1

THEORIES OF PERCEPTION

In order to receive information from the environment we are equipped with sense organs eg eye,ear, nose. Each sense organ is part of a sensory system which receives sensory inputs andtransmits sensory information to the brain. A particular problem for psychologists is to explainthe process by which the physical energy received by sense organs forms the basis of perceptualexperience. Sensory inputs are somehow converted into perceptions of desks and computers,flowers and buildings, cars and planes; into sights, sounds, smells, taste and touch experiences.

A major theoretical issue on which psychologists are divided is the extent to which perceptionrelies directly on the information present in the stimulus. Some argue that perceptual processesare not direct, but depend on the perceiver's expectations and previous knowledge as well as theinformation available in the stimulus itself. This controversy is discussed with respect to Gibson(1966) who has proposed a direct theory of perception which is a 'bottom-up' theory, andGregory (1970) who has proposed a constructivist (indirect) theory of perception which is a 'top-down' theory.

EXPLAINING PERCEPTION - A TOP-DOWN APPROACH

Helmholtz (1821-1894) is considered one of the founders of perceptual research. He argued thatbetween sensations and our conscious perception of the real world there must be intermediateprocesses. Such processes would be, for example, 'inferential thinking' - which allows us to gobeyond the evidence of the senses (these inferences are at an unconscious level). Thus Helmholtzwas an early Constructivist who believed perception is more than direct registration ofsensations, but that other events intervene between stimulation and experience.

An early illustration that supports the idea of perceptions as modifiable constructions rather thanthe direct responses to pattern of stimulation is the 'Ames Room'. This room is of an irregularshape with a receding rear wall and decorated in a special manner.

C

B A

One explanation for the Ames Room illusion is that the perceiver is in a situation of having tochoose between two beliefs built up through experience - (a) rooms that look rectangular andnormal, usually are just that, (b) people are usually of 'average' size. Most observers choose (a)and therefore consider the people to be 'odd'.

The interesting thing about the Ames Room illusion is that it does not disappear when you learnthe true shape of the room.

The true wall, AC on the diagram, isdecorated so as to appear to be in theposition AB. Viewed from the frontpeephole with one eye the room appears tobe rectangular but a person moving fromA to C will appear to shrink.

Viewing Point

2

PERCEPTIONS AS HYPOTHESES - R L GREGORY (B 1923)

Gregory proposes that perceiving is an activity resembling hypothesis formation and testing. Hesays that signals received by the sensory receptors trigger neural events, and appropriateknowledge interacts with these inputs to enable us to makes sense of the world.

Gregory has presented evidence in support of his theory, some of which is outlined below:

1. 'Perception allows behaviour to be generally appropriate to non-sensed objectcharacteristics'.

For example, we respond to certain objects as though they are doors even though we can only seea long narrow rectangle as the door is ajar.

2. 'Perceptions can be ambiguous'

The Necker cube is a good example of this. When you stare at the crosses on the cube theorientation can suddenly change, or flip'. It becomes unstable and a single physical pattern canproduce two perceptions.

Gregory argues thatsurely to do this wemust be using more thanjust sensory inputs.

How do we know fromthis stimulus alone thatthis is a door?

3

3. 'Highly unlikely objects tend to be mistaken for likely objects'.

Gregory has demonstrated this with a hollow mask of a face. Such a mask is generally seen asnormal, even when one knows and feels the real mask. There seems to be an overwhelming needto reconstruct the face, similar to Helmholtz's description of 'unconscious inference'.

What we have seen so far would seem to confirm that indeed we do interpret the information thatwe receive, in other words, perception is a top down process. However:….

EVALUATION OF THE TOP-DOWN APPROACII TO PERCEPTION

1. The Nature of Perceptual HypothesesIf perceptions make use of hypothesis testing the question can be asked 'what kind of hypothesesare they?' Scientists modify a hypothesis according to the support they find for it so are we asperceivers also able to modify our hypotheses? In some cases it would seem the answer is yes.For example, look at the figure below:

4

This probably looks like a random arrangement of black shapes. In fact there is a hidden face inthere, can you see it? The face is looking straight ahead and is in the top half of the picture in thecentre. Now can you see it? The figure is strongly lit from the side and has long hair and a beard.

Once the face is discovered, very rapid perceptual learning takes place and the ambiguouspicture now obviously contains a face each time we look at it. We have learned to perceive thestimulus in a different way.

Although in some cases, as in the ambiguous face picture, there is a direct relationship betweenmodifying hypotheses and perception, in other cases this is not so evident. For example, illusionspersist even when we have full knowledge of them (e.g. the inverted face, Gregory 1974). Onewould expect that the knowledge we have learned (from, say, touching the face and confirmingthat it is not 'normal') would modify our hypotheses in an adaptive manner. The currenthypothesis testing theories cannot explain this lack of a relationship between learning andperception.

