Y1 Cog

cognitive psychology

the scientific study of mental processes

the information processing approach

• based on analogy between the mind and computer

• assumes brain processes information from the environment in a similar manner as a computer

a powerful approach to understanding psychological behaviour in labs and real world settings

interested in how we process information, where and when, the implications, errors and applications

clear assumption that information is represented in nervous system

the mind as a computer

hardware - physical system / nervous system

software - mental processes (memory, attention, reasoning, perception)

propositional representations

token mental representations with semantic properties (tokens with meaning)

indirect realism

we access external reality through representations

information - the amount of entropy/disorder in a system

very mathematical and involves predictive probability in a system

cognitive processes aim to process this surprise and filter out any noise

environmental information is processed by a variety of different processing systems (modularity)

discrete systems transform and sort the information collected from the environment

history of cognitive psychology

70 years old

structuralist approach - introspection (Wundt)

• problematic, can’t be verified, different reports, can alter thought processes, assumes mental events are conscious

behaviourism - study of observable, measurable events

• all behaviour can be explained with the learned relationship between stimuli and response

• classical - passive stimuli response association

• operant - pos/neg reinforcement

• mental variables unimportant

• basic drives of eating, breathing, reproduction

• Watson - need a scientific approach to psychology, with same foundations as other physical sciences

cognitive revolution

• paradigm shift with new concepts of attention, skill and capacity

  • behaviourism was too simplistic

• influence of information theory from maths, engineering and computers

information flow 1 - bottom up (data driven)

begins with an analysis of the sensory input eg. light on retina

perception is built up from low level information

perception directs cognition (starts at the bottom, with observable patterns, which inform our higher order cognitive processes)

information flow 1 - top down (concept driven)

high level cognitive influences

knowledge and experience influence our perceptions of the world

our behaviour is influenced by conceptual data eg. can walk around house in dark

information flow 2 - serial

piecemeal processing

bottle neck

sequential/limited processing

information flow 2 - parallel

bulk processing

late/no bottle neck

consecutive

approaches to studying cognition - experimental cognition

very common approach

experimenter controls the variable in attempt to study only 1 particular variable/system (eg. memory capacity)

structures are deduced indirectly as a result of measurements of accuracy and reaction time

approaches to studying cognition - cognitive neuropsychology

uses tests and experiments with patient sample, comparing to non-patient sample / other patients

how cognitive systems work is deduced based upon brain injuries/abnormalities

• involves very small samples and relies heavily on double dissociations

assumes modularity of mind - one area damages one particular function

• functional specialisation - brain has specialised areas for specific task

approaches to studying cognition - cognitive neuroscience

brain imaging techniques: EEG, PET, fMRI, TMS

approaches to studying cognition - computational cognition

artificial intelligence (physical electronics and computer programs)

connectionism

abstract associate networks

applications of cognitive psychology

product design eg. phones

visual behaviour eg. driving and road safety

object/face recognition eg. airport security

social interaction eg. ingroups/outgroups

representational accounts

assume we have activity in our nervous system that indirectly represents the world around us

non-representational accounts

don’t assume we have activity in our nervous system that indirectly represents the world

instead, information is directly picked up as we engage with the world

eg. Gibson’s ecological, Alva Noe’s access

sensation

the physical stimulation of the sensory apparatus

eg. effect of light on retina, vibrations on ear drum, surface pressure on skin

detecting stimulations

information is captured by the sensory systems and converted to neuronal activity

vision

the dominant sense in humans (we are still multi-sensory)

sensory information is captured, converted into electrical energy, carried along multiple specialized routes and processed in multiple locations for different purposes

perception

the ability of the mind to refer sensory information to an external object and its cause

the experiential (consciousness) component of experiencing/sensation

accessing and capturing the outside world, transforming it and representing it in the brain

cohesive experience

sensory information is transformed from one kind of energy, to electrical brain activity

• then the brain has to make sense of this

about making sense of the world, forming a cohesive whole

perception is adequate, not accurate (Matlin and Foley)

• physical stimuli are rich in information

• human sensory system is really good at gathering information

• concepts (higher level processes) shape our perception

sensation → perception is a combination of…

top down processes (concepts/expectations) and

bottom up processes (structures in the visual pathway)

consciousness

the state of awareness of our own existence, sensations, thoughts and surroundings

contrasting unconscious

Bor and Seth:

