Chapter 9: Knowledge

Why does the definitional approach to categorization initially seem like a good way of thinking about categories, but then become troublesome when we consider the kinds of objects that can make up a category?

(+) Works for some things, such as geometric objects. (-) Doesn't work for natural objects (i.e. birds, trees, and plants, games). The problem is that not all of the members of everyday categories have the same features.

Activation of units in a network depend on what two things

(1) the signal that originates in the input units (2) the connection weights throughout the network.

Why is the use of categories so important for our day-to-day functioning?

1. Allows us to understand what things are and their functions when part of the bigger picture. 2. We can explain what makes a certain thing different from others in a category.

How does the exemplar approach to categorization differ from the prototype approach, and how might the two approaches work together?

1. People use both approaches to categorization. 2. Prototypes may be more important as people initially learn about categories; later, exemplar information may become more important. 3. Exemplars may work best for small categories (such as U.S. presidents), and prototypes may work best for larger categories (such as birds).

Semantic network approach

1. Proposes that concepts are arranged in networks that represent the way concepts are organized in the mind (hierarchically) 2 Collins and Quillian's model is a network that consists of nodes connected by links. 3. Concepts and properties of concepts are located at the nodes. Properties that hold for most members of a concept are stored at higher-level nodes. This is called cognitive economy.

What experiments did Rosch do that demonstrated connections between prototypicality and behavior?

1. Quantified prototypicality 2. Presented subjects with "bird" or "furniture" and a list of about 50 members of the category. 3. Subjects then rate the extent to which each member represented the category title on a scale. 4. Result: There is a strong relationship between family resemblance and prototypicality.

Describe the example of connectionist network with canaries

1. Start with initially weak and undifferentiated activation of property units, with many errors (i.e. input "canary" causes activation of the property unit tall) 2. Error signals sent back, adjusts connection weights 3. Small adjustments in connection weights x repetition = correct properties assigned to canary.

List the four approaches to explaining how concepts are represented in the brain.

1. sensory-functional hypothesis 2. semantic category approach 3. multiple-factor approach 4. embodied approach.

cognitive economy

A feature of some semantic network models in which properties of a category that are shared by many members of a category are stored at a higher level node in the network. For example, the property "can fly" would be stored at the node for "bird" rather than at the node for "canary."

How are connection weights adjusted?

A learning process that involves 1. erroneous responses in the property units cause an error signal to be sent back through the network, by a process called back propagation. 2. This provides info s.t. the correct property units will be activated.

back propagation

A process by which learning can occur in a connectionist network, in which an error signal is transmitted backward through the network. This backward-transmitted error signal provides the information needed to adjust the weights in the network to achieve the correct output signal for a stimulus.

sentence verification technique

A technique in which the participant is asked to indicate whether a particular sentence is true or false. For example, sentences like "An apple is a fruit" have been used in studies on categorization.

categories

All possible examples of a concept. Once you know something is in a category, you know a lot of general things about it and can focus your energy on specifying what is special about this particular object. (i.e. cats include Persian cats, siamese cats, etc.)

exemplar

An actual member of a category that a person has encountered in the past.

connection weight

Determines how signals sent from one unit either increase or decrease the activity of the next unit (modulator). High connection weights result in a strong tendency to excite the next unit, lower weights cause less excitation, and negative weights can decrease excitation or inhibit activation of the receiving unit.

error signal

During learning in a connectionist network, the difference between the output signal generated by a particular stimulus and the output that actually represents that stimulus.

What did Wittgenstein propose to deal with the problem of the definitional approach to categorization?

Family resemblance; the idea that things in a particular category resemble one another in a number of ways and allows for some variation within a category (i.e. chairs)

What arguments did Rosch present to support the idea that one of these levels (different levels within a category) is "privileged"?

Indicated that a basic level of categories (such as guitar, as opposed to musical instrument or rock guitar) is a "privileged" level of categorization that reflects people's everyday experience.

What is true of high-prototypical objects (Rosch and Mervis)?

It means that a category member closely resembles the category prototype (it is like a "typical" member of the category) (a) They have high family resemblance (b) statements about them are verified rapidly (proven via sentence verification technique) (c) they are named first (d) they are affected more by priming; the results of the priming experiments support the idea that subjects create images of prototypes in response to color names.

Prototype approach

Membership in a category is determined by comparing the object to a prototype that represents the category, called a prototype. For example, the prototype for the category "birds" might be based on some of the birds you usually see, such as sparrows, robins, and blue jays, but doesn't necessarily look exactly like any one of them

concepts

Mental representations of a class or individual and the meaning of objects, events, and abstract ideas. They provide the rules for creating categories.

priming

Presentation of one stimulus facilitates the response to another stimulus that usually follows closely in time

How has research on categorization by experts led to modifications of Rosch's ideas about which category is "basic" or "privileged"?