2. Perceptual Development

A perplexing question for the constructivists who propose perception is essentially top-down innature is 'how can the neonate ever perceive?' If we all have to construct our own worlds basedon past experiences why are our perceptions so similar, even across cultures? Relying onindividual constructs for making sense of the world makes perception a very individual andchancy process.

The constructivist approach stresses the role of knowledge in perception and therefore is againstthe nativist approach to perceptual development. However, a substantial body of evidence hasbeen accrued favouring the nativist approach, for example:

Newborn infants show shape constancy (Slater & Morison, 1985); they prefer their mother'svoice to other voices (De Casper & Fifer, 1980); and it has been established that they prefernormal features to scrambled features as early as 5 minutes after birth.

5

4. Sensory Evidence

Perhaps the major criticism of the constructivists is that they have underestimated the richness ofsensory evidence available to perceivers in the real world (as opposed to the laboratory wheremuch of the constructivists' evidence has come from).

Constructivists like Gregory frequently use the example of size constancy to support theirexplanations. That is, we correctly perceive the size of an object even though the retinal image ofan object shrinks as the object recedes. They propose that sensory evidence from other sourcesmust be available for us to be able to do this.

However, in the real world, retinal images are rarely seen in isolation (as is possible in thelaboratory). There is a rich array of sensory information including other objects, background, thedistant horizon and movement. This rich source of sensory information is important to thesecond approach to explaining perception that we will examine, namely the direct approach toperception as proposed by Gibson.

A DIRECT APPROACH TO PERCEPTION - GIBSON 1966

Gibson claimed that perception is, in an important sense, direct. He worked during World War IIon problems of pilot selection and testing and came to realise:

In his early work on aviation he discovered what he called 'optic flow patterns'. When pilotsapproach a landing strip the point towards which the pilot is moving appears motionless, with therest of the visual environment apparently moving away from that point.

The outflow of the optic array in a landing glide.

According to Gibson such optic flow patterns can provide pilots with unambiguous informationabout their direction, speed and altitude.

Three important components of Gibson's Theory are 1. Optic Flow Patterns; 2. InvariantFeatures; and 3. Affordances. These are now discussed.

6

1. Light and the Environment - Optic Flow Patterns

Changes in the flow of the optic array contain important information about what type ofmovement is taking place. For example:

2 Any flow in the optic array means that the perceiver is moving, if there is no flow theperceiver is static.

3 The flow of the optic array will either be coming from a particular point or moving towardsone. The centre of that movement indicates the direction in which the perceiver is moving. Ifa flow seems to be coming out from a particular point, this means the perceiver is movingtowards that point; but if the flow seems to be moving towards that point, then the perceiveris moving away. See above for moving towards an object, below is moving away:

The Optic Flow pattern for a person looking out of the back of a train.

2. The role of Invariants in perception

We rarely see a static view of an object or scene. When we move our head and eyes or walkaround our environment, things move in and out of our viewing fields. Textures expand as youapproach an object and contract as you move away. There is a pattern or structure available insuch texture gradients which provides a source of information about the environment. This flowof texture is INVARIANT, ie it always occurs in the same way as we move around ourenvironment and, according to Gibson, is an important direct cue to depth. Two good examplesof invariants are texture and linear perspective.

7

3. Affordances

Are, in short, cues in the environment that aid perception. Important cues in the environmentinclude:

OPTICAL ARRAY The patterns of light that reach the eyefrom the environment.

RELATIVE Objects with brighter, clearerBRIGHTNESS images are perceived as closer.

TEXTURE The grain of texture getsGRADIENT smaller as the object recedes. Gives the

impression of surfaces receding intothe distance.

RELATIVE SIZE When an object moves further awayfrom the eye the image gets smaller.Objects with smaller images are seenas more distant.

SUPERIMPOSITION If the image of one object blocks theimage of another, the first object isseen as closer.

HEIGHT IN THE Objects further away areVISUAL FIELD generally higher in the visual field.

Texture Gradientgiving theappearance ofdepth

Linear Perspective

Parallel lines, eg railway tracks,appear to converge as they recedeinto the distance.

8

EVALUATION OF GIBSON'S DIRECT APPROACH TO PERCEPTION

Visual Illusions

Gibson's emphasis on DIRECT perception provides an explanation for the (generally) fast andaccurate perception of the environment. However, his theory cannot explain why perceptions aresometimes inaccurate, eg in illusions. He claimed the illusions used in experimental workconstituted extremely artificial perceptual situations unlikely to be encountered in the real world,however this dismissal cannot realistically be applied to all illusions.

For example, Gibson's theory cannot account for perceptual errorslike the general tendency for people to overestimate verticalextents relative to horizontal ones.

