General Knowledge

General Knowledge Organization

Introduction

New topic: What we do with information once we have it.
Topics:
- General knowledge.
- Thinking and reasoning.
- Problem-solving and creativity.
Focus: How we organize information in our minds.

Semantic Knowledge

Semantics: Meaning.
Categories and Concepts:
- Categories: How we organize information.
- Concepts: Mental representation of categories.
Prototypes: Canonical, obvious examples of a category.
Parallel Distributed Processing: Representing knowledge across neuron-type units.
Schemas and Scripts: Efficient ways to represent knowledge.

Semantic Memory

Definition: Representation of meaning; organized knowledge about the world.
Vocabulary:
- English-speaking adults: Meanings of 20,000 - 100,000 words.
  Note: Uni students likely at the top end due to constant learning of jargon and new words.

Types of Knowledge in Semantic Memory

General Knowledge

Definition: Facts tested in quizzes (e.g., pub quizzes, Trivial Pursuit).
Examples:
- $π = 3.1416$
- Kim and Kanye's first child is named North.

Lexical Knowledge

Definition: Representation of information about words.
Example: "Spirit" can relate to ghost, personal feelings, or alcohol.

Conceptual Knowledge

Definition: Organizing information to make storage efficient; knowledge of concepts.
Example: Knowledge of mammals includes that they are warm-blooded.

Concepts

Definition: Mental representation of a category.
Example: The concept of fruit includes all known fruits.
Abstract: Includes personal experiences and opinions about the category (e.g., fruit = yum or yuck).
Efficiency: Stores information about many objects within a single word concept.

Prototypes

Definition: the most obvious, easily recalled member of a category.
Example: Sparrow as a prototypical bird.
Less Prototypical Examples: Desert owls, penguins, flamingos.
*Emu as a non-prototypical bird.

Prototype Approach

Some members of categories are more obvious than others.
Examples: Apples, bananas, and oranges are prototypical fruits in Australia.
Prototypes: Idealized, representative members of a category.

Priming Studies

How exposure changes response.
Prototypical Fruits: Presenting bananas & apples facilitates responses to oranges as fruits.
Non-Prototypical Fruits: Custard apple, dragon fruit, papaya.
Tomatoes & Avocados: Technically fruits, but often categorized as vegetables.

Alternative Theories

Necessary and Sufficient Features: All category members are equal if they possess defining features.
Example: Bachelor = male and unmarried.
Prototype Approach: Some members are more prototypical (e.g., James Bond is a more prototypical bachelor than a young boy or the Pope).

Prototype Research

Graded structure confirms that category members are not created equal.
Representative Members: Are easily recalled and quickly responded to in tasks.
Minor Bird: A prototypical bird in Australia.
Puffin: A non-prototypical bird.

Concepts and Categories

Overlapping definitions.
A category is a group of objects.
A concept is your mental representation of a category.

Examples & Judgments

Common responses are typically prototypical.
Example: If asked to name a bird, people commonly say "robin".
Judgement speed: Faster to identify prototypes as belonging to a category.
Example: "A robin is a bird" is verified faster than "A penguin is a bird".
Prototypes are processed fluently and easily.

Semantic Priming

Presenting related items to see the effects on each other.
Closely related items, especially prototypes, yield faster responses.

Color Representation

Prototypical colors have prototypic representations.
Roch (1970s) conducted semantic priming experiments using color patches.
Task involved identifying if paired colors were the same or different after a color name prime.
Prototypical colors are recognized faster when primed.

Color Patch Experiment

Two patches of slightly different shades of red were shown after the word red was presented.
Participants had to determine if they were the same shade or different shades.
The first example closer to the prototypical red and the second more of a brick red.
Priming with the name of a color worked better when it was a prototypical color.
Semantic Priming only works for prototypical colours

Family Resemblance

No single attribute is shared by every category member.
Examples share at least one attribute in common.
Study: Asked participants to rate transport types for prototypicality.
Prototypical Items: Car, bus, bicycle share the most attributes with other members.

Category levels

Superordinate Level: Highest, most general (e.g., animal, vegetable, transport).
Basic Level: Moderate specificity (e.g., cat, car, chair).
Subordinate Level: Very specific (e.g., British Shorthair, Mazda 3).

Usage & Communication

Basic level categories are common in descriptions.
Example: "Look at my cat" rather than "Look at my animal" or "Look at my British Shorthair".
Basic level names identified more quickly.

Prototype vs. Basic Level

Prototype: Best example of a category.
Basic Level: Moderately specific; Goldilocks zone.
Linguistic Links : Categories are linked with linguistics.
Knowledge representation differs with language.
Japanese group bottles, trees and cups together while English speakers do not.

Prototype Always at Basic Level

Prototypes can exist in any category
Superordinate and subordinate categories can both still have prototypes

Network Approach

Emphasis: Interconnections between items.
Word Association Test: Reveals connections between concepts (e.g., cow → farm → milk).
Semantic Networks: Representation of how knowledge differs among individuals.

Semantic Network Model

Collins and Quillian: Developed in the 1960s.
Hierarchical Structure: Superordinate categories (animal) → basic categories (bird, fish) → subordinate categories (canary, ostrich, shark, salmon).
Monotreme: An example of a category tree
Economical Information Storage: Storing information at a higher level avoids repetition.

Nodes & Relationships

Nodes: Represent concepts; linked with relationships.
Arrows: Indicate relationships (e.g., "is a" relationship).
Attributes: Stored at the lowest level they are true for all lower nodes.

Collins and Quillian's Hierarchy

A canary is a type of bird, that is a type of animal
Each arrow indicates a type of relationship
However the researchers made a mistake an ostrich says it cannot fly even though the bird says it can fly

Sentence Verification

Checking whether sentences are true or false.

Types of Sentences Used

A robin is a robin: baseline reaction time (Tautological).
A robin is a bird: verified quickly.
A robin is an animal: takes longer.
A robin is a fish: marked as false.

Time & Response

Hierarchical processing: responds to bird faster than animal.
Researchers will use the baseline reaction of a robin is a robin to determine a baseline of how fast that person responds. The older you get the slower you get.
These nodes could represent groups of neurons not just a single neuron.

Problems with Collins and Quillian's Model

Hierarchy Limitation: Not everything fits neatly into a hierarchical structure (e.g., truth, love, justice).
Confound: Frequency of co-occurrence affects responses.
Incorrect Predictions: Faster to verify "a dog is an animal" than "a dog is a mammal", despite hierarchy.

Parallel Distributed Processing (PDP)

Alternative to Collins and Quillian's Model.
Spreading Activation: Activation of one concept spreads to related concepts.
Nodes: Each is a node concept, these nodes also have links with each other.
Activation spreads when the nodes are triggered.
Connections in the network are weighted and have experience.
Will inhibit or activate other units and that explains the semantic priming.
Hierarchical has simply over dimensions of activation model and could represent mirror neurons.