2002/03/05psyc202-005, copyright jason harrison 20021 cognition transforming perceptions into action...

2002/03/05 Psyc202-005, Copyright Jason Harrison 2002

1

Cognition

transforming perceptions into action and thought

2002/03/05 Psyc202-005, Copyright Jason Harrison 2002 2

Perception and Cognition• Attention• Memory• Language• Cognitive Modules• Consciousness• Animal Cognition• Cognitive Science

2002/03/05 Psyc202-005, Copyright Jason Harrison 2002

3

Attention

selection, processing, and consciousness of specific sensations


Attention

• recognized as very important to perception– how much can we do at once? why?

• difficult to describe in objective terms– clarity of perception; intensity; consciousness

• recent work: emphasis on selection• selective access:

– only certain parts of [visual] input sent on for further processing

– examples: selective looking and selective listening


Example 1:Selective Looking (Neisser; Simons)• Watch two teams of players

– one in white shirts– one in black shirts

• Each player throws basketball to others on their team --- teams do not interact

• Select one of the colours (white or black)– Task: count number of times ball is

exchanged


From http://www.wjh.harvard.edu/~viscog/grafs/demos/gorilla.shtml

• Pick the white or black team• Count the number of times their

basketball is exchanged

QuickTime™ and aCinepak decompressor

are needed to see this picture.

Neisser; Simons


Neisser; Simons• Result:

– over half the observers do not see the person in the gorilla suit

• inattentional blindness:– if we don’t attend to something we won’t see it

• Instead of a complete, detailed world, we only see a small part of it– the part we are attending to!

• This is how magicians make things (dis)appear


Card Trick

• Pick a card


I’ve removed your card


Selective Looking andInattentional Blindness• Does anything get through?

– picture of a smiley face (but not a sad one)– name of observer (Sara, George)

but not close names (Sura, Geosge)– highly familiar stimuli– highly predictable or expected stimuli

• Some things “draw attention” to themselves– processed “without attention”, or

very easily pass “perceptual filters”


brain

paper

light

aircraft

shirt

...

fruit

morning

planet

door

swan

...

each audio stream is independent of each other (streams don’t interact)

Example 2:Selective Listening (Cherry and many others)• Two audio streams: one input per ear

repeat (“shadow”) what is being said


after the listener has finished shadowing,test what they remember from other stream…

brain

paper

light

aircraft

shirt

...

fruit

morning

planet

door

swan

...

brain

paper

light

aircraft

shirt

...

fruit

morning

planet

door

swan

...

Shadowing left ear

brain, paper, light,

brain

paper

light

aircraft

shirt

...

fruit

morning

planet

door

swan

...

Shadowing right ear

fruit, morning, planet,

Cherry• Select one of the streams (left or right)

repeat (“shadow”) what is being said


Cherry

• Result– Listeners acquire almost nothing from

unattended stream

• Inattentional deafness:– if we don’t attend to something, we won’t

hear it

• Instead of a complete, detailed world, we only hear a small part of it– yes dear, of course dear, you’re so right dear


Selective Listening andInattentional Deafness• Does anything get through from unattended

channel?– large change in volume– change in gender of speaker– not a change in language

• Again, some things” draw attention” to themselves– processed “without attention”

• Related:– “cocktail-party problem” -- multiple speakers– “could you repeat that?” -- short term auditory memory


Example 3:Selective Access (Posner)• Focus visual attention to an area by using a

cue– “Spotlight” or “zoom lens”

• measure time to identify target item when:– observer does not know where item will appear– observer does know where item will appear

• cue is a briefly presented dot at the location of target


Posner: Example 1• no cuing: what letter appears?


Posner: Example 1• no cuing: what letter appears?

A


Posner: Example 2• with cuing: what letter appears?



N


Time

No Cue Cue

Cue reduces time

Advance knowledge of location improves performance

Results: Cue versus No Cue


Amount of reduction depends on distance from cue

Reductionof time

Cued location

Greatest reduction at location of cue

-attention is like a spotlight or a zoom lens

Results: Relative position of Cue


Cue triggers “formation of a spotlight”• whatever is in spotlight is attended• more it is attended, the better it is

processed• size, shape of spotlight can be controlled


Different types of cues

a) exogenous (outside generating)• low-level “reflexes”• sudden changes (e.g., flash or movement)• draws attention automatically

b) endogenous (inside generating)• high-level control• instruction (via some kind of visual sign or

pattern)• sends attention to requested location


Exogenous Cuing


Exogenous Cuing

No interpretation needed - cue is at target location

N


Endogenous Cuing


Endogenous Cuing

Observer need to interpret the cue


Endogenous Cuing

C

Observer need to interpret the cue


Effect of Cues

• Both types of cues control the same attentional mechanism (spotlight) but reflect different strategies

a) exogenous (low-level control)• bottom-up control of attention• based on what’s actually happening in

environment

b) endogenous (high-level control)• top-down control of attention• based on what observer believes


Summary of Selection

• Selective Looking – inattentional blindness

• Selective Listening – inattentional deafness

• Selective Access – cuing, exogenous, endogenous


Attention: Selective Processing

• Selection of sensations (left ear, right ear)• Selective access• Selective construction


Selective Access

• Only certain parts of [visual] input sent on for further processing– some parts ignored, some filtered, some

selected

• Perception:coordinated interaction of different perceptual systems

• Even if attention allows observer to see,how well are things put together?