Neither can Gibson's theory explain naturally occurring illusions. For example if you stare forsome time at a waterfall and then transfer your gaze to a stationary object, the object appears tomove in the opposite direction .

Bottom-up or Top-down Processing?

Neither direct nor constructivist theories of perception seem capable of explaining all perceptionall of the time. Gibson's theory appears to be based on perceivers operating under ideal viewingconditions, where stimulus information is plentiful and is available for a suitable length of time.Constructivist theories, like Gregory's, have typically involved viewing under less than idealconditions.

Research by Tulving et al manipulated both the clarity of the stimulus input and the impact of theperceptual context in a word identification task. As clarity of the stimulus (through exposureduration) and the amount of context increased, so did the likelihood of correct identification.However, as the exposure duration increased, so the impact of context was reduced, suggestingthat if stimulus information is high, then the need to use other sources of information is reduced.One theory that explains how top-down and bottom-up processes may be seen as interacting witheach other to produce the best interpretation of the stimulus was proposed by Neisser (1976) -known as the 'Perceptual Cycle'.

PERCEPTUAL SET

The concept of perceptual set is important to the active process of perception. Allport, 1955defined perceptual set as:

"a perceptual bias or predisposition or readiness to perceiveparticular features of a stimulus".

Perceptual set is a tendency to perceive or notice some aspects of the available sensory data andignore others. According to Vernon, 1955 set works in two ways: (1) The perceiver has certainexpectations and focuses attention on particular aspects of the sensory data: This he calls a'Selector'. (2) The perceiver knows how to classify, understand and name selected data and whatinferences to draw from it. This he calls an 'Interpreter'.

9

It has been found that a number of variables, or factors, influence set, and set in turn influencesperception. The factors include:

• Expectations• Emotion• Motivation• Culture

1. EXPECTATION

(a) Bruner & Minturn, 1955 illustrated how expectation could influence set by showingparticipants an ambiguous figure '13' set in the context of letters or numbers e.g.

The physical stimulus '13' is the same in each case but is perceived differently because of theinfluence of the context in which it appears. We EXPECT to see a letter in the context of otherletters of the alphabet, whereas we EXPECT to see numbers in the context of other numbers.

(b) We may fail to notice printing/writing errors for the same reason. For example:

1. 'The Cat Sat on the Map and Licked its Whiskers'.2.

(a) and (b) are examples of interaction between expectation and past experience.

(c) A study by Bugelski and Alampay, 1961 using the 'rat-man' ambiguous figure alsodemonstrated the importance of expectation in inducing set. Participants were shown either aseries of animal pictures or neutral pictures prior to exposure to the ambiguous picture. Theyfound participants were significantly more likely to perceive the ambiguous picture as a rat ifthey had had prior exposure to animal pictures.

Oncein a

a lifetime

10

2. MOT1VATION AND EMOTION

Allport, 1955 has distinguished 6 types of motivational-emotional influence on perception:

(i) bodily needs (eg physiological needs)(ii) reward and punishment(iii) emotional connotation(iv) individual values(v) personality(vi) the value of objects.

(a) Sandford, 1936 deprived participants of food for varying lengths of time, up to 4 hours, andthen showed them ambiguous pictures. Participants were more likely to interpret the pictures assomething to do with food if they had been deprived of food for a longer period of time.Similarly Gilchrist & Nesberg, 1952, found participants who had gone without food for thelongest periods were more likely to rate pictures of food as brighter. This effect did not occurwith non-food pictures.

(b) A more recent study into the effect of emotion on perception was carried out by Kunst-Wilson & Zajonc, 1980. Participants were repeatedly presented with geometric figures, but atlevels of exposure too brief to permit recognition. Then, on each of a series of test trials,participants were presented a pair of geometric forms, one of which had previously beenpresented and one of which was brand new. For each pair, participants had to answer twoquestions: (a) Which of the 2 had previously been presented? ( A recognition test); and (b)Which of the two was most attractive? (A feeling test).

The hypothesis for this study was based on a well-known finding that the more we are exposedto a stimulus, the more familiar we become with it and the more we like it. Results showed nodiscrimination on the recognition test - they were completely unable to tell old forms from newones, but participants could discriminate on the feeling test, as they consistently favoured oldforms over new ones. Thus information that is unavailable for conscious recognition seems to beavailable to an unconscious system that is linked to affect and emotion.

3. CULTURE

(a) Deregowski, 1972 investigated whether pictures are seen and understood in the same way indifferent cultures. His findings suggest that perceiving perspective in drawings is in fact aspecific cultural skill, which is learned rather than automatic. He found people from severalcultures prefer drawings which don't show perspective, but instead are split so as to show

11

both sides of an object at the same time. In one study he found a fairly consistent preferenceamong African children and adults for split-type drawings over perspective-drawings. Split-type drawings show all the important features of an object which could not normally be seenat once from that perspective. Perspective drawings give just one view of an object.Deregowski argued that this split-style representation is universal and is found in Europeanchildren before they are taught differently.