• consciousness level - scale of awareness from zero contents to fully aware

• consciousness content - moment to moment, here and now experience

consciousness and attention are not the same thing

• they are related but involve different parts and processes in the brain

may come from synchrony and co-ordination

Baumiester and Masicampo - difference between levels of content experience

low-level - basic here and now awareness

high-level - reasoning, self-reflecting (humans only)

David Charmer’s 2 problems:

easy/soft problem - consciousness comes from brain activity (for stimuli that we process with and without awareness is different, but present)

• where does it come from

hard problem - how does brain activity become conscious awareness, where does the experience come from

• how, where and why we do it

Merikle - measuring when we are conscious

subjective threshold (own report)

• point at which individuals can/can’t report awareness of a stimulus

objective threshold (observer)

• point at which individuals can/can’t make accurate forced choice decisions about a stimulus/adjust a stimulus

the binding problem

understanding how the brain puts all the distributed processing information back together, to give a cohesive experience

consciousness is not unitary, but it feels that way

incoming information is extracted and broken down to be processed in multiple locations

synchrony is a possible solution

selective attention can explain this

Feldman - multiple difficulties

• how does broken down information remain contagious/connected over time, as a cohesive

synchrony

precise - use timings of single cell firing to time stamp information

• problem - computationally expensive

general - general pattern of cell firing is used to bind

• problem - loss of detail

selective attention

we attend/process things/events together, so they are bound together in the brain

attentional clusters/grouping

Gestalt

laws/bias for putting information together in a coherent way

similarity - similar things go together

proximity - things close together go together

good continuation - when objects intersect, assume the object continues behind it

why are we conscious

perceptual

• perceiving your environment may help you better interact with it

action control

• free will, control and agency

social communication

• understanding and interacting with others (theory of mind)

information integration

• reflect on your own experience, organised information allows you to integrate experiences as you have them, but we don’t know why

the global workspace model

selective attentional processes affect how much of what we process we are aware of

• eg. filtering, augmenting

what we see/hear is selected

conscious experience comes from:

• here and now contents:

  • self, intentions, expectations, perceptual context

• incoming sensory stimuli (environmental information)

• external output (speech and actions)

• unconscious processes

  • interpretations, memories, language, automatic biases

consciousness comes from synchronised and integrated activity across multiple brain regions

• pre-frontal cortex, anterior gyrus and parts of parietal lobe

support for the global workspace model

fMRI and ERP evidence for differential brain activity for objects/words that are below conscious/subjective perception

behavioural results suggest unconscious processing is possible up to and including semantic level

• response affected by ‘meaning’ of the target, even though you don’t see/hear it

problems with the global workspace model

zero level of consciousness is tricky to define, and impossible to measure

vegetative states and sleep

• patients with limited brain activity appear to have some level of awareness

types of attention

people decide to engage in either focused or divided attention

usually determined by goal-driven or top-down attentional control processes

bottom up attention

passive modes of attention

exogenous

alterness/arousal

reflexive eg. towards a bolt of lightning

top down attention

active modes of attention

endogenous

selective/focused eg. you choose whether to listen/look or not

• selectively attend to certain stimuli in environment, while ignoring others

• present 2+ stimuli inputs, instruction to respond to just 1

  • necessary to reduce overloading our senses and cognitive processes

divided attention

multitasking

ability to undertake several tasks at once

• present at least 2 stimulus inputs, instruction to respond to all

may tell us something about attentional mechanisms and their capacity

the eye as a camera

optics:

• ciliary muscle

• iris

• cornea

• conjunctiva

• pupil

• lens

film

processor

• retina

• fovea

• optic nerve

the retina

film

made up of photoreceptor cells

broken down into:

• fovea - more cells with most of them cones

• parafovea - less cells with most of them rods

acuity/sharpness is highest at the centre

cells in the optic nerve

magnocellular cells

parvocellular cells

koniocellular cells

acuity/sharpness

highest at the centre of the retina

• as something appears further from the centre of the retina, acuity drops off steeply

• but we still perceive a world where everything is sharp in focus

our brain fills in the gap and uses our frequent eye movements to update and hold visual information - like the blind spot