Selective Construction

• Only certain parts of [visual] input are put together

• Attention both allows conscious vision and builds up visual representations

• Severe limits to what can be accessed or constructed…


Example 1: Visual Search- what visual properties are formed immediately?

- diagnostic: visual search

- e.g., is there a blue dot?


-Result: for some targets, search always fast

- target “pops out” of display

Reactiontime (ms)

Number of items (set size)

300

400

500

600

Search slope = 0(“pop-out”)

- e.g., is there a blue dot?

Example 1: Visual Search


-Explanation:- for some properties, a unique value will

draw attention (exogenous cue)

Search slope = 0(“pop-out”)

- e.g., where’s the blue dot?-e.g., a single large item among small ones

-e.g., a single curved item among straight ones


Proposed application: data visualization


- e.g., is there a blue vertical line?

-However, this is not always the case



- e.g., is there a blue vertical line?

Reactiontime (ms)

Number of items (set size)

300

400

500

600

search slope(30-100 ms/item)

-Result: for some targets, search is slow

- effort needed



- need to combine properties to detect target

-combination is not automatic

-spotlight of attention is needed to “weld”properties together

-search is a serial process - time neededdepends on number of items

-spotlight travels at about 50 ms/item

Example 1: Visual SearchExplanation


Recall:-visual input via high-resolution fovea

Retina has good acuity only in central 4° of vision

-”clear vision” only in this range

Fovea-highest density of cones

Retina

180°

4°

Eye

Example 2: Perception of Scenes


When perceiving a scene, can only get “pieces” of it at any instant

-need to move eyes around to see scene

Example 2: Perception of Scenes

Need to move eyes around to see the world - eye movements make jumps called saccadesSaccades made about 3 times / second

- jump to various parts of the world

-by moving eyes around, a complete scan of world is made

-eyes are almost always moving (unless something really interesting is seen)

Saccades


Question: How do the “pieces” (individual fixations)

get put together?Possible answer: The information from each fixation

is collected into a visual buffer somewhere in the brain

Construction of whole from saccades

(eyes being moved around)

Visual BufferAccumulates information

However, no evidence of this has ever been found

Visual BufferAccumulates information

Visual Buffer: Does not exist• If it did, it should be easy to see any

changes made to the image


Seeing changes• sometimes very difficult to do…• insert brief blank interval between

original and changed images– eye blink– saccade or eye movement– occlusion– movie scene cut– see examples at

http://www.wjh.harvard.edu/~viscog/lab/demos.html


http://www.wjh.harvard.edu/~viscog/grafs/demos/door.shtml

Seeing changes in real-life

QuickTime™ and aSorenson Video decompressorare needed to see this picture.

• Observers have great difficultly seeing a change made during an interruption– Change Blindness

• This can happen in real life


Explanation of Change Blindness• Attention is needed to see change

– without it, observers are “change blind”

• Attention combines pieces of the image– forms them into a coherent representation:

token– tokens support perception (detection) of change– tokens describe actual objects in world

(basis for action, e.g., grasping)– tokens can be “mis-mapped” across an

interruption


1. In everyday life, we usually see change

-if something changes, creates a motion signal-motion is an exogenous cue; draws attention

-if something interferes with drawing of attention (flicker, occlusion) then attention doesn’t go to change

-> won’t see it

-attentional distraction is a major cause of traffic accidents (e.g., cell phone usage)

Notes on Change Blindness


2. There is no accumulation of built-up representation

-if observer looks at picture for several seconds before flicker sequence begins -> no effect-representation built up by attention (token) stays built up only as long as attention stays on it

-token “remapped” after flicker/blind/saccade

Attention acts like a hand:-token exists as long as pieces of input are “held”

-if hand picks up drops and later picks up something else, token “remapped” to new item


-token “dissolves” after attention withdrawn


3. If we only see a few objects at a time, why do we have the impression of seeing lots?

-representation of object formed on demand-whenever it is needed, it is built-virtual representation -> “just in time” system

-example: refrigerator light-only on when needed -> looks like it’s always on

-example: object representation-only there when needed

-> looks like it’s always there



Summary of Attention

• Selection is key:– selection of stimuli: looking, listening– processing of stimuli: access, construction

• Blindness comes in many forms– inattentional blindness– change blindness

• Tasks:– counting passes of ball, shadowing words,

letter recognition, visual search, changes across blinks, flashes and interruptions.


You might be wondering

• Where in my head does this all happen?• How do we know about these locations?• Time for

Neural Sites of Visual Attention!

Michael Posner


Neural Sites of Visual Attention

• Studied via– human brain lesions (strokes, traumas)– brain scans of humans– cell recordings in monkeys

• Two different networks (Posner)– Posterior: expression of attention– Anterior: control of attention


Posterior Network

• parietal lobe (and south)• movement and engagement of attention• unconscious processing• three interacting systems:

– disengagement– movement– engagement


Posterior Systems• disengagement

– posterior parietal lobe– permits processing of attended source to end

• movement– superior colliculus

• eye movements: location map for tectopulvinar system

– allows attention to be drawn to new location

• engagement– thalamus (pulvinar)– blocks input from unattended sources


Anterior Network• control of attention• provides contents of awareness• conscious processing• network includes:

– anterior cingulare cortex (prefrontal)– frontal eye fields

• handles executive control– voluntary shifts of attention– responses to endogenous cues

• location of free will?


Summary of Neural Sites

• Posterior– automatic movement and engagement of

attention

• Anterior– voluntary shifts of attention

2002/03/05psyc202-005, copyright jason harrison 20021 cognition transforming perceptions into action...

Documents