Elephant drawing split-view and top-view perspective. The split elephant drawing was generallypreferred by African children and adults.

(b) Hudson, 1960 noted difficulties among South African Bantu workers in interpreting depthcues in pictures. Such cues are important because they convey information about the spatialrelationships among the objects in pictures. A person using depth cues will extract a differentmeaning from a picture than a person not using such cues.

Hudson tested pictorial depth perception by showing participants a picture like the one below. Acorrect interpretation is that the hunter is trying to spear the antelope, which is nearer to him thanthe elephant. An incorrect interpretation is that the elephant is nearer and about to be speared.The picture contains two depth cues: overlapping objects and known size of objects. Questionswere asked in the participants native language such as:

‘What do you see?’‘Which is nearer, the antelope or the elephant?’

‘What is the man doing?’'

The results indicted that both children and adults found it difficult to perceive depth in thepictures.

12

The cross-cultural studies seem to indicate that history and culture play an important part in howwe perceive our environment. Perceptual set is concerned with the active nature of perceptualprocesses and clearly there may be a difference cross-culturally in the kinds of factors that affectperceptual set and the nature of the effect.

VISUAL CONSTANCIES

Perceptual constancies involve seeing visual objects accurately, regardless of their distance awayfrom us, or other factors that distort the retinal image. For example, the door is 'seen' as arectangular shape even when open and the retinal image is of a trapezium.

13

1. SIZE CONSTANCY

When we observe an object, the light falling on the retina is known as the 'retinal image'. Lightrays enter through the lens in the front of the eye and are focused on a particular area of theretina at the back of the eye. The intriguing question is 'how does the brain interpret the lightimage on the retina to arrive at an accurate perception?' This is particularly interesting when wetry to explain how objects are perceived as the same size even when seen at a distance. This canbe demonstrated:

Hold both hands in front of you, your left hand at arms length and your right hand about halfway to your face. Both hands are 'perceived' as the same size, but the retinal image of the righthand will be much larger. Now move your right hand so that it overlaps the left hand, stin withthe left at arms length and the right hand halfway towards your face. You should find that theright hand 'swamps' the left - this is because the two hands are actually stimulating differentsized retinal images. The diagram below shows the different sizes of retinal images projected bythe same sized hand at different distances.

How is it that we 'perceive' the hand as being the same size in spite of these differences in theretinal image? One explanation for this (ie for 'size constancy') is that the brain receivesinformation both about size of the retinal image and distance of the object. The visual system

Note that thenearer handhas a muchlarger imageon the retina.

Lens

Hands, near and far.

Retina

14

seems to automatically make allowances for distance. For example, even thought the retinalimage of the left hand is small, distance cues inform the brain that it is further away than theright hand and this can explain the smaller retinal image. Taking account of both size anddistance in the visual system the brain would probably conclude that the hands are the same size.

Several distance cues have been identified which could aid the process of constancy scaling, andtwo of these are:

A. RETINAL DISPARITY

The retina in each eye receives a slightly different image. To demonstrate this: close one eye andline your finger up with the corner of the room. Now close that eye and open the other eye - thefinger appears to move. Normally, when using both eyes the visual system calculates how faraway the finger is by combining information from the differences between the two images on theretinas.

B. MOTION PARALLAX

To a moving observer distant objects appear to move more slowly than near objects. Forexample, when you look out of the window of a moving train nearer objects like telegraph polesflash by faster than distant telegraph poles. The visual system can use this information tocalculate how far away the telegraph poles are.

2. SHAPE CONSTANCY

Knowledge of the 'real' shape of an object means that it is still perceived as being the sameregardless of the angle from which it is viewed. For example, I 'perceive' the wall clock ascircular even though from the angle I am now looking at it the retinal image is of an ellipticalshape.

3. COLOUR AND BRIGHTESS CONSTANCY

This is where familiar objects retain their colour (or hue) in a variety of lighting conditions.Knowledge of the 'real' colour of the object means that it is still perceived as being that colour,regardless of the actual colour wavelength of the light that reaches the eye. Thus at night we stillperceive our red car as 'red' even under the night light.

4. LOCATION CONSTANCY

Knowledge that objects don't generally move means that things are seen as remaining in thesame place even when the observer moves around and the retinal image changes. As we movearound the environment we produce a constantly changing pattern of retinal images yet we donot perceive the world as spinning; this is due to kinaesthetic feedback. The brain subtracts theeye movement commands from the resulting changes on the retina and this helps to keep us andthe environment stable.

