Week 8 Flashcards
What is Knowledge?
- Connections to truth, facts or principles
- Gained from study, inspection or experience
- Perception & Discovery
Conceptual Knowledge
Enables u to recognise objects & events
Helps us make inferences about things
Mental representation of objects events and abstract ideas (Kiefer & Pulvermuller 2012)
Category and Categorisation
- Category - all possible features of a concept; pointers to knowledge
- Categorisation - Assigning objects and events into a category
- Allows us to understand what we experience in the environment
- Enables us to use objects to achieve our goals
Categories
- Allow us to apply knowledge to new situations
- Ususally an automatic process
- Can be complicated process
Definitional Approach
- Placing Objects into Categories
- Definition of category determines which category it belongs to
- Most natural objects are not easy to define
e.g. how to define a cat?
Categories - Family Resemblance
- Wittgenstein 1953
- Items in a category resemble each other somehow
- This allows some variation within a category
Prototype Approach to Categorisation
- Belonging to a category is determined by comparison to an ideal image that represents it
- Generated using an average style that is common for it.
- Rosch 1973
- Close resemblance creates the category
Quantifying Typicality
- Presented participants with a Category
- Rate Items based on the category from 1-7
- Shows that sometimes objects are considered better or worse examples of categories
- Rosch & Mervis 1975 found that high resemblance categories can overlap in features
Sentence Verification Technique
Smith et al 1974
Subjects ask to respond yes/no quickly to a statement
Responses were faster for common objects
Prototype Approach to Categorisation
- Rosch 1975b
- Primed subjects with a words then asked to decide if samples matched
- These included similar and non similar stimuli
- Subjects faster to respond to the stimuli when they were good examples of the primer
- Priming helps identification of highly typical objects
Exemplar Approach to Categorisation
- Rather than taking a simple “average” Exemplar takes into account many samples
- Object is compared to every sample to determine belonging
- Used to support prototypical approach
- Sparrows are more birdlike than penguins so they are categorised faster.
Exemplars or Prototype Approach
- May allow better performance with highly atypical objects.
- Deals better with categories that are highly variable like games
e.g. penguins and birds
How are categories organised
- We break down categories into sub categories
- Rosch 1976
- Basic level category that is more psychologically important than other levels?
- Global (Superordinate)
- Basic
- Specific (Subordinate)
Categories Organised
- Subjects listed more items for the categories were more specific
- Rosch 1976 argued that basic category is special
- Because provides large increase of information over global level
- But does not lose much information ad the specific level
Basic Categories
- Subjects respond faster when using basic level categories
- Basic Level categories prime subjects better than global level categorise
e.g. cars is faster that vehicles
Knowledge can influence Categorisation
- Tanaka & Taylor
- Asked bird experts and non experts to name objects from lots of categories
- Also had 4 bird pictures
- Experts identified Bird Species and non experts said bird
- Experts pay attention to features of the birds; non experts were not aware of them
- The Level for special is different for each person depending on their expertise
Conceptual Knowledge is . . .
Enables us to recognise objects and make inferences about objects
Categorisation Allows . . .
- Understand what is happening in the environment
- Draw large amount of knowledge from a stimulus
- Use objects in the environment to achieve our goals
- Definitional approach limits us because most natural objects don’t conform perfectly with their definitions
Family Resemblance Allows . . .
Variation in a category by using shared features
Prototype Approach does
Compares the object to an average prototype of the category
Exemplar Approach does
Compare an object to every other instance of a category
Exemplar or Prototype Approach to Categorisation
- Both approaches may be required
- Rosch 1976
- Basic level of category is psychologically special
- Provides a large increase of information over the global level
- Does not lose much information from the specific level.
- People tend to identify objects using basic level categories.
- However, experts tend to use more specific levels
Perceptual Categorisation
- Decisions experts make can be conceptualised
- This is called Categorisation Judgement
- Skin cancer: Is the mole cancer or not?
- X-rays: Is the bone broken or not?
- Radiologists: Is that scan cancer or not?
- Fingerprints: Do these prints match or not?
Perceptual Categorisation – Fingerprints
- Could categories on single features like loops & whirls
- Do experts do this?