Show that the basic level of categorization can depend on a person's degree of expertise.

Indicate the basic idea behind the embodied approach and the evidence for and against each one.

States that our knowledge of concepts is based on reactivation of sensory and motor processes that occur when we interact with the object. For example, when a person uses a hammer, sensory areas are activated in response to the hammer's, size, shape, and color, and, in addition, motor areas are activated that are involved in carrying out actions involved in using a hammer. For: Mirror neurons, Against; isn't well suited to explain our knowledge of abstract concepts such as "democracy" or "truth".

Indicate the basic idea behind the multiple factor approach and the evidence for and against each one.

The idea of distributed representation focusing on searching for more factors that determine how concepts are divided up within a category. For: Mechanical devices have a widely distributed semantic representation that includes regions important for the representation of both living things and artifacts. Because animals tend to be more similar then artifacts, these patients find animals harder to recognize Against:

hierarchical organization

The kind of organization in which larger, more general categories are divided into smaller, more specific categories

categorization

The process by which things are placed in categories.

Indicate the basic idea behind the semantic category approach and the evidence for and against each one.

The semantic category approach proposes that there are specific neural circuits in the brain for some specific categories. For: There is a genetic basis for the mechanisms that support face recognition, Against:

What did Smith and Coworkers find about the sentence verification technique?

The typicality effect; The ability to judge highly prototypical objects more rapidly

prototype

a typical member of a category, formed by averaging category members a person has encountered in the past.

input unit

activated by stimuli from the environment. They are represented by the pattern of activity that is distributed across the other units.

spreading activation

activity that spreads out along any link that is connected to an activated node. Anything part of it and especially when primed, is easier to retrieve.

connectionism

an approach to creating computer models for representing cognitive processes.

parallel distributed processing model

concepts are represented by activity that is distributed across a network.

Prototypicality

describes how well an object resembles the prototype of a particular category.

unit (connectionist)

inspired by neurons found in the brain.

examplar approach to categorization

involves determining whether an object is similar to other objects. For example, if a person has encountered sparrows, robins, and blue jays in the past, each of these would be an exemplar for the category "birds." (+) 1. Doesn't discard information about atypical cases within a category, such as penguin in the "bird" category. 2. The exemplar approach can also deal more easily with categories that contain widely varying members, such as games. 3. Explains the typicality effect; objects more similar to exemplars are judged faster.

conceptual knowledge

knowledge that enables us to recognize objects and events and to make inferences about their properties

Describe how a connectionist network learns, considering specifically how connection weights are adjusted. Also consider how the way information is represented in a connectionist network differs from the way it is represented in a semantic network.

learn from (1) correct distributed pattern for a particular concept (2) through a gradual learning process that involves (3) adjusting the weights that determine how activation is transferred from one unit to another. (4) beings with incorrect connection weights, which are slowly modified in response to error signals (gradual).

Connectionist network

proposes that concepts are represented in networks that consist of input units, hidden units, and output units, and that information about concepts is represented in these networks by a distributed activation of these units. This approach is also called the parallel distributed processing (PDP) approach 1. Units are like that of neurons 2. Lines are where information is transmitted, representing axons.

output unit

receives input from hidden units

hidden unit

receives signals from input units

crowding

refers to the fact that animals tend to share many properties

Indicate the basic idea behind the sensory-functional hypothesis and the evidence for and against each one.

states that our ability to differentiate living things and artifacts depends on a semantic memory system that distinguishes sensory attributes and a system that distinguishes function. For: Explains Warrington's and Shallice's patients with a category-specific memory impairment; lost ability to identify one type of object but retained the ability to identify other types of objects. Against: Some patients who couldn't recognize artifacts have impaired sensory ability or some could couldn't identify living things but had both sensory and functionally impaired abilities.

What is the evidence for and against the Collins and Quillian model?

supported by the results of experiments using the sentence verification technique. The spreading activation feature of the model is supported by priming experiments. criticized for several reasons: It can't explain the typicality effect, the idea of cognitive economy doesn't always hold, and it can't explain all results of sentence verification experiments.

definitional approach to categorization

the idea that we can decide whether something is a member of a category by determining whether the object meets the definition of the category

basic level

the middle level, and often the first level learned, within a category hierarchy, such as "dog" in the animal/dog/poodle example

global (superordinate) level

the most general level within a category hierarchy, such as "animal" in the animal/dog/poodle example

specific (subordinate level)

the most specific level within a category hierarchy, such as "poodle" in the animal/dog/poodle example

What does it mean to say that there are different levels within a category?

via hierarchical organization. The following are the levels; 1. global (superordinate) 2. basic 3. specific (subordinate)