15

VISUAL ILLUSIONS

Constancy scaling seems to happen automatically; it doesn't require us to think about it.Normally the visual system receives accurate information about the size and distance of objectseg by the use of distance cues. Psychologists have been particularly interested in instances wherethe visual system makes errors as it does when conflicting information is received. Look at thePonzo Illusion:

In the Ponzo Illusion the top horizontal line looks longer than the line below it despite the factthat they are the same size and therefore must have the same retinal image. One explanation forwhy we perceive the top line as longer (paradoxically, we usually perceive the nearest line aslonger) is that these type of illusions contain false depth cues which trigger the size constancymechanism inappropriately. In the case of the Ponzo Illusion the converging lines are the falsedepth cues which suggest the top line is further away than the bottom line. The eye is tricked bythe depth cues in the converging lines into 'thinking' the top line is further away. An object whichis further away produces a smaller retinal image. The size constancy mechanism thereforeexpands the perceived size of the top line. In most cases automatic triggering of the sizeconstancy mechanism by a simple depth cue would result in an accurate perception. It is whenperception goes wrong that psychologists have been given insight into how the automatic scalingmechanism might operate.

Which horizontal linelooks longer?

16

ILLUSIONS IN THE NATURAL WORLD

Illusions are relatively rare in the natural world and so there has been no evolutionary pressure toproduce a perceptual system that overcomes this. One illusion that does occur in the naturalworld is the 'Moon Illusion'.

The moon (and sun) appears larger when low down on the horizon than when high in the sky.The size of the retinal image does not change. You can 'black out' the moon by holding a 1/4inch disc at arm's length, whether the moon is high or low in the sky. Why then does it appear tobe much larger when it is near the horizon? One explanation is constancy scaling. When themoon is high in the sky, there is no depth/distance information visible, so you see the moon at itscorrect (ie retinal image) size. However, when the moon is low down, near the horizon, depthcues operate. The horizon is as far away as it is possible to see so constancy scalingautomatically increases the size. If the moon (whose retinal image remains the same) appears tobe further away when it is closer to the horizon then we conclude it must be larger.

FOUR TYPES OF ILLUSIONS

Gregory, 1983 has identified 4 types of illusions:

Distortions (eg Ponzo Illusion) where we make a perceptual mistake;

Ambiguous figures (eg the Necker Cube) where the same input results in different perceptions;

17

Paradoxical Figures (eg the Penrose Trident) we assume this is a 3-dimensional object;

Fictions (eg the Kanizsa triangle) where we see what is not in the stimulus, ie a second triangle.

ATTENTION

"Everyone knows what attention is. It is taking possession by the mind, in clear and vivid form,of one of what seems several simultaneously possible objects or trains of thought.... it implieswithdrawal from some things in order to deal effectively with others "

W James 1890.

Do we attend simultaneously to everything in our environment or do we attend selectively tocertain types of information at any one time? The topics of perception and attention merge into

18

each other since both are concerned with the question of what we become aware of in ourenvironment. We can only perceive things we are attending to, we can only attend to things weperceive. Therefore some of the same questions are at issue, notably the question of whetherattention is governed by 'bottom-up' sensory processes (as proposed by the informationprocessing models) or whether 'top-down' processes like memory/expectations etc play animportant part in attention.

HUMANS AS INFORMATION PROCESSORS

When we are selectively attending to one activity, we tend to ignore other stimulation, althoughour attention can be distracted by something else, like the telephone ringing or someone usingour name. Psychologists are interested in what makes us attend to one thing rather than another(selective attention); why we sometimes switch our attention to something that was previouslyunattended (e.g. Cocktail Party Syndrome), and how many things we can attend to at the sametime (attentional capacity).

One way of conceptualising attention is to think of humans as information processors who canonly process a limited amount of information at a time without becoming overloaded. Broadbentand others in the 1950's adopted a model of the brain as a limited capacity informationprocessing system, through which external input is transmitted.

INFORMATION PROCESSING SYSTEM

STIMULUS RESPONSE

Information processing models consist of a series of stages, or boxes, which represent stages ofprocessing. Arrows indicate the flow of information from one stage to the next.

Input processes are concerned with the analysis of the stimuli.

Storage processes cover everything that happens to stimuli internally in the brain and caninclude coding and manipulation of the stimuli.

Output processes are responsible for preparing an appropriate response to a stimulus.

ATTENTION THEORIES are concerned with how information is selected from incomingstimuli for further processing in the system - therefore operate at the input processes end of themodel.

Basic Assumptions of the Information Processing Approach to CognitiveProcesses

The information processing approach is based on a number of assumptions, including:

(1) information made available by the environment is processed by a series of processingsystems (eg attention, perception, short-term memory);

Inputprocesses

Storageprocesses

Outputprocesses

19

(2) these processing systems transform or alter the information in systematic ways;

(3) the aim of research is to specify the processes and structures that underlie cognitiveperformance;

(4) information processing in humans resembles that in computers.

