C9- Concepts and Categories

Cognition: Concepts and Generic Knowledge

Lecture Outline

  • Overview of chapter focused on concepts and generic knowledge.
    • Understanding Concepts
    • Prototypes and Typicality Effects
    • Exemplars
    • Difficulties with Categorization via Resemblance
    • Concepts as a Set of Conceptual Knowledge
    • The Knowledge Network
    • Integrating Concepts

Inferences Based on Theories

  • Role of Categorization:

    • Helps in applying general knowledge to novel cases.
    • Allows for drawing broad conclusions from previous experiences, facilitating learning and reasoning.
  • Category-Based Inferences:

    • Influenced by the typicality effect, which posits that it is more likely to draw inferences from typical cases as opposed to atypical cases.
    • Example: An inference about robins being a type of bird can easily extend to ducks, but not vice versa.
  • Theories interact with the conceptual knowledge regarding related concepts, enhancing our understanding of the world around us.

Different Profiles for Different Concepts

  • Natural Kinds vs. Artifacts:

    • Natural Kinds:
    • Examples include organisms such as skunks and raccoons.
    • They possess essential properties that are relatively stable over time.
    • Artifacts:
    • Examples include manufactured objects like toasters and coffee pots.
    • Their categorization does not rely on intrinsic properties but more on functionality and cultural context.
  • Forming Categories Based on Different Reasons:

    • Categories can arise from various criteria, including:
    • Features: Based on shared attributes or properties.
    • Goal-derived categories: Such as classifying items for a specific purpose (e.g., exercise equipment).
    • Relational categories: Based on relationships between items (e.g., hunting).
    • Event categories: Based on specific events or timelines (e.g., dates).

Knowledge Network

  • Knowledge representation takes the form of a complex web of associations connecting all relevant pieces of information, indicating that:
    • Concepts are dependent on intertwined beliefs and requisite background knowledge.
Evidence for Knowledge Representation in a Network
  • Derived from the sentence verification task, where participants are required to quickly evaluate the truthfulness of various statements:

    • Examples of Statements:
    • "Robins are birds."
    • "Robins are animals."
    • "Cats have hearts."
    • "Cats are birds."
  • Processing Links:

    • Some statements require multiple connections (e.g., "Cats have hearts" requires two links), whereas others may require just one (e.g., "Cats have claws").
Visualization of Knowledge Network
  • Figure 9.5 illustrates the interconnected relationships among concepts, showcasing how overarching categories (e.g., BIRDS and ANIMALS) link to specific examples (e.g., ROBIN and CAT).
  • Response Time Analysis:
    • Mean response time can be influenced by the number of link traversals required, as demonstrated in the specific data points regarding response times at various traversal levels.

Propositional Networks

  • Definition:
    • Propositions represent the smallest unit of knowledge that can be assigned a truth value (true or false).
  • Network Structure:
    • Nodes in propositional networks represent concepts, while the links signify relationships between these concepts, allowing for complex knowledge constructs.
Problems in Sentence Verification
  • Problem #1:

    • Responses are generally faster when the sentence relates to a prototypical stimulus (e.g., "A canary is a bird" is quicker compared to "An ostrich is a bird").
  • Problem #2:

    • The principle of "nonredundancy" does not consistently apply; for instance:
    • Comparing the sentences "Birds have feathers" and "Peacocks have feathers" can yield differing verification times.

Distributed Processing

  • Local Representations:
    • In propositional networks, each node represents a single concept or idea.
  • Distributed Representations:
    • In connectionist networks, concepts are represented by a pattern of activation throughout the network.
    • Parallel Distributed Processing (PDP): Theoretical framework supporting the distributed nature of cognitive processing across the network.

Putting the Pieces Together

  • Complexity of Concepts:
    • Concepts play a pivotal role in human reasoning, though they involve intricate thought processes.
    • Utilization of prototypes and exemplars is frequent, especially when fast associations or classifications are necessary.
    • Theories of concepts are often structured through networks of interrelated knowledge, indicating that beliefs contribute to overall cognition and influence categorization and reasoning.

Chapter Questions

  • Question 1:

    • According to Wittgenstein:
    1. We do not possess a real general concept for each known category; instead, category members are learned individually.
    2. We assess memberships based on family resemblance probabilistically.
    3. Rigid features defining a category can only be identified post extensive study.
    4. Prototypical members form the basis for comparison with other potential category members.
  • Question 2:

    • Correct Statement:
    • b) Just because people use prototypes does not mean it is their only source of information.
  • Question 3:

    • In a production task, the __________ category members mentioned yield the slowest reaction times in sentence verification:
    • a) first
  • Question 4:

    • Categorization based on similarity to stored instances is termed __________ theory:
    • d) exemplar
  • Question 5:

    • Which statement about category knowledge is NOT true?
    • b) Judgments of resemblance will override our beliefs about categories.

Credits

  • Concludes the Lecture PowerPoint presentation for Chapter 9 of Cognition.
  • For additional resources, visit: https://digital.wwnorton.com/cognition8
  • Copyright © 2022 W. W. Norton & Company.