Chapter 9- Conceptual Knowledge Flashcards

1
Q

Conceptual knowledge

A

Knowledge that enables us to recognize objects and events to make inferences about their properties. When we encounter a new item or event, how do we know what it is? How do we distinguish items from each other?

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Concepts

A

Defined in a number of ways- the mental representation of a category or individual, or categories of objects, events, or ideas, or “glue” that holds our mental world together. Concepts are the form in which knowledge exists that lets you recognize objects or understand what’s going on around you. Describing a cat as an animal that meows is describing an aspect of a concept.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Category

A

Learned organizations or groupings of concepts. Includes all possible examples of a specific concept. The category “cats” includes all of the different types of cats. Mental representations of objects (concepts) provide our rules for how we place things in each category.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Categorization

A

The process by which things are put into categories. This refers to new things, like an unclassified instance or exemplar. Once we recognize something as being part of a category, we know a lot about it “pointers to knowledge”. For example, recognizing an animal as a cat provides a lot of information about its behavior. Categories allow us to take action to interact with our environment. To begin spreading butter on bread, we need to know all the different things we need to do to accomplish this and all the rules about what bread, butter, and knives are like.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

How do categories help us understand behavior?

A

Certain behaviors make sense once we can place a person in a certain category. For example, a man wearing black and gold face paint might not make sense until you realize he’s heading toward the football stadium. Then, you realize he can be categorized as a Pittsburgh Steelers fan.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Classical/definitional approach to categorization

A

States that we can decide whether something is a member of a category by determining whether a particular object meets the definition of the category. For example, a square is defined as a shape with 4 equal sides

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Why don’t definitions work for categories?

A

Definitions work for some things, like geometric shapes. However, not all members of a category will have the same features. A chair might be defined as furniture with a back and arms, but not all chairs will meet that definition.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Family resemblance

A

Wittgenstein- refers to the idea that things in a category resemble one another in different ways. This approach allows for some variation in a category, unlike the dictionary definition. All chairs must resemble the other chairs in the category in some way or share some similar characteristic

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Prototype approach to categorization

A

Membership in a category is determined by comparing the object to a prototype that represents that category.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

For the prototype approach, what determines if something is a typical member of a category?

A

Rosch- stated that the typical prototype is based on the average of members of a category that are commonly experienced. A prototype for bird might look like a combination of birds you typically see, but it doesn’t have to exactly resemble any of them. The ideal category member will serve as a reference point

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Typicality

A

The degree to which a category member resembles the prototype of the category. Differences in typicality are the differences within categories (there are many different types of birds, and some are very different from others). High typicality means that a category member closely resembles the prototype of that category.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Rosch typicality experiment

A

Participants were given a list of items in different categories, and were asked to rate the extent to which each item represented the category. The low rating for sparrow, for example, showed that participants considered sparrows to be good examples of birds, while penguins were considered bad examples.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

How well do good and poor examples of a category compare to other items in the category? (Rosch and Mervis)

A

Participants were given a list of objects and were asked to list as many attributes or characteristics of the object as they could. Similar objects, such as chair and sofa, were given many of the same characteristics, meaning that the family resemblance was high. With objects like mirror and phone, resemblance was low, even though they both can be classified as furniture. It was concluded that there’s a strong relationship between family resemblance and prototypicality- good examples of the category share many attributes with other members of the category.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Sentence verification technique (Smith)

A

Participants were presented with statements and asked to answer yes or no depending on if they thought that the statement was true or not. Participants responded faster for objects that are high in prototypicality than for objects that are low in prototypicality.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Typicality effect

A

The ability to judge highly prototypical objects more rapidly. Typical exemplars also tend to be listed prior to less typical exemplars.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

When participants are asked to name objects in a category, which objects tend to be named first?

A

Prototypical objects are named first. Therefore, a sparrow would be named before a penguin when participants were asked about birds

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

Priming

A

Occurs when presentation of one stimulus facilitates the response to another stimulus that usually follows closely in time, if it contains some of the information needed to respond to the stimulus. Prototypical members of a category are more effected by a priming stimulus.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

Rosch priming experiment

A

Participants heard the prime first, which was the name of a color. They then saw a pair of colors side by side and indicated as quickly as possible whether the colors were the same or different. Colors were either the same and good examples of the category, the same and poor examples (light blue, etc) or different. The same judgements for the prototypical colors were faster than same judgements for non prototypical colors. The prime facilitates the participants’ responses to the prototypical stimulus because when they hear the name of a color, they imagine a prototypical example of that color.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

How did Rosch’s research provide an advance over the definitional approach?