A number of Models of attention within the Information Processing framework have beenproposed including:

Broadbent's Filter Model (1958), Treisman's Attenuation Model (1964) Deutsch and Deutsch'sLate Selection Model (1963)

and these will be outlined and evaluated. However, there are a number of evaluative points tobear in mind when studying these models, and the information processing approach in general.These include:

1. The information processing models assume serial processing of stimulus inputs.

Serial processing effectively means one process has to be completed before the next starts.

Parallel processing assumes some or all processes involved in a cognitive task(s) occur at thesame time.

Attention

Selective Attention Divided Attention

(Processes only oneinput)egBroadbent/TreismanDichotic ListeningTasks

(Processes all inputs)

eg Kahneman DualTask Experiments

Auditory Visual Task Similarity Practice

20

There is evidence from dual-task experiments (examples are given later) that parallel processingis possible. It is difficult to determine whether a particular task is processed in a serial or parallelfashion as it probably depends (a) on the processes required to solve a task, and (b) the amountof practice on a task. Parallel processing is probably more frequent when someone is highlyskilled; for example a skilled typist thinks several letters ahead, a novice focuses on just 1 letterat a time.

2. The analogy between human cognition and computer functioning adopted by the informationprocessing approach is limited. Computers can be regarded as information processing systemsinsofar as they:

(i) combine information presented with stored information to provide solutions to avariety of problems, and(ii) most computers have a central processor of limited capacity and it is usually assumedthat capacity limitations affect the human attentional system.BUT -(i) the human brain has the capacity for extensive parallel processing and computersoften rely on serial processing;(ii) humans are influenced in their cognitions by a number of conflicting emotional andmotivational factors.

3. The evidence for the theories/models of attention which come under the informationprocessing approach is largely based on experiments under controlled, scientific conditions.Most laboratory studies are artificial and could be said to lack ecological validity. In everydaylife, cognitive processes are often linked to a goal (eg you pay attention in class because youwant to pass the examination), whereas in the laboratory the experiments are carried out inisolation form other cognitive and motivational factors. Although these laboratory experimentsare easy to interpret, the data may not be applicable to the real world outside the laboratory.More recent ecologically valid approaches to cognition have been proposed (eg the PerceptualCycle, Neisser, 1976).

Attention has been studied largely in isolation from other cognitive processes, althoughclearly it operates as an interdependent system with the related cognitive processes of perceptionand memory. The more successful we become at examining part of the cognitive system inisolation, the less our data are likely to tell us about cognition in everyday life.

4. The Models proposed by Broadbent and Treisman are 'bottom-up' or ‘stimulus driven’models of attention. Although it is agreed that stimulus driven information in cognition isimportant, what the individual brings to the task in terms of expectations/past experiences arealso important. These influences are known as 'top-down' or 'conceptually-driven' processes. Forexample, read the triangle below:

Eg Shadowing, fateof unattendedstimuli

Task DifficultyEg effects onautomaticity

Parisin the

the Spring

21

Expectation (top-down processing) often over-rides information actually available in thestimulus (bottom-up) which we are, supposedly, attending to. How did you read the text in thetriangle above?

MODELS OF ATTENTION

BOTTLENECK MODELS OF ATTENTION

A bottleneck restricts the rate of flow, as, say, in the narrow neck of a milk bottle. The narrowerthe bottleneck, the lower the rate of flow. Broadbent's, Treisman's and Deutsch and DeutschModels of Attention are all bottleneck models because they predict we cannot consciously attendto all of our sensory input at the same time. This limited capacity for paying attention is thereforea bottleneck and the models each try to explain how the material that passes through thebottleneck is selected.

BROADBENT’S FlLTER MODEL

Donald Broadbent is recognised as one of the major contributors to the information processingapproach, which started with his work with air traffic controllers during the war. In that situationa number of competing messages from departing and incoming aircraft are arriving continuously,all requiring attention. The air traffic controller finds s/he can deal effectively with only onemessage at a time and so has to decide which is the most important. Broadbent designed anexperiment (dichotic listening) to investigate the processes involved in switching attention whichare presumed to be going on internalb in our heads.

Broadbent argued that information from all of the stimuli presented at any given time enters asensory buffer. One of the inputs is then selected on the basis of its physical characteristics forfurther processing by being allowed to pass through a filter. Because we have only a limitedcapacity to process information, this filter is designed to prevent the information-processingsystem from becoming overloaded. The inputs not initially selected by the filter remain briefly inthe sensory buffer, and if they are not processed they decay rapidly. Broadbent assumed that thefilter rejected the non-shadowed or unattended message at an early stage of processing.

Broadbent (1958) looked at air-traffic control type problems in a laboratory.