• our perception of the visual world is made by amalgamating visual snapshots and knowledge about our environment

limits on how we encode information (the eye), mean what we actually see is a more fuzzy than what we perceive

saccadic suppression

saccades - rapid movements of the eye between fixation points

• we do not perceive our own saccades

this happens to supress the motion blur during the saccades, to perceive a stable world

• just a decrease of the visual sensitivity

the eye travels up to 900 visual degrees per second

saccades can last up to 50ms

• we are ‘blind’ for over 3 hours of the day

overt attention

attended information (in/around the fovea)

slow, 3-4 saccades per second

covert attention

unattended information (everything apart from in/around the fovea)

faster, 50ms to shift

when we attend somewhere/to something without moving our eyes

• acts as a filter - selects stimuli for further processing

• is limited capacity

includes attentional spotlight and zoom lens

Hemholtz

• observed if we can enhance perception if we focus our attention on a location in the visual field

• but it comes at the expense of other areas in the visual field

attentional spotlight

focused visual attention resembles a spotlight

• enhanced attention within a small region of the visual field

• but not outside the beam of attentional spotlight

zoom lens

attention is directed to a given region of the visual field

the area of focal attention can be increased/decreased, depending on task demands

voluntary orienting (expectancy)

results in faster reaction times

posner task

primary themes characterising attention

capacity limitation

• our limited ability, to carry out various mental operations at the same time, needs a way to prioritise information

perceptual gating (selection)

• conscious perception is always selective

• but selection is not always conscious

attentional modalities

vision

• limit on how much we can take in because things in the environment are placed in different spatial location

auditory

• streams of sound from different locations

  • can’t listen to all at once, so selectively listen - cocktail effect

  • use top-down processing to fill in the gaps

Welford (1952)

presented 2 signals in rapid succession - psychological refractory period paradigm (PRP)

participants make speeded response to both

• reaction time to 2nd stimulus depends on how close it is presented to the 1st stimulus

• the closer the presentation, the slower the reaction time

saw this as evidence for a bottleneck (early selection)

• as processing of 1 stimulus must be completed before processing of the next one can begin

claimed this as good evidence for a central limit on human processing capability

Cherry’s cocktail party effect (1953)

we are able to follow one conversation while several people are talking

due to selective auditory attention

• ability to selectively attend to one auditory message, while filtering out others in a noisy environment

shadowing task - dichotic listening procedure

• shadow = repeat aloud the message told to attend to

• participants were able to:

  • perform the task and repeat the correct message

  • filter out the unshadowed message with little information remembered about this message

  • process the unshadowed physical characteristics

• participants were not able to:

  • detect the meaning of the 2nd stream

  • detect if the 2nd stream was a foreign language/reversed speech

  • repeat any words in the 2nd stream

testing the bottleneck - Broadbent (1958)

3 digits presented to one ear at the same time as another 3 digits presented to the other ear

participants would recall numbers ear by ear better than in pairs

• suggests the stimuli are accessed in parallel, by a sensory buffer

  • this filter stimuli on the basis of the physical characteristics

  • the other input remains in the buffer

attention allows us to selectively process information to filter out irrelevant information

• his filter theory argues that we cannot identify/process something without attention

evaluation:

• theory is based on physical properties = inflexible

attenuation - Treisman (1964)

unattended things are sometimes processed, when they leak through the filter, as they are processed enough to reach the threshold on conscious awareness

where the filter occurs depends on task demands:

• physical cues, syllabic pattern, grammatical structure, meaning

attentional competing hypotheses

early selection

• broadbent

• physical characteristics of messages are used to select one message for further processing, and all others are lost

attenuation

• treisman

• physical characteristics are used to select one message for full processing

• other messages are given partial processing

late selection

• deutsch and deutsch

• all messages get through, but only one response can be made

inattentional blindness

what we don’t attend to, we are not aware of

a grat demonstration that attention is required for identification

change blindness

interleaving screens where large changes happen, but can only see these changes when we are attending to them

Rensinck

• explored ‘are change blindness and inattentional blindness the same?’

