Psych200 Chapter 9

Conceptual knowledge

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

Concept

a mental representation of a class of objects, events, or ideas, or of an individual item.

Category

 all possible examples or members of a particular concept

Definitional approach to categorization

category membership is determined by whether an item meets a set of necessary and sufficient defining features

Family resemblance

category members share overlapping features, but there may be no single feature common to every member

Prototype approach to categorization

we classify an item by comparing it with a mental representation of the category’s most typical member

Prototype

a mental average or best example of the category’s most typical member, based on commonly shared features

Typicality

how representative or typical an item is of its category

Exemplar approach to categorization

 we classify an item by comparing it with stored examples of category members

Exemplar

a specific remembered example of a category member

Hierarchical organization

organization in which broad categories are divided into increasingly specific subcategories

Global level

the broadest, most general level of categorization

Basic level

the intermediate level that usually provides the most useful everyday information

Specific level

the narrowest, most detailed level of categorization

Semantic network

a model representing concepts as nodes connected by links that show relationships

Node

a representation of a concept in a semantic network

Link/edge

a connection between nodes representing a relationship between concepts

Cognitive economy

storing shared properties once at a general level rather than repeating them at every subordinate category

Spreading activation

activation of one concept that spreads through connected concepts in a network

Priming

facilitation of processing a stimulus because a related stimulus was encountered or activated beforehand

Connectionism

an approach that models cognition using networks of simple, interconnected processing units

Connectionist model/parallel distributed processing model

a network in which many units process information simultaneously and knowledge is represented across patterns of connections.

Input unit

a unit that receives information from the environment or the problem

Hidden unit

an intermediate unit that combines and transforms input before it reaches the output

Output unit

a unit that produces the network’s response or decision

Connection weight

the strength and sign of the influence one unit has on another

Back propagation

a learning procedure that sends output error information backward through a network to adjust connection weights

Graceful degradation

the gradual, rather than all-or-none, loss of performance when part of a distributed system is damaged

Sensory-functional hypothesis

the proposal that living things are identified especially using sensory features, whereas artifacts are identified especially using functional information

Multiple-factor approach

the view that category knowledge depends on multiple kinds of information and their distribution across concepts, not just a living/artifact split

Distributed representation

a representation in which a concept is encoded by a pattern of activity across many units  or brain areas

Crowding

difficulty distinguishing items when they share many features or when representations/features overlap

Semantic category approach

the proposal that some categories may rely on partly specialized neural systems because of their importance or distinct processing demands

Embodied approach

the view that conceptual knowledge is grounded in, and partly reactivates, sensory, motor, and bodily experiences

Semantic somatotopy

a correspondence between the body/action-related brain new areas activated and the body part or action described by a concept

Hub-and-spoke model

 model in which modality-specific systems (“spokes”) interact with a central integrative conceptual system (“hub”)

Anterior temporal lobe (ATL)

a brain region proposed to act as a central hub that integrates information from different modalities into coherent concepts

graceful degradation

A participant responds faster to a target word after seeing a related word than after seeing an unrelated word. Which concept best explains this finding in the semantic network approach?

Which is a criticism of Collins and Quillian's hierarchical model?

It does not adequately explain the typicality effect

the hub-and-spoke model

A person with damage to the anterior temporal lobe (ATL) has difficulty understanding concepts across several categories, while damage to a more specialized region may produce a narrower impairment. Which model is most consistent with this pattern?

The embodied approach, because conceptual knowledge can involve perceptual and action-related systems

A researcher proposes that understanding the concept "hammer" can involve reactivating systems used to perceive and manipulate hammers. Which theoretical approach does this proposal best represent?

In a semantic network, a concept such as "canary" is represented by a node connected to other concepts. What does the connection between canary and bird represent?

A relationship between the concepts represented by the two nodes

spreading activation

A trained connectionist network suffers damage to some of its units, but it can still produce a partially correct response. This is an example of:

Why is the definitional approach difficult to use for some everyday categories?

Many categories do not have a single set of defining features that applies to each member

cognitive economy

In Collins and Quillian's hierarchical model, the property "has roots" is stored at a general node representing plants rather than separately at every individual plant node. What principle does this illustrate?