22

Broadbent wanted to see how people were able to focus their attention (selectively attend), and todo this he deliberately overloaded them with stimuli - they had too many signals, too muchinformation to process at the same time. One of the ways Broadbent achieved this was bysimultaneously sending one message (a 3-digit number) to a person's right ear and a differentmessage (a different 3-digit number) to their left ear. Participants were asked to listen to bothmessages at the same time and repeat what they heard. this is known as a 'dichotic listening task'.

In the example above the participant hears 3 digits in their right ear (7,5,6) and 3 digits in theirleft ear (4,8,3). Broadbent was interested in how these would be repeated back. Would theparticipant repeat the digits back in the order that they were heard (order of presentation), orrepeat back what was heard in one ear followed by the other ear (ear-by-ear), He actually foundthat people made fewer mistakes repeating back ear by ear and would usually repeat back thisway.

SINGLE CHANNEL MODEL

Right Ear756

Left Ear483

1. Order of presentation745863

(ii) Ear by ear756483

23

Results from this research led Broadbent to produce his 'filter' model of how selective attentionoperates. Broadbent concluded that we can pay attention to only one channel at a time - so his isa single channel model.

In the dichotic listening task each ear is a channel. We can listen either to the right ear (that's onechannel) or the left ear (that's another channel). Broadbent also discovered that it is difficult toswitch channels more than twice a second. So you can only pay attention to the message in oneear at a time - the message in the other ear is lost, though you may be able to repeat back a fewitems from the unattended ear. This could be explained by the short-term memory store whichholds onto information in the unattended ear for a short time.

Broadbent thought that the filter, which selects one channel for attention, does this only on thebasis of PHYSICAL CHARACTERISTICS of the information coming in: for example, whichparticular ear the information was coming to, or the type of voice. According to Broadbent themeaning of any of the messages is not taken into account at all by the filter. All SEMANTICPROCESSING (processing the information to decode the meaning, in other words understandwhat is said) is carried out after the filter has selected the channel to pay attention to. Sowhatever message is sent to the unattended ear is not understood.

BROADBENT’S FILTER MODEL

Because we have only a limited capacity to process information, this filter is designed to preventthe information-processing system from becoming overloaded. The inputs not initially selectedby the filter remain briefly in the sensory buffer store, and if they are not processed they decayrapidly. Broadbent assumed that the filter rejected the non-shadowed or unattended message atan early stage of processing.

EVALUATION OF BROADBENT'S MODEL

(1) Broadbent's dichotic listening experiments have been criticised because:

(a) The early studies all used people who were unfamiliar with shadowing and so found it verydifficult and demanding. Eysenck & Keane (1990) claim that the inability of naive participants toshadow successfully is due to their unfamiliarity with the shadowing task rather than an inabilityof the attentional system.

(b) Participants reported after the entire message had been played - it is possible that theunattended message is analysed thoroughly but participants forget.

Sensese.g. eye,ear

Short TermMemoryStore

FILTER Selected inputfor attention.

Selection on thebasis of physicalcharacteristics only.

Inputchannels

24

(c) Analysis of the unattended message might occur below the level of conscious awareness. Forexample, research by von Wright et al (1975) indicated analysis of the unattended message in ashadowing task. A word was first presented to participants with a mild electric shock. When thesame word was later presented to the unattended channel, participants registered an increase inGSR (indicative of emotional arousal and analysis of the word in the unattended channel).

More recent research has indicated the above points are important: eg

Moray, N (1969) studied the effects of practice. Naive subjects could only detect 8% of digitsappearing in either the shadowed or non-shadowed message, Moray (an experienced 'shadower')detected 67%.

2. Broadbent's theory predicts that hearing your name when you are not paying attention shouldbe impossible because unattended messages are filtered out before you process the meaning -thus the model cannot account for the 'Cocktail Party Phenomenon'.

3. Other researchers have demonstrated the 'cocktail party effect' under experimental conditionsand have discovered occasions when information heard in the unattended ear 'broke through' tointerfere with information participants are paying attention to in the other ear. For example, Gray& Wedderburn (1960) found that students could put material from both ears together so that itmade sense. This implies some analysis of meaning of stimuli must have occurred prior to theselection of channels. In Broadbent's model the filter is based solely on sensory analysis of thephysical characteristics of the stimuli.

GRAY & WEDDERBURN'S EXPERIMENT

Gray & Wedderburn found that participants were able to give a category by category response(ie one that made sense of the material heard, as shown in the diagram below). Broadbent's FilterModel predicts this would not be possible. It is now certain that the unattended message can beprocessed far more thoroughly than was allowed for in Broadbent's theory.