- Thompson & Tangen
- Yes they can distinguish print under pressure
Perceptual Categorisation – Global Approach
- Focus on individual features like in prototype approach
- May be relying on global information shared between instances
- Radiologists presented with scans of tumours may do this
- Evans et al 2013 –
- Exemplar Approach sufficient exposure to scans examples they can identify the ‘gist’ of the category
How much info do we need to categorise
- Searston et al 2014 say not much
- Subjects required to report if two examples are from the same or different subcategory
- Subjects able to distinguish above chance levels down to 2x2 resolution
- Global style/redundant inforation is sufficient to distinguish between categories
Perceptual Categorisation – Bees
- Humans can perceive a lot with little information
- Apparently so can Bees
- We et al 2013
- Trained honey bees to distinguish between paintings by Picasso and Monet
- Wu et al 2013
- Bees trained on 5 paintings
- Could distinguish pairs in colour and black & White after 5 days
- Also could distinguish novel paintings based on style
Summary Categorisation
- Experts are able to distinguish categories based on little detail
- Evidence for both Prototype and Exemplar approach
- Redundant Global information can be enough for us to distinguish categories
- The ability is not just found in humans
Sensory Functional Hypothesis
- Some Patients with sensory deficits can identify animals but not tools
- Sensory attributes =/= animals
- Some can identify some tools but not others.
- Functional Attributes =/= Artifacts
Multiple Factor Approach
- Distributed Representation
- Members of specific categories share perceptual attributes
- Many different attributes need to be considered to group objects
- Colour, size, shape
- Hoffman & Lambon-Ralph 2013 - Rate 160 items on 7 features
- Animals more likely to be associated with motion and colour
- Artifacts more likley to be associated with performed actions
- Mechanical objects and instruments overlapped with artifacts & Animals
- Involved both action knowledge and sensory attributes
Multipe-Factor Approach - Crowding
- Animals share many features like Eyes, Legs, Movement
- Artifacts share fewer features eg: Cars versus boats
- Suggested Patients with Category Specific impairment may lose their ability to identify animals
- They have difficulty distinguishing things that share similar features
Semantic Category Approach
- Specific Neural circuits in brain exist for specific categories
- Research shows specific brain area respond to certain stimuli
- Different categories activate different brain location
- Huth et al 2006
- Subjects listened to a story under fMRI
- Different words activated different areas of the cortex
Semantic Categoery Approach - Brain Map
- The brain’s response to single items are distributed over a number of different cortical areas.
- Identifying faces activates areas that respond to:
- Face area in temporal lobe
- Emotions
- Facial expressions
- Where the face is looking
- Attractiveness
- Hammer activates areas related to:
- Shape and colour
- Actions
The Embodied Approach
- Barasalou 2008
- Knowledge of concepts based on reactivation of sensory and motor processes
- This occurs when we interact with that object
- when we use a hammer sensory areas for shape, size, colour
- also activates motor area
- When we see the word hammer the same sensory and motor areas activated to represent the hammer
Embodied Approach
- Hauk et al 2004
- Measured brain activity under fMRI
- Subjects asked to:
- Move their left foot, left/right index finger, or tongue.
- Read “action words” such as kick, pick, lick.
- Areas of cortex activated by actual movement similar to those caused by reading
Embodied Approach - Patient AA
- Garcia et al. 2013
- Patient AA stroke affected ability to produce actions associated with various objects
- Could not indicate how to use hammers, scissors, feather dusters
- Embodied approach would predict that he should not be able to recognise these
- He could identify pictures of those objects
- Ability to represent motor activity is not needed to recognise objects
The Hub & Spoke Model
- Some people with damaged anterior temporal lobes (ATL) suffer from Semantic Dementia
- This model proposes that areas of the brain associated with specific functions are connected to the ATL
- ATL Serves as a hub that integrates information from these areas
- Purple: ATL
- Yellow: Valence (Strong/Weak)
- Red: Auditory
- Dark Blue: Praxis (Manipulability)
- Light Blue: Functionality
- Green: Visual
- Pink: Speech
Probic et al 2010 - Hub & Spoke Model
- Presented pictures of living things and objects
- Measured time for naming pictures using TMS applied to ATL
- ATL usually activated when using an object
- Parietal inactivation slowed reaction times for man mad objects but not living things
- Specifically reaction time slowed for objects with high manipulability
- ATL inactivation slowed for both man made and living things
Hub and Spoke Results
Support the idea that parietal area was specifically related to manipulating objects
ATL is a hub with general functions
Most researchers agree ATL plays role in integrating stimuli from different brain areas