A

It provided a lot of experimental evidence that all items within a category are not the same.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

Exemplar approach to categorization

A

Involves determining whether an object is similar to other stored objects. The standard for this approach involves many exemplars, and each known exemplar acts as a reference point. This approach explains the typicality effect by proposing that objects that are more like exemplars are classified faster. A sparrow is similar to many exemplars, so it’s classified faster because there are more reference points for that item

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

Exemplars

A

Actual members of a category that a person has encountered in the past. If a person has encountered sparrows, robins, and blue jays in the past, they would be exemplars of the bird category.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

Advantages of the exemplar approach

A
  1. It can take atypical cases into account by using real examples. Flightless birds are not typical but can be exemplars of this category, rather than getting lost in the “average of a bird” proposed by the prototypical approach
  2. It can deal more easily with variable categories like games- we only need exemplars of some of the varying examples to recognize them
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
23
Q

Are prototypes or exemplars better for recognizing objects?

A

People might use both approaches. When we initially learn about a category, we might average exemplars into a prototype. Later on, the exemplar information strengthens. We know what cats are (a prototype) but we know our own cat best (an exemplar).

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
24
Q

Hierarchical organization

A

Where general categories are divided into smaller, more specific categories. Some chairs can be categorized as kitchen chairs and others as dining room chairs.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
25
Q

3 levels of categories in hierarchical organization (Rosch)

A
  1. Superordinate level- also known as the global level, “furniture”.
  2. Basic level- “table”
  3. Subordinate level, or the specific level. Example- “kitchen table”
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
26
Q

Rosch basic level

A

Participants were asked to list features of different levels- for furniture, tables, and kitchen tables. They listed only a few features that were common to all furniture, and more that were common to all tables or all kitchen tables. Rosch proposed that going above the basic level (to furniture) results in a loss of information, and going below it (to kitchen table) does not provide a gain in information. This category is most commonly used in language, when naming things

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
27
Q

How do the names of things relate to basic level categories?

A

When asked to name objects, participants often gave their basic level name. For example, “guitar” rather than electric guitar (specific) to musical instrument (global). When participants were shown a category label and then shown a picture, participants identified the object as being part of the category more quickly if they had been given the basic category label.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
28
Q

How does knowledge affect categorization?

A

Nonexperts on a category are likely to describe an object as belonging to the basic category. However, experts on a category (plants, birds) are more likely to assign those objects to a specific category. This is because they have learned to pay attention to features that nonexperts aren’t aware of. To categorize objects, the learning and experience of the people categorizing them matters.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
29
Q

Semantic network approach

A

Proposes that concepts are arranged in networks in the mind.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
30
Q

Collins and Quillian’s semantic network model

A

The network consists of nodes that are connected by links. Each node represents a category or concept, and related concepts are connected by directional associations (links). Properties are indicated for each concept using arrows. The connected related concepts indicate how each concept is related to each other in the mind- canary and bird, and bird and animal, are connected in the mind.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
31
Q

Hierarchical model

A

A model that consists of levels arranged so that more specific concepts, like “canary”, are at the bottom, and general concepts are at higher levels. There are directions to these relationships. This describes Collins and Quillian’s semantic network model.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
32
Q

Cognitive economy

A

Storing properties of an object just once, at a higher level node (in the hierarchical model). This saves space and makes the network more efficient. It does create a problem saying “can fly” at the bird node for example, because not all birds can fly. Exceptions can be added at lower nodes with the unique properties of that node.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
33
Q

What is the purpose of Collins and Quillian’s semantic network model?

A

The model indicate how concepts and properties are associated, and makes predictions about how we retrieve information. It does not correspond with physiology.

34
Q

Collins and Quillian’s semantic network experiment

A

It was predicted that the time is takes for a person to retrieve information about a concept should be determined by the distance a person has to travel through the semantic network. It was found that statements that required traveling a farther distance from a specific node resulted in a longer reaction time.

35
Q

Spreading activation

A

Activity that spreads out along any link that is connected to an activated node. This means that when the “canary-bird” pathway is activated, other concepts connected to “bird” are activated too. The additional concepts that are activated are now primed and can be retrieved more easily from memory.

36
Q

Lexical decision task

A

Participants read stimuli, some are words and some are not words. They indicate as quickly as possible whether the stimulus is a word or nonword.

