There are two main types of memory within the human brain, or mind. They are ‘episodic memory’ and ‘semantic memory’. Episodic memory is the memory that we have of specific events that we describe as a memory; this type of memory is typically subjective memory, or first-person memory. This type of memory is typically expressed through phrases such as “I remember when...”, or “I remember that...”. Semantic memory is a different type of memory in that it is a collective thought, or knowledge; it is the memories that we have stored that we claim as knowledge. This type of memory entails things such as knowledge that birds have wings, 2+2=4, etc. It is inherent memory that allows us to have a basic knowledge of the world.
Collins and Quillian
One of the leading theories of semantic memory that we are going to explore is the “Collins and Quillian (and Loftus) Model” (256); this theory deals explicitly with “semantic memory, comprehension and meaning” (256).
The first part of this theory attempts to explain what the structure of semantic memory looks like. This theory introduced an explanation of the structure of semantic memory that is called a ‘network’ (257). A network within semantic memory is the structure by which semantic memory works; it is what stores long-term information and allows it to become accessible in the mind. Some parts of the network include ‘nodes’ and ‘pathways’; nodes are sections in the network in which stored information stays, and the pathways are the roads that transmit the information from one node to another and allows the network to work effectively so that we can accurately retrieve information that has been stored in the memory.
The second part of the theory from Collins and Quillian tells us that the “major process that operates on this structure is spreading activation, the mental activity of accessing and retrieving information from this network,” (257); this method of access explains why we need the networks inside of our memory, it tells us what the networks are used for. Spreading activation is a process of memory retrieval that is prompted by words or concepts; when we hear, read or experience a word or concept, we access the idea in our memory via the networks within the brain. The next step in this retrieval is our associations and reactions to the idea; for instance, our semantic memory retrieves the idea of a bird, it then goes through all of our other memories to find associations (robin, penguin, eagle; or wings, beak, feathers, etc) that go with that concept in order to interpret and understand the idea fully. The following diagram illustrates spreading activation in the mind:
FIGURE 7-1 p 258
Collins and Quillian discovered that once an idea is activated, it opens up possibilities of other ideas being activated in our minds and often creates a domino effect of acknowledgment; by activating one idea, we are able to connect and expand concepts through the network.
The third part of this theory is called “intersection search” (258); this follows the spreading activation process. This search is what happens when one concept is activated in the brain and begins travelling through pathways within the network which intersect with other pathways that connect with different concepts. After the pathways intersect, our mind must make a decision on which pathway we must follow to come to the correct or at least semi-correct conclusion. Although these activations, connections and pathways are intricate and allow us to retrieve and store information, that information does not stay active in our brains long-term, but continues to be stored in our semantic memory waiting for another instance where it may be activated.
Smith
In contrast with Collins and Quillian’s model, we have “Smith’s Feature Comparison Model” (260). Smith’s model takes a different approach to the retrieval of memory than Collins and Quillian’s model. The first, and “most basic structural element was the feature list,” (260). The feature list is also known as ‘semantic features’. In contrast to the ‘network’ theory of Collins and Quillian, Smith suggests that our minds are simply compiled of lists. According to Smith, these lists are organized by “definingness” (260); this means that our minds categorize the semantic feature lists based on priority, which means that when we are given a concept, we automatically switch to the best related concept on a list for that concept. We rank concepts in order of priority; the features listed at the beginning of our memory are referred to in Smith’s model as defining features and the features that we consider later on are considered characteristic features. For instance, when discussing the concept of a tree the defining features are that it has a trunk, some branches, roots, wood, etc; and the characteristic features would be concepts such as ‘paper’, ‘sawmills’, etc. These characteristic features still relate to the core concept, which is why they are still involved in the list, but they are not essential to the description of the concept or object.
Information retrieval in this model is known as “feature comparison”; this method is slightly difficult to explain without examples, but I will do my best with the help of the following diagram:
FIGURE 7-2 p 261
When we are given two concepts, we bring up the feature lists of each concept to compare and contrast the features so that we can determine whether or not the two concepts are compatible. In this figure, which is shown on page 261 of the Cognition book, shows the feature lists of the words ‘robin’ and ‘bird’; this illustrates how our minds determine whether or not the statement ‘A robin is a bird’ is true or false. Without realizing it, we automatically state that this statement is true, but our mind has gone through many processes to get to this conclusion; our semantic memory has been at work. The feature lists in this diagram illustrate that ‘robin’ and ‘bird’ have four out of five of the same defining features, and the ‘robin’ list has the ‘red-breasted’ characteristic. Since eighty percent of the features listed in our features list are the same, we decide that a robin is indeed a bird. There are two stages, however, of this comparison. We have just seen and explained Stage I of this comparison; Stage II is when a concept is not as simple as ‘a robin is a bird’ and becomes something like “a chicken is a bird”. This example as found in page 262 is slightly more ambiguous in that a chicken does not have the same characteristics as most other birds, and therefore makes this decision more difficult as only a few defining characteristics are similar. In essence, Stage I allows for fast reactions and quick decisions, as the defining features typically line up quickly and easily; Stage II causes slower reactions and decisions, as the defining features may be less determining and more difficult to place.
Empirical Tests
There are different types of tests that are done to ensure the truth of the models of semantic memory. The most common task involved in these tests is called the “sentence verification task”; this is a task whereby subjects need to respond true or false to a simple statement. When subjects are tested using such simple concepts, their reaction time is faster than tests that involve more advanced decision-making. This test proves the theory posed by Collins and Quillian by showing that when given a concept, the mind will spread to other concepts almost simultaneously; this test does not however support Smith’s model of memory.
Works Cited
Ashcraft, M., Klein, R. (2010). Cognition. Toronto: Pearson.