There may be another hub though
How much this hub is responsible for representation of concepts is not known
Information about concepts is distributed across many structures in the brain
Each of the 4 approaches emphasises different types of information
Sensory Functional Approach
We have memory system that distinguishes sensory attributes and another that distinguishes functions
Multiple Factor Approach
Emphasises the role of many different features and properties
Category Specific Approach
Emphasises specialised areas of the brain and networks connecting these areas
Embodied Approach
Activity caused by sensory and motor properties of objects
Hub & Spoke Model and the ATL
- ATL acts as a hub to integrated information from other areas in the brain
- These areas all have different functions
Semantic Networks
- Collins & Quiglan 1969
- Concepts are arranged into networks
- Nodes - represent a concept
- Links - Represent connections between nodes
- Heirarchical
- Levels arranged so that more specific concepts are at the bottom
- More general concepts are at higher levels
Lexical Decision Task
- Max & Schvaneveldt 1971
- Read stimuli and asked to decide if the stimulus is a word or non word
- Reaction times were faster when the two words were associated
- e.g. Doctor/Nurse
Lexical Decision Task - Spreading Activation
- When a node in a network is activated it spreads to nearby connected nodes
- Activated node is Primed and more easily retrieved from memory
- e.g. Robin will also activate bird, which may in turn activate canary
Connectionist Approach
- McClelland & Rogers 2003
- Properties represent as individual Units
- Activated by the pattern of activity in representation in the hidden network
- When the Canary concept and can relations are activated the spread of activation creates a pattern of hidden units which will activate related words
- Matsumoto & Juang 2013/17
- Culture is a unique meaning and information system, shared by a group and transmitted across generations, that allows the group to meet basic needs of survival, pursue happiness and well-being, and derive meaning from life.
Cultural Effects on Cognition and Perception
- Hofstede 1980
- Culture is Mental Programming that shapes our experience
Psychology Theories of Perception
- Generally assumed that perception, cognition, emotion and personality are universal
- How might culture influence these processes
- Is psychology theory biased towards a particular culture
Mueller-Lyer Illusion 1889
- Culture can affect Perception
- Even though perception is believed to be universal across cultures as it is based in physiology
- Cultures that do not focus on right angled geometry don’t perceive this illusion
Perceptual Illusions
- Muller-Lyer Illusion - 1889
- Eye and Brain Angles Gregory - 1966
- Shepard’s Table - Shepard 1990
Cross Cultural Differences on Perception
- Muller-Lyer Illusion is stronger for industrialised cultures vs non industrialised cultures
- Pollack & Silver 1967 say Mueller-Lyer Illusion declines with age
- Non Europeans with more retinal pigmentation less able to detect contours
- This seems to have been debunked it seems African American and European American children from similar cultures perceive it the same
Cross Cultural Differences - The Effect of Experience
- Stewart 1973 - Black & White children living in America showed no differences
- Children from urban Zambia had stronger illusion effect that rural Zambian children
- Not necessarily physiological differences but environmental ones that shape our perception
Masuda & Nisbett 2001
Attention: What do you attend to?
- Compared Japanese vs American subjects
-
Analytic:
- Focus on specifics
- Salient object/person
- Context Independent
-
Holistic:
- Focus on the whole
- Attend to relationship between object & Context
- Context Dependent
- Found no difference for memory of focal objects
- Japanese students remembered more background objects
Background Influenced Recognition
- Masuda & Nisbett 2003
- Original background:
- Japanese > Americans
- Novel Background
- Japanese < Americans
- Collectivist Japanese Culture - focus on relationship between objects & background
- Individualist US Culture - focus mainly on salient objects
Culture and Dialectical Thinking
- Peng & Nisbett 1999
- Studied Chinese people and American people
-
Dialectical Thinking
- Addresses issues from both sides
- Attempt to reconcile contradictions
-
Non-Dialectical Thinking
- One side is more right that the other
- argues strongly from one viewpoint
- Collectivist responses more dialectical
- Individualist responses more nondialectical
Prototypical Approach to Categorisation
- Takes a single average of a prototype and sets it as the standard
- Other examples are compared to it and measured based on salience
Exemplar approach to Categorisation
- Takes many samples of an object
- compares many of it’s features to see if it belongs in a category
- Better with Highly Atypical object
Rosch 1976 - Special Basic Category
- Basic category level is more psychologically special
- Provides a large increase of information from the global category
- Not much increase in the next specific level