of her “donut pillow” and think it makes a valid point as to why the Collins & Quillian hierarchical network model is flawed. One thing that was not presented in Collin & Loftus’s model was the link of phrases to concepts. They discusses that you need a category such as “bird” to think of “can fly”. I believe that commonly know phrases need to have a place in the semantic model. For example, when I saw the phrase “can fly” the first thing that pops into my head is “You Can Fly! You Can Fly!” from the Disney Peter Pan film. I think it would be interesting to see how powerful the connection between culturally known phrases is within the semantic network model.
4 comments:
The semantic network model of Collins and Loftus allows for more "freedom" than that of Collins and Quillian and for that reason seems to "work" better for me. One of its biggest strengths in my opinion is that it recognizes that many connections in our mind are not logic-based at all and are instead formed through personal experiences. This is very true for me at least- if someone were to say donut I would not necessarily think of the food first but instead think of a pillow (because of my favorite donut pillow). The connection is not logical but it makes sense to me because of my personal experiences. Another strength of Collins' and Loftus' semantic network model is that allows for varying degrees of strength between connections. As pointed out in the article, although an ant and a canary are both animals it would take one longer to connect an ant to animal than it would with the canary. The article however points out that this theory's weakness is that allows for very little prediction of reaction times. Before being able to predict someone's time in a reaction test you would have to map out their thought network first.
The hierarchical network is not learned and part of a formal education but rather how we associate concepts and properties. The connections made are from the subjects, their functions and what they can be linked to. The strengths of the activation model can be classified under different categories, such as, the variety of information gathered around something when brainstorming or the relations of subjects and how they support and affect each other. From the diagram in the document, Bird and Airplane are linked together with Can fly, and Bird is also linked to Animal. The link shows that the concept and property of Airplane can be defined differently. The airplane can fly, which is true, or the airplane is a kind of bird, which is false. Those kinds of relations in a single network can be confusing at times. The weakness of the spreading activation can be observed when making a research on a computer and it gives the opposite result of what the human brain is thinking. This is because the information was automatically merged together with other concepts and properties. When we hear a word, the activation that spreads could give us irrelevant information, or the process to come to the right one could take a long time. The length of the links in the network makes the information travels faster but might not directly relate.
I agree with Collin’s and Loftus understanding of the semantic network model. I believe that it allows the individuality of the person to explain the semantic networks in their brain as opposed to general logic. I agree with Kristen’s example of her “donut pillow” and think it makes a valid point as to why the Collins & Quillian hierarchical network model is flawed. One thing that was not presented in Collin & Loftus’s model was the link of phrases to concepts. They discusses that you need a category such as “bird” to think of “can fly”. I believe that commonly know phrases need to have a place in the semantic model. For example, when I saw the phrase “can fly” the first thing that pops into my head is “You Can Fly! You Can Fly!” from the Disney Peter Pan film. I think it would be interesting to see how powerful the connection between culturally known phrases is within the semantic network model.
Can someone please delete this blog? Or delete my postings.
Post a Comment