Untitled Quiz

Introduction to Cognitive Science

• Course title: KMF 1014 Introduction to Cognitive Science
• Unit: 2 - Part 1

Cognition

• Cognition (Latin): acquiring knowledge through mental powers (knowing)
• All processes involving sensory input:
  • Transformation
  • Reduction
  • Elaboration
  • Storage
  • Retrieval
  • Use

Pre-Scientific Experimentation

• Introspection: understanding mental processes
• How we think
• How we think about thinking
• Cognitive psychologists look for explanations for mental functions.

Cognitive Processes: Underlying Mechanisms

• Learning
• Reasoning
• Language
• Perception
• Imagery
• Memory
• Problem Solving

Cognitive Capabilities

• Obtaining sensory information (visual, auditory, perceptual illusions)
• Storing information as representations:
  • Concepts
  • Propositions
  • Rules
  • Analogies
• Utilizing stored information for day-to-day activities
  • Comparing similar situations.
  • Specifying relations between propositions.

The Psychological Approach (Cognitive Psychology)

• Emphasis on internal mental operations
• Computer as a metaphor for MIND
• Mental functioning is representation and computation
• Tasks for Cognitive Psychologists:
  • Developing theoretical frameworks to explain brain workings
  • Devising experiments to illuminate frameworks

Computer = Mind?

• IP Model = Computer Model
• SR (Sensory Register)
• STM (Short-Term Memory)
• LTM (Long-Term Memory)
• Input Devices
• Central Processing Unit
• Hard Drive Storage

Knowledge in LTM

• Retrieval from LTM
  • Knowing "what" (general knowledge, semantic, episodic memory, implicit memory—concepts, proposals schemas)
  • Knowing "how" (procedural knowledge, automated, scripts)

Have you ever experienced these?

• Trying to remember something misplaced
• What thought processes occurred for searching?
• Trying to recall a person's name

Reflection - How do we organize our experience of the world?

• Concepts, notions, ideas of things

Concept?

• Idea about something providing a means of understanding the world (mental representation)
• Often captured in a single word
• Related to other concepts, derived from specific instances
• Mastering a concept involves categorizing objects and events of a specific domain
• Coherence (consistency)

Categories?

• Set of objects belonging together
• A class of similar things
• Enables predictions and organizes aspects of equivalence with other concepts
• Based on common features or similarity to a prototype

Concept? Category?

• Category - Food (Protein)
• Concept - Different kinds of fish
  • White fish (cod, haddock, plaice)
  • Oily Fish (salmon, trout, herring, eels)
  • Shellfish (lobsters, prawns, crabs)

Concepts & Categories: Connection?

• Central to how humans represent knowledge
• Concept is a mental representation of a category
• Category is a set of objects belonging together; each member is a concept.
• Basic process for organizing and functioning in the world

Reflection - Organize classmates into categories

• Reflect on the categories and their basis of organization

Recap: Concept & Category

• Concept: mental idea or notion (mental representation of a class or individual, psychological state signifying thoughts of concepts).
• Category: Set of objects belonging together (all real-world entities appropriately categorized as concepts)

Mechanism for Concept & Category Formation?

• Is the ability to categorize innate?
  • Infants' ability to distinguish perceptions suggests innate categorization ability.
• Basic types of categories
  • Sensory inputs (Perceptual, taste, sound, etc.)
  • Abstractions (Emotions, other kinds of abstractions: intelligence/stupidity, democracy/dictatorship, melodic/atonal)

Categorization: Human Cognition

• Categorization is fundamental to human cognition.
• Stimuli properties are involved in categorization.
• Sorting/classifying is to keep order

Categorization Study: Approaches

• Key Question: What criteria do we use to assign concepts to a category?
• Theoretical Approaches to Categorization:
  • Exemplar Approach
  • Feature Approach
  • Prototype Approach

The 'Exemplar' Approach

• Exemplar = every instance of a category, stored in memory.
• Stores all encountered category members from past experiences.
• Classification based on similarity to stored exemplars (faster classification of similar objects).

Exemplar Approach

• Categorization based on past experiences.
• Developed throughout childhood and beyond.
• We construct categories based on experience

Feature Approach

• Form categories by specifying necessary and sufficient features for membership.
• Entity qualifies as a member if it possesses these features.
• Defining characteristics of a category.

Feature Approach

• Categorization based on necessary and sufficient characteristics.
• Example: Dogs – size, color of ears, animate, four legs, tail, bark, furry

Prototype Approach

• Representation formed by average values of features
• The best or most representative member of a category.
• New objects compared to prototypes for categorization.
• Fast categorization due to similarity to prototypes.

Prototype Approach

• "Average" of that type
• Fits our expectations.
• Example: Bird—Canary VS Ostrich

Summary

• No single approach for categorization is sufficient.
• We use a combination of strategies

Categories: Experimental Evidence

• Perceptual categories, studies of color categories (Sapir-Whorf Hypothesis).

Experiment (by Berlin & Kay, 1969)

• Researchers presented participants with color chips (violet to red).
• Participants categorized the colors.
• Identified boundary chips.