  • CB = spotting transitions

  • IB = identifying something that is ‘irrelevant’ to the task

• change blindness suggests there is a failure of visual short-term memory (VSTM), enabling us to compare between scenes (irrelevant for inattentional blindess)

  • CB and change detection tasks show there is a clear limit in the information we can hold in VSTM = limit in what we can hold across a temporal gap

limits of attention

capacity

our idea of what the world is like is incorrect

gestalt psychology

means whole in german

• scientists in 1910 suggested that perception could not be done by breaking it down into parts, but by considering the whole experience

links to object identification

the whole is other than the sum of its parts

laws of perceptual organisation

• eg. law of pragnanz - the percept you see should be the simplest interpretation of the scene (good figure)

• eg. law of proximity, similarity, closure, good continuation, common fate

• provide us some framework on how to separate figure from the ground

theories of object recognition

imaged based models

structural description models

marr’s vision

recognition by component (RBC)

• one of the most difficult tasks the visual system has to perform

imaged based models - theory of object recognition

specific views are ‘stored’ and recognition performance is based on generalisation from these

encode structured templates of viewpoint dependent representations

structural description models - theory of object recognition

information about the 3D structure of an object is extracted from a single view

marr’s vision - theory of object recognition

object parts are represented independently of their spatial cofiguration and viewpoint

everything can be broken down into sub parts

• and looked how they link together

recognition by component (RBC) - theory of object recognition

objects are represented using basic shape units called geons (36 in total)

• geons = variations in a small number of basic parameters called non-accidental properties (NAPs)

  • these are basic features that define variation in shapes, which are view independent

non-accidental properties (NAPs)

curvilinearity - curviness in the 2D image caused by curve on object

parallelism - lines in parallel in 2D objects caused by parallel lines on object

cotermination - 2+ edges that terminate at the same point

symmetry - axis of symmetry in 2D image reflect the axis of symmetry on object

collinearity - a straight line in the 2D image is caused by straight line in object

role of disparity and view generalisation during object recognition

clear evidence of viewpoint generalisation

• accuracy is always better at the learnt rather than interpolated viewpoints, whilst extraploated viewpoints are the most difficult to discriminate

  • argues for image based models

effect of stereo depth information

• although not significant in all experiments, subjects find the 3D viewing condition easier to perform, particularly in the extrapolated viewpoints

  • argues for structural description models

both theories correct

canonical viewpoint`

most recognisable view eg. horse from the side, rather than the front

Palmer, rosch and chase (1981)

• participants shown views of an object and asked to rate how much each one looked like the objects they depict

• in recognition task, reaction time correlated with the ratings

• canonical views are recognised faster at the entry level

why canonical views are recognised faster at the entry level

frequency hypothesis

• it is the view we’ve mostly seen in our lifetime

• the easiness of recognition is related to the number of times we’ve seen the objects from each viewpoint

maximal information hypothesis

• viewpoint contains the most amount and most informative information about the object

• some views provide more information than others about the objects

• best views tend to show multiple sides of the object with all its parts

edelman and bulthoff

• created new objects to control for familiarity

• when presenting all view points with the same frequency, observers had preference for specific viewpoints

• when few viewpoints were presented, recognition was better for previously seen viewpoints

face recognition

our visual system is active in its attempts to organise our perceptual input without any conscious effort

involves ‘within-category’ discrimination

• discrimination between members of the same basic level category

• discrimination of patterns which share the same essential features eg. eyes, mouth, nose

errors in face recognition can have catastrophic consequences eg. eye-witness testimony

yin

• participants are better at recognising upright faces than they are other objects

• but worse for inverted faces than they are for inverted objects

face recognition hypotheses

featural hypothesis

• faces are primarily remembered due to their facial features

configurational/spacing hypothesis

• emphasis on the relationship/spacing amongst the facial features

holistic hypothesis

• takes the face as a perceptual whole, where both configurational and featural information are required for accurate recognition

configuration is more important than the features

the thatcher illusion

tells us that we perceive faces in terms of the global configuration of facial features

we are unable to detect/accurately process the properties of local individual face parts if upside-down

evidence that faces are processes holistically

functional face recognition model - bruce and young

faces are structurally encoded

encoded information activates face recognition units (FRUs)

if match between encoding and FRU, then semantic information can be accessed

personal identity nodes (PINs) contain information about that person

evidence suggesting face recognition is special

johnson and morton

• new born babies will preferentially view faces from 9 minutes old

meltzoff and moore

• expression analysis seems to be innate (independent of object recognition)

face agnosia (prosopagnosia) without object agnosia, object agnosia without prosopagnosia

human fMRI result of objects vs faces:

