easurement

semantic memory is not amenable to any forgetting because it embodies general knowledge that we all possess. In sentence verification task, the participants are asked to indicate whether the given sentences are true or false. Faster the participants respond, better retained is the information needed to verify those sentences (see Activity . for use of this task in measurement of semantic knowledge).

c) Priming (for measuring information we cannot report verbally) : We store many kinds of information that we can’t report verbally - for instance, information necessary to ride a bicycle or play a sitar. Besides, we also store information that we are not aware of, which is described as implicit memory. In priming method, participants are shown a list of words, such as garden, playground, house, etc. and then they are shown parts of these words like gar, pla, ho, along with parts of other words they had not seen.

Participants complete parts of seen words more quickly than parts of words they had not seen. When asked, they are often unaware of this and report that they have only guessed. get organised in schema . They are mental frameworks which represent our knowledge and assumptions about the world.

For example, think of a schema of a drawing room. It will have different objects/things, like a sofa set, center table, paintings, etc., which are found in a drawing room and where they are found in the drawing room. So far we have examined the concept as the basic level at which knowledge is represented in the long-term memory and the notions of category and schema as the first level at which concepts are organised. Let us now look at a higher level of organisation that concepts acquire in the long-term memory.

In the year , Allan Collins and Ross Quillian published a landmark research paper in which they suggested that knowledge in long-term memory is organised hierarchically and assumes a network structure. Elements of this structure are called nodes. Nodes are concepts while connections between nodes are labelled relationships , which indicate category membership or concept attributes. In order to verify the proposed network structure of long-term knowledge, participants of experiments were asked to verify the truth of the statements such as ‘canary is a bird’ or ‘a canary is an animal’ (answer was in Yes/ No).

These were generally class-inclusion statements in which the subject was word ‘canary’ (perhaps, you know, it is a bird) and the predicate took the form ‘is a’. A critical finding of such experiments was that as the predicate became hierarchically more remote from the subject in a sentence, participants took longer time to verify that it is true or false. Thus, people took longer to verify that a ‘canary is an animal’ compared to that which said ‘canary is a bird’ because bird is an immediate superordinate category in which canary is subsumed while animal is a superordinate category which is more distant and remote from the concept canary. According to this view, we can store all knowledge at a certain level that ‘applies to all the members of a category without having to repeat that information at the lower levels in the hierarchy’.

This ensures a high degree of cognitive economy , which means maximum and efficient use of the capacity of long-term memory with minimum redundancy.