chapter 9 concepts

Question 1

Why do our definitions of basic concepts like “dog” often fail?

Answer 1

It seems that for every definition (the way you describe something) there is an exception that can be made for all concepts. a shoe is work on the foot made for walking in etc, but what about a wooden shoe.

Question 2

Explain the meaning of the term “family resemblance”

Answer 2

There are features that are common in a family but are not all found in all family members, usually can see in two or three members. Say there is an ideal that has all the features, then the other family members will have a characteristic

Question 3

Describe prototype theory

Answer 3

A prototype is the average of the various category members, all the dogs are combined into an average prototype of what a dog is.

Question 4

What is meant by graded membership?

Answer 4

Membership in a category depends on the resemblance to the prototype, and resemblance is a matter of degree (not all dogs are the same).

Question 5

How can the prototype notion be tested?

Answer 5

Sentence verification; participants represented with a series of sentences and then they had to decide if they were true or false. Penguin is a bird and a robin is a bird.

Question 6

What are basic level categories?

Answer 6

Basic level of categorization, instead of specifying something specifically we use a broader category, such as how do you get to work? We don’t specify exact train, or type of car etc we just say car, train, walk ect.

Question 7

What are exemplars?

Answer 7

Is a specific remembered instance, in essence an example. Something specific, not just a chair but uncle jerrys chair

Question 8

How are exemplars similar to and different from prototypes?

Answer 8

Prototype is an average representing the whole category and an exemplar is whatever example of the category is that comes to mind (different examples may come to mind on different occasions.

Question 9

Explain how we rely on both exemplars and prototypes in our thinking about categories

Answer 9

Prototypes are useful for a quick summary of information about a category and examplars provide information that the prototype has lost providing more information and variability within the category.

Question 10

Describe a situation in which categorisation and typicality do not go “hand in hand”

Answer 10

That the judgement of a membership category depends on judgement of typicality. If something is typical of the category, then it will show strong resemblance and if not then vice versa.

Question 11

What else do we use to make category judgments besides typicality and how do we know?

Answer 11

Can a kettle be a toaster/ can a recoon be a skunk. Thing that are man made are judged differently to living things. As living is process with deep properties and no just mere appearances. The deep properties rely on a web of beliefs, beliefs arise as you have a broader understaning about other things, eg to understand what a racoon is then, you need to hace understanding of some biology, lifecycle and hereditary.

Question 12

3. Describe the explanatory theories that people hold and why we need them for categorisation

Answer 12

Theories for a concept was also crucial for out decision of resemblance, which one matter in guiding resemblance and which ones don’t matter. Eg we all have a theory of drunkenness and, a person jumping into a pool at a party fully clothed may fit into the theory of drunkenness.

Question 13

4. How do these theories affect our learning and the inferences that we make?

Answer 13

Categorisation enables the ability to apply general knowledge to new cases you encounter, and draw broad conclusions from your experience.
People were willing to make inferences about typical cases to the whole category but not from atypical cases to the category.

Question 14

5. What are some of the different profiles that people have for different concepts?

Answer 14

People may think about different concepts in different ways. People think about natural things in a pretty stable way, what a bush, water, stone is, but arifacts that are man made, can be what you would like then to be, for example a table could have 15 or 4 legs and be whatever you d like. People reason differently from natural to artifacts as they have different beliefs to why they are the way they are

Question 15

How is knowledge stored? How is it retrieved?

Answer 15

Knowledge is stored through a series of networks, with associative links and connecting nodes, associative links don’t just tie together the various bits of knowledge, they also represent the knowledge

Question 16

Describe Collins and Quillian’s (1969) experiment and what it tells us about the structure of the knowledge network

Answer 16

Tells us that it is quicker to retrieve knowledge involving closely related ideas, it will take more time to retrieve more distant ideas.
Participants were given sentences are asked to note if they were true or false. Eg a cat is a bird they had to traveling through a connection of nodes to find the answer. If two nodes are directly linked then the reaction time is quicker, the more nodes required the slower the time.

Question 17

What are propositional networks

Answer 17

Proposition’s, defined as the smallest units of knowledge that can either be true or false, eg ‘children love candy’ is a proposition but ‘children’ is not.

Question 18

What are the similarities and differences between a network with local representations and a connectionist network?

Answer 18

Local representation is when one node represents one idea, so when that node is activated you think about that idea. Whereas connectionist networks, rely on distributed representation where a pattern of activation occurs across a network and not just one node. Cant find the meaning in one node all the sequenced nodes need to fire to find the meaning

Question 19

What is parallel distributed processing (PDP)?

Answer 19

the step that bring about connectionist network all occur simultaneously with each other so that one entire representation can smoothly trigger the next. Qw q result these models are good at generalizing what they have learned to newer never seen before variations of the pattern.

Question 20

How does learning occur in a connectionist network?

Answer 20

Activation of one pattern = George Washington then simultaneously activation pattern for = president.
This theory learning involves adjustment of connections among nodes, so that after learning, the activation will flow in a way that represents new knowledge.