• subject viewed faces for a while, then pictures of objects

• one area becomes more active during face-viewing (fusion face area - FFA)

• another area becomes more active during object-viewing (lateral occipital complex - LOC)

inversion effect

• healthy participants are better at recognising upright, rather than inverted faces

  • effect is not as strong with objects

• prosopagnosic participants are better at recognising inverted rather than upright faces

Jennifer Aniston cell

single cell recording in live patients, who were shown hundreds of pictures

neurons in the medial temporal lobe (MTL) responded to only jennifer aniston - but not when she was photographed next to brad pitt

• this part of the brain is involved in memory

greebles

participants learned them and became experts

• they are mini figures varying in elements

the fusion face area (FFA) of the brain behaved like they were faces

after this special training, brain activity shifted from object area to face area

evidence suggesting face recognition is NOT special

bruyer et al

• a prosopagnosic farmer who could identify his cows

McNeill and warrington

• patient with prosopagnosia who could distinguish between his sheep

ellis and young

• such cases might simply reflect specialities in processing for many types of object

• after training people to recognise non-face objects, they show face recognition-like performance

language

a system of symbols and rules that enable us to communicate

• the rules specify how words are ordered to form sentences

words, both written and spoken, are symbols that stand for other things

collections of signs, sounds and gestures put together to use rules (grammar) and structure (syntax) to convey message

panbanisha - the bonobo great ape

a lifetime in captivity receiving language training

taught to use a keypad of 400 geometric patterns to communicate

• keypresses prompted a synthetic voice

he understood english at the level of a 2.5 year old human

had a 3000 word vocabulary at 14 years

could produce a grammatically correct sentence with keypresses

shows apes can grasp aspects of language

is language innate or learnt

a critical debate in the development of the cognitive revolution

skinner = learnt

chomsky = innate

innate universal grammar

an innate set of grammatical principles found in all human languages

• multiple linguistic universals

lexical categories

• nouns, verbs, adjectives etc

  • some categories are missing in some languages however

word order

• subject, object verb (SOV)

• subject, verb, object (SVO)

  • almost universal - majority has the subject preceding the object

recursion

• embedding clauses within sentences

  • lacking in some languages eg. Piraha - Amazonian language

justification for innate universal grammar

offers explanation for:

• why only humans develop language fully

• the broad similarities across the various languages

• the rapid speed in which children learn to speak

criticisms of innate universal grammar

there aren’t linguistic universals

• languages differ enormously across the world

children learn language quickly because it’s invented by humans, with the limitations of human performance in mind

language reflects pre-existing and hence non-language specific, human learning and processing mechanisms

natural language

any language that has developed naturally through human interaction and use

• can take different forms eg. speech, signing

fundamental to human cognition and social integration

a flexible, generative, symbolic representation system for communicating meaning

• grammar and syntax are key mechanisms

• meaning is subject to inference and pragmatics

it is complex and resource intensive

• subject to inteference and error

• we have a range of shortcuts

• context is very important

language follows developmental sequence:

• simple → complex

• babbling → words → sentences

natural language appears to be unique to humans

only humans have the vocal equipment and unique brain areas for spoken language

generative and flexible

• finite set of symbols, infinite set of messages

grammar

• combination rules contain subject, verb, object

very young children are sensitive to grammar

importance of natural language

as a social animal, language is necessary to convey simple and complex meaning/ideas

may have been a key factor for human brain development - encephalisation

• human brain is 3x bigger than it should be, given our body size

• correlates with sociality

• only humans have the vocal equipment and unique brain areas for spoken language

non-humans able to convey complex message, but is not the same as natural language