37
Q

Meyer and Schvaneveldt’s lexical decision task experiment

A

Some word pairs were closely associated (bread and wheat) while others were more weakly associated. The reaction time was faster when the two words were associated. This could have occurred because retrieving one word from memory triggered a spread of activation to other nearby locations in the network

38
Q

Problems with Collins and Quillian’s semantic network model

A

The model can’t explain the typicality effect- the model predicts equally fast reaction times for specific typical and atypical examples. The concept of cognitive economy is also questionable because people might store the properties of concepts right at the relevant node. With the sentence verification task, there were longer reaction times for shorter distances in the network

39
Q

Connectionism

A

An approach to creating computer models for representing cognitive processes. These models are called parallel distributed processing

40
Q

Connectionist network

A

These networks contain circles called units- units are inspired by the neurons found in the brain. The lines are connections that transfer information between units, and they represent axons. Contains input units, output units, and connection weights

41
Q

Input units

A

Units that are activated by stimuli from the environment, like neurons

42
Q

Hidden units

A

Input units send signals to the hidden units, which send signals to output units

43
Q

Connection weights

A

Determines how signals sent from one unit either increase or decrease the activity of the next unit. This corresponds to what happens at the synapse as signals are transmitted between units. Similar neurons firing at a fast rate, high connection weights cause a strong tendency to excite the next unit. Negative weights can decrease excitation or inhibit activation of the receiving unit.

44
Q

Activation of units in a network depends on (2)

A
  1. The signal that originates in the input units
  2. The connection weights throughout the network
45
Q

A stimulus presented to input units is represented by

A

The pattern of activity that is distributed across the other units. This is similar to distributed representation in the brain.

46
Q

How does the learning process occur in a connectionist network

A

Networks must be trained, or activity in one unit will spread out across the whole network. The learning process occurs when the erroneous responses in the property units cause an error signal to be sent through the network through back propagation (signals are sent back through the unit). The error signals sent back to the hidden units and the representation units provide information about how the connection weights should be adjusted so the correct property units will be activated. Patterns will be strengthened over time

47
Q

How are connectionist networks different from Collins and Quillian’s semantic network model?

A

Connectionist networks are created by a learning process that shapes the networks. This means that information about each concept is contained in the distributed pattern of activity across multiple units. In the semantic network model, concepts and the properties of concepts are represented by the activation of different nodes. Representation in a connectionist network is more complex because many units are involved for each concept.

48
Q

Graceful degradation

A

Information in a connectionist network is distributed across many units, so damage to the system doesn’t completely disrupt its function. Damage occurs gradually. This process is called graceful degradation. This is similar to what occurs with brain damage, where only part of the brain stops working. These results support the idea of connectionism.

49
Q

How do connectionist networks explain generalization of learning?

A

Similar concepts have similar patterns. Therefore, training a system to learn the properties of one concept provides information about other concepts that are related (like types of birds). This is similar to how we learn about concepts, because learning about a type of bird (like canaries) lets us predict the properties of other types of birds. This provides evidence for connectionism.

50
Q

Sensory-functional hypothesis

A

States that our ability to differentiate living things and artifacts depends on a memory system that can distinguish sensory information and a system that can distinguish functions.

51
Q

Category-specific memory impairment

A

When patients lose the ability to identify one type of object but are able to identify other types of objects. They might have trouble identifying animals but no trouble identifying nonanimals

52
Q

How did Warrington and Shallice come up with the sensory-functional hypothesis?

A

They stated that living things were usually identified by their sensory features, but artifacts (non living things) were distinguished by their function.

53
Q

Problems with the sensory-functional hypothesis

A

Some patients with brain damage are better at identifying animals than artifacts, which contradicts the hypothesis. Some patients are better at identifying certain types of artifacts than others, meaning that artifacts are not a homogenous category as predicted by the hypothesis

54
Q

Multiple-factor approach

A

Related to the idea of distributed representation, searches for factors beyond sensory and functional that determine how concepts are divided in a category

55
Q

Multiple-factor approach experiment (Hoffman and Ralph)

A

Participants were presented with a list of objects and asked how well they associated the objects with various features. It was found that animals are more highly associated with motion and color, and artifacts are more highly associated with performed actions (consistent with SF hypothesis). However, it was found that mechanical devices overlapped with both animals and artifacts

56
Q

Crowding

A

The fact that animals tend to share many properties (like eyes, legs, ability to move). However, separate categories of artifacts tend to share fewer properties. Some researchers propose that patients don’t actually have a category-specific impairment, they have difficulty distinguishing between similar objects

57
Q

Semantic category approach

A

Proposes that there are specific neural circuits in the brain for some categories. Researchers have proposed that these categories are innately determined because they are important for survival. The approach also states that the brain’s response to items from a specific category is distributed across multiple areas of the cortex- identifying faces depends on the area in the temporal lobe as well as areas identifying emotion and other characteristics

58
Q

Semantic category experiment (Huth)

A

Created a map of the brain based on spoken language using fMRI scans. The maps were very similar for each of the 7 participants.