Categories: Experimental Evidence

• Findings: 11 basic color categories exist in all languages (disproving Sapir-Whorf Hypothesis)
• If there are fewer than 11 basic color categories, strict limitations on the categories exist.
• Standard terms for white and black are universal
• Third color term will be red; fourth is green or yellow

Categories: Experimental Evidence

• Categorization using hierarchy (superordinate, basic, subordinate level: Example fruit hierarchy)
• Organization based on free sorting using the principle of family resemblance.
• When people confront the world they find two types of categories: already constructed and must construct themselves.

Reflection: What approach did you use to sort concepts?

• Assess the approach used during sorting concepts.

Category Loss

• Difficulty recognizing and describing stimuli due to brain injury (e.g. Agnosia, Prosopagnosia)

Recap

• Human acquire concepts and engage in categorization process.
• Concepts and categories are intimately connected
• Humans construct categories from instances
• Concept represents totality of a category
• Concepts and categories are stored in the brain

How are such concepts represented in our brain?

• Mental Representation

Cognitive Abilities

• Obtain sensory info, store representations, and utilize stored information. Examples are visual, auditory, perceptual illusions. Also concept, proposition, rules, analogies, comparison between similar situations. Specifying relationships between propositions.

Knowledge in LTM

• Retrieval from LTM
• Knowing "what" (general knowledge, semantic, episodic memory, implicit memory concepts, proposals, schemas)
• Knowing "how" (procedural knowledge, automated, scripts, images)

Types of Information in Memory

• Semantic: abstract, general knowledge about facts, concepts
• Episodic: personal experiences
• Procedural: skills, knowledge about "how to"

Representation of Concepts

• Introspection: concepts are related using a single link, use as inquiry method
• Observation -hypothesis. Concepts are represented in the brain as a network
• One concept may lead to another because aspects are linked to features of other concepts

Semantic Network

• Also known as propositional network - concepts as nodes, associations as links.

Do you have the same schema?

• Generalized knowledge about situations or events

Representation of Propositions

• Relationship between concepts that is complete in nature.
• Thoughts are expressible verbally and linguistic in nature.
• Smallest unit of meaning (McKoon & Ratcliff 1980)
  • True or false unit of meaning.

Proposition & Propositional Network

• Subject, predicate, object (relationships between concepts)

Truth Value?

• Proposition expressed in language-like fashion.
• Language means of encoding concepts, propositions, and themselves.
  • Example: "That is a fudge cake." (True/False; Concepts: Fudge, Cake)

Example of truth values in propositions

• Example sentences (robin, wings, uncle): true/false

Propositional Representations

• Entities not yet converted to ordinary language are translatable when needed.
• Talk about; think about concepts

Propositional Representations: Hypothesis

• Meaning/propositions are easier to remember than precise language
  • Meaningful combinations of concepts and relations are central to propositions.

Propositional/Semantic Networks

• Thought rests on mental representation of propositions (thought is associative)
• Associative likely due to shared elements
• How do associative processes work?
  • Hypothesis: meanings and links are stored in a propositional/semantic network

H1: Propositional Networks

• Concepts = nodes
• Association = links
• Theory of spreading activation

H1: Theory of Spreading Activation

• Originated from computer programs simulating human memory search.
• Now supported by neuroscientific evidence.
• Several key assumptions:
  • Concept nodes are activated when processed.
  • Activation spreads in network, decreasing at weaker links (and strengthening stronger ones)

Spreading activation theory

• How does a proposition/semantic network work?
  • Spreading activation
  • Node representing a concept is activated.
• Example: Bread → butter

Spreading Activation Theory

• Activation is strong enough to be attended to.
• Frequently used links and typicality (activation travels faster between nodes)
• Distances between items, irrelevant paths, and interfering information affect the spread.
• Example: Fruit/apple VS Fruit/tomato; Bread/butter VS bread/milk

Spreading Activation Theory

• Example relations: animal, bird, canary, ostrich

Assumptions; Theory of Spreading Activation

• Activation released from concept node during use, hearing, and reading.
• Activation gradually decays from concept node if other activities intervene. Repetition strengthens activation.
• Greater number of stimulated concepts leads to less activation for each concept. - Activation sums from multiple sources to reach a threshold for awareness

Recap

• A person can categorize once the concept is learned
• Two types of mental representation (concept visual/perception; proposition meaning)
• Propositional networks consist of nodes and links (concepts associated)
• Spreading activation model: activation spreads to neighboring concepts

Memory: Implications on Learning

• Organize information into meaningful chunks; verbal & imagery coding; elaboration (interconnectedness); meaningful learning

Reflection; Unit 2

• Difference between concepts and categories; how they are acquired
• Theoretical approaches to categorization
• Representation of concepts/propositions in the brain
• Meaning of mental imagery & examples
• Differences in memory stores (Atkinson & Shriffin 1968), sensory, STM, LTM
• Key ideas in Tulving's Model of Memory

Ideas for Group Project (Assignment 3)

• Problem solving (workspace, university)
• Memory (workspace)
• Memory & Learning
• Experts/novices in the workspace