• eg. waggle dance, bird song

3 processes involved in language production - Levelt

conceptualisation

• planning the message

• desired concept to be communicated

formulation

• transforming the intended message into speed sounds and sentences

  • including grammatical, morpho-phonological and phonetic encoding

articulation

• words turned into speech

• movements from the tongue, lips, jaw, lungs, larynx and glottis

these processes are incorporated into models of speech production, but they differ in number of processing levels

tip of the tongue phenomenon

a failure in retrieval

cannot recall the exact word, but can recall words of a similar form/meaning

• recall can be guided by partial word information

• info can often be retrieved if given sufficient additional time to retrieve it

occurs when we can access the concept of a word, but cannot access the phonological form of the world

• only the conceptualisation process occurs, not formulation or articulation

phonology

sounds of language

semantics

meaning encoded in language

meaning conveyed by words, phrases and sentences

grammar

rules of language

govern syntax

SVO, plus specific rules for each language

syntax

structural order of words

eg. colourless green ideas sleep furiously

• this has correct grammatical structure - syntactically acceptable

• but does not have clear semantic structure/meaning

phonological structure

organised into phonemes

morphophonological structure

• how the phonemes are arranged/grouped to form words and morphemes

syllabic structure

• how spoken words can be deconstructed into syllables

prosodic structure

• assigning stress and intonation within a sentence

phonemes

constitutes the spoken sounds of a sentence

what we hear

smallet unit of sound in a word than can convey meaning

• eg. swapping the ‘m’ phoneme to ‘n’ in ‘dime’ changes the meaning of the word (dime → dine)

graphemes - the written alternative to a phoneme (what we see)

• phonemes can form graphemes in multiple ways (multiple spellings)

  • eg. shun (sound) can be spelt as tion (writting) in ration for example

• graphemic content within a word can have multiple phonological representations and sounds/pronounciations

  • eg. ough can be dough, tough, plough, thought, through

orthography

the conventional spelling system of a language

ghoti is an alternative spelling for fish

most language have a shallower orthography than english

written production

similar to speech production

• conceptualisation - planning the message

• formulation - transforming the message into sounds and sentences

• articulation - words turned into writing

2 ways of generating the spelling of a word to be written:

• sublexical route

• lexical route

sublexical route

word is prepared as in speech

• until the point where phonemes can be converted to graphemes

useful for words with unfamiliar spellings

also known as sound-to-sound spelling and phoneme-to-grapheme conversion

lexical route

direct retrieval of the orthographic word

• stored knowledge about how the word is written

very important for languages with deep orthographies

allows us to spell words with poor sound-to-spelling consistency, or words with the same spelling eg. they’re/their/there

whorfian hypothesis - language and thought

language determines/influences thinking

• perception of the world is affected by the words we use to describe it

• principle of linguistic relativity

  • supposedly saw similarities with Einstein’s theory of general relativity

strong version - language determines thinking

• intertranslatability - can translate a statement in one language to one in another language, just might not be the same number of words

• contradicts the argument that a thought expressible in one language is not in another

• eg. the word Mokita means ‘truth everybody knows but nobody speaks out’

  • in English there is no word with this meaning, but we understand the concept

weak version - language imposes constraints on thinking

weakest version - language influences memory

winawer et al

studied categorical perception in english and russian subjects

• russian has distinctive words for dark blue (siniy) and light blue (goluboy) 

examined performance in a speeded colour discrimination task  

• russian speakers were faster to discriminate 2 colours if they can be categorised separately  

categories in language can affect performance of basic perceptual colour discrimination tasks 

support for whorfian hypothesis

roschian hypothesis - language and thought

language does not determine/influence thinking

• perception of the world is not affected by the words we use to describe it

language is only used to describe our perceptions of the world

pragmatics and inference

literal meaning vs intended meaning of the sentence

• knowledge about how to use language appropriately in the right contexts

• eg. can someone tell me the time?

  • literal = yes

  • intended = 12:30

concerned with practical language use and comprehension

• grasping the speakers understanding and beliefs with accompanying gestures (vocal and facial)

can be considered as meaning minus semantics

emphasis on figurative language - not intended to be taken literally

• idiom eg. kick the bucket

• metaphor eg. fishing for compliments

  • the non-reversibility of metaphors is an important phenomenon

    eg. my surgeon is a butcher vs my butcher is a surgeon

a phrase looked up in a dictionary can give us definitions

• but cannot give the intended meaning of the sentence

• more information is required, such as speakers tone, stress of words, context