59
Q

Embodied approach

A

States that our knowledge of concepts is based on reactivation of sensory and motor processes that occur when we interact with the object

60
Q

Embodied approach experiment (Hauk)

A

Measured participants’ brain activity under fMRI for two conditions: as participants moved their foot, index finger, or tongue, and as participants read “action words” (verbs). The activation caused by reading the words occurs in approximately the same area of the brain as doing the action, although doing the action creates more brain activity.

61
Q

Semantic somatotopy

A

The correspondence between words related to specific parts of the body and the location of brain activity.

62
Q

Problems with the embodied approach

A

Some patients are able to easily recognize objects and are just impaired in doing the necessary actions to use the object, in contrast with this approach. This approach also doesn’t explain our knowledge of abstract concepts

63
Q

Semantic dementia

A

Causes a general loss of knowledge for all concepts, people tend to have equal trouble with living things and artifacts. Usually caused by anterior temporal lobe damage.

64
Q

Hub and spoke model of semantic knowledge

A

States that areas of the brain that are associated with specific functions are connected to the ATL, which acts like the hub to integrate information for these areas. Evidence- damage to one of the specialized brain areas (the spokes) causes specific deficits, but damage to the hub causes general deficits, like with semantic dementia.

65
Q

Transcranial magnetic stimulation

A

A pulsating magnetic field is placed over a person’s skull, and brain function is temporarily disrupted in that area. The disrupted area of the brain is likely involved in whatever behavior is disrupted. Experiments using this technique have been used to provide evidence for the hub and spoke model.

66
Q

Forms that categories can take (5)

A
  1. attribute / entity categories
  2. relational categories
  3. ad hoc categories
  4. abstract coherent categories
  5. thematic & taxonomic categories
67
Q

Attribute / entity categories

A

All members in a category look similar

68
Q

Relational categories

A

Objects used for similar activities, for example, glass and water would be in the same category

69
Q

Ad hoc categories

A

“After the fact”. Some categories are defined based on how you’re going to achieve your current goal. “What things do I want to sell in a garage sale?”- you would sort things based on what you do or don’t want to cell

70
Q

Categorization frameworks (Medin)

A

Participants freely sorted exemplars (alien figures) into categories. Most participants preferred to sort the aliens by only one attribute. This exposes an issue with the family resemblances idea. Participants usually won’t build categories through family resemblances when a simpler approach is possible.

71
Q

Issues with the prototype approach to categorization (4)

A
  1. How can you categorize something you have never seen before?
  2. Where is the edge or boundary of the category?
  3. Some categories, like games, do not seem to possess a prototypical structure due to such high variability.
  4. Surrounding context can influence the way certain items are classified. Prototypes might change in conjunction with context shifts
72
Q

Issues with the exemplar approach

A
  1. A lot of load on memory
  2. What about when items can be categorized at multiple levels?
  3. Abstracted (superordinate level) categorizations should be effortful, but people are extremely fast at identification
  4. Some categories are easier to learn than others
73
Q

Shepard’s law of generalization/similarity

A

For any two items in “psychological space”, as the distance between them increases, the perceived similarity between them decreases exponentially. This decay in similarity forms a gradient. Objects with greater similarity are more likely to be classified in the same category and are therefore more confusable.

74
Q

Within-category compression

A

When members of the same category come to look more similar after a category has been learned. The perceptual difference is smaller within categories

75
Q

Between-category expansion

A

When members of different categories tend to look more dissimilar after a category has been learned. Perceptual/psychological difference is larger between categories

76
Q

How are speech categories an example of between-category expansion and within-category compression?

A

The da and ta sounds began sounding more dissimilar at a lower time and a higher time. However, there was a period of time in the middle where the sounds were very similar and difficult to distinguish

77
Q

How are concepts/categories organized in the mind?

A

Like a network. Diversity in language can be organized in a network, with more similar languages being grouped together

78
Q

2 approaches to develop a network in the mind

A
  1. Semantic network
  2. Connectionism
79
Q

How can we test the semantic network approach?

A

Using the sentence verification task. Judgements with longer distances should take a longer time to respond, because you need to travel past more nodes.

80
Q

Brain basis of concept representation (4)

A
  1. Semantic Category Approach
  2. Embodied Approach
  3. Sensory-Functional Hypothesis
  4. Multiple-Factor Approach