Cognitive Psychology: Concepts and Generic Knowledge

Which of the following benefits does a hierarchical network provide?

It is efficient because information is stored only once

In a PDP model, learning can happen in all of the following ways EXCEPT

random, synchronous firing of nodes leads to a weakening of a connection

Anomia

An inability to name common objects. Damage in the grain's temporal pole had difficult naming persons but were daily able to name animals and tools. Patients with damage in the inferotemporal region had difficulty naming persons and animals but did somewhat better naming tools. Patients with damage in the lateral occipital region had difficulty naming tools but did reasonably well naming animals and persons.

Basic level categories

Are usually represented in our language via a single word, wile more specific categories are identified only via a phrase. The odds are good that you would answer by saying "It's a chair", using the basic-level description, rather than the superordinate label (It's a piece of furniture") or a more specific description ("It's an upholstered armchair"), even though these other descriptions would certainly be correct. Rosch and others have argued that there is a "natural" level of categorisation, neither too specific nor too general, that people tend to use in their conversations and their reasoning.

Propositions (Anderson model)

At the centre of conception is the idea of proposition, defined in this setting as the smallest units of knowledge that can be either true or false. e.g. Children love candy. Propositions are easily represented as sentences, but this is merely a convenience. Propositions can also be represented in various nonlinguistic forms, including a structure of nodes and linkages. Individual ideas are represented with local representations: Each node represents one idea, so that when that node is activated, you're thinking about that idea, that node is activated.

It has been suggested that a rigid definition for a category is not possible and that resemblance (much like a family resemblance)may be more appropriate. Why is this the case?

Categorization is very often a matter of degree, not an all-or-none process

Resemblance

Does depend on shared properties, but more precisely resemblance depends on whether the objects share import an, essential properties. On this basis, you regard plums and lawn mowers as different from each other because the features they share are trivial or inconsequential. Your decisions about which features are essential (cost or weight) vary from category to category and they vary in particular according to your reliefs about what matters for that category. BLUE GNU is an example where you've decided that colour isn't a critical feature, and you categorise despite the contrast on this dimension. But you also know that colour does matter for other categories, and so you know that something's off if a jeweller tries to sell you a green ruby. Whether in judging resemblance or in deciding on categorisation, the features that you consider depend on the specific category.

Explanatory Theories

Each of us has something that we can think of as a "theory". These theories are less precise, less elaborate, than a scientist's theory, but they serve the same function: They provide a crucial knowledge base that we rely on in our thinking about an object, event, or category; and they enable us to understand new facts we might encounter about the object or category. In the case that someone jumps in the pool fully clothed at a party. Each of pus has certain beliefs about how drunks behave; we have, in essence, a "theory" of drunkenness. This theory enables us to think through what being drunk will cause someone to do and not to do, and on this basis we would decide that yes, someone who jumped into the pool fully clothed probably was inebriated.

Concepts

Even simple terms, denoting concepts we use easily and often, resist being defined. In each case, we can come up with what seems to be a plausible definition, but then it's easy to find exceptions to it. For example what is a "dog"? Most people include has fur ind the definition, but what about the hairless Chihuahua? For most concepts, definitions are not available. For many purposes, though, you don't need a definition and can rely instead on a mix of prototypes and exemplars. More-typical category members are being privileged.

Graded membership

Objects closer to the prototype are "better" members of the category than objects farther form the prototype.

Sentence verification task

Participants are presented with a succession of sentences; their job is to indicate (by pressing the appropriate button) whether each sentence is true or false. In this procedure, participants' response is slower for sentences like "A penguin is a bird" than for sentences like "A robin is a bird". When there is a close similarity between the test case and the prototype, participants can make their decisions quickly; judgements about items more distant for the prototype take more time.

Inferences Based on Theories

People are willing to make inferences from the typical case to the whole category, but not from an atypical case to the category. Participants that were told a new fact about robins were also infer to ducks but not in reverse. Also find it easy to infer that property in gazelle's blood can be transmitted to lions via the food chain, relying on their beliefs about how these concepts are related to each other.

Connectionist networks

Rely on distributed representation in which each idea is represented, not by a certain set of nodes, but instead by means of a specific pattern of activation across the network. Must use processes that are similarly distributed, so that one widespread activation pattern can have broad enough effects of evoke a different but equally widespread pattern. The steps bringing this about must all occur simultaneously in parallel. PDP models have a excellent capacity for how the pattern is implemented. Thus, the models are impressively able to generalise what they have learned to new, never-seen- before variations on the pattern. Learning must involve some sort of adjustments of the connections among nodes, so that after learning, activation will flow in a fashion that can represent the newly gained knowledge.

Categorization is plainly influenced by judgements about resemblance, but it is also influenced by factors other than resemblance. Which of the following is NOT part of the evidence indicating the importance of these other factors?

Resemblance plays a key role in the use of prototypes but plays no role in the well-documented use of exemplars.

Maxine has sustained brain damage to her left temporal lobe, which influences her ability to categorise efficiently. Which of the following is most likely to describe the problems that Maxine will face?

She will lose the ability to discriminate some categories but others will remain unaffected.

A lemon that has been painted red, whit, and blue and then run over by a car is still likely to be categorised as a lemon. Which of the following is NOT an accurate description of why this might be?

Superficial things like colour do not play a role in categorisation

Exemplar-based reasoning

That some cases categorisation relies on knowledge about specific category members rather than the prototype. The standard is provided by whatever example of the category comes to mind and, of course, different examples may come to mind on different occasions). You assess the similarity between a candidate object and the standard. If the resemblance is great, you judge the candidate as being within the relevant category; if the resemblance is minimal, you seek some alternative categorisation.

Prototype theory

The best way to identify a category is to specify the "centre" of the category, rather than the boundaries. This ideal prototype will be an average of the various category members you've encountered. Thus the prototype dog will be the average colour of the dogs you've seen, the average size of the dogs you've seen. Different people each with their own experiences, will have slightly different prototypes.

Family resemblance

There is probably no "defining features" for your family - features that every family member has. Nonetheless, there are features that are common in the family, if we consider family members two or even three at a time, w can usually find some shared attributes. Many concepts have the same character, with many features shared among the instances of the concept, but no features shared by all of the instances. And the more of these features an object has, the more likely you are to believe it is the category. It is a matter of degree, not all or none.

Judgments about which category members are typical

are daily shifted by changes in context or changes in perspective

On study found that if participants were told a new fact about robins, they would also believe that the new fact was true of ducks. However, if told a new fact about ducks, participants would not extrapolate this information to robins. This suggests that

participants are willing to apply inferences from atypical case within a category to the whole category but will not apply inferences from an atypical case to the whole category