Right EarJack

2Jill

Left Ear1

and3

Possible Responses

25

ANNE TREISMAN’S (1964) ATTENUATION MODEL

Selective attention requires that stimuli are filtered so that attention is directed. Broadbent'smodel suggests that the selection of material to attend to (that is, the filtering) is made early,before semantic analysis. Treisman's model retains this early filter which works on physicalfeatures of the message only. The crucial difference is that Treisman's filter ATTENUATESrather than eliminates the unattended material. Attenuation is like turning down the volume sothat if you have 4 sources of sound in one room (TV, radio, people talking, baby crying) you canturn down or attenuate 3 in order to attend to the fourth. The result is almost the same as turningthem off, the unattended material appears lost. But, if a nonattended channel includes yourname? for example, there is a chance you will hear it because the material is still there.

TREISMAN'S ATTENUATION

Treisman agreed with Broadbent that there was a bottleneck, but disagreed with the location.Treisman carried out experiments using the speech shadowing method. Typically, in this methodparticipants are asked to simultaneously repeat aloud speech played into one ear (called theattended ear) whilst another message is spoken to the other ear.

In one shadowing experiment, identical messages were presented to two ears but with a slightdelay between them. If this delay was too long, then participants did not notice that the samematerial was played to both ears. When the unattended message was ahead of the shadowedmessage by upto to 2 seconds, participants noticed the similarity. If it is assumed the unattendedmaterial is held in a temporary buffer store, then these results would indicate that the duration ofmaterial held in sensory buffer store is about 2 seconds.

In an experiment with bilingual participants, Treisman presented the attended message in Englishand the unattended message in a French translation. When the French version lagged onlyslightly behind the English version, participants could report that both messages had the same

Ear by EarJack 2 Jill

1 and 3 Category by Category1 2 3

Jack and Jill

Order of PresentationJack 12 andJill 3

Sensese.g. eye,ear

AttenuatingFilter

SemanticAnalysisFilter

Selected inputfor attention.

Inputchannels Inputs, including

attenuated inputsare passed on forsemantic analysis.

26

meaning. C1early, then, the unattended message was being processed for meaning andBroadbent's Filter Model, where the filter extracted on the basis of physical characteristics only,could not explain these findings. The evidence suggests that Broadbent's Filter Model is notadequate, it does not allow for meaning being taken into account.

Treisman's ATTENUATION THEORY, in which the unattended message is processed lessthoroughly than the attended one, suggests processing of the unattended message is attenuated orreduced to a greater or lesser extent depending on the demands on the limited capacityprocessing system. Treisman suggested messages are processed in a systematic way, beginningwith analysis of physical characteristics, sy11abic pattern, and individual words. After that,grammatical structure and meaning are processed. It will often happen that there is insufficientprocessing capacity to permit a full analysis of unattended stimuli. In that case, later analyseswill be omitted. This theory neatly predicts that it will usually be the physical characteristics ofunattended inputs which are remembered rather than their meaning. To be analysed, items haveto reach a certain threshold of intensity All the attended/selected material will reach thisthreshold but only some of the attenuated items. Some items will retain a permanently reducedthreshold, for example your own name or words/phrases like 'help' and 'fire'. Other items willhave a reduced threshold at a particular moment if they have some relevance to the mainattended message.

EVALUATION OF TREISMAN'S ATTENUATION MODEL

1. Treisman's Model overcomes some of the problems associated with Broadbent's Filter Model,e.g. the Attenuation Model can account for the 'Cocktail Party Syndrome'.

2. Treisman's model does not explain how exactly semantic analysis works.

3. The nature of the attenuation process has never been precisely specified.

4. A problem with all dichotic listening experiments is that you can never be sure that theparticipants have not actually switched attention to the so called unattended channel.

EARLY VS LATE SELECTION MODELS OF ATTENTION

When does selectivity occur? Does it happen in the early stages of recognition - whenconstructing a description of the input - or during the later stages, when comparing the input'sdescriptions to those of stored objects? The issue is important because it concerns whether wecan selectively ignore something before we know what it means - EARLY SELECTION - oronly after we know its meaning - L ATE SELECTION.

Broadbent and Treisman agree that selection of a single channel occurs at an early stage beforerecognition processes begin and so their models are called EARLY SELECTION MODELS.

An alternative view is that information from all channels is transmitted to the semantic analysisrecognition stage and it is only after this that a selection is made. The general framework for alate selection theory of this kind was first proposed by Deutsch and Deutsch (1963) and was laterelaborated by Norman (1968).

DEUTSCH AND DEUTSCH’S LATE SELECTION MODEL (1963)

27

DEUTSCH AND DEUTSCH (1963) solved the problems posed by the Broadbent model in adifferent way to Treisman. Their model suggests that all inputs are subject to high level semanticanalysis before a filter selects material for conscious attention. Selection is therefore laterbecause it occurs after items have been recognised rather than before as in Broadbent's model.Selection is also 'top-down' as opposed to Broadbent's and Treisman's Models which are knownas 'bottom-up' in that an item which has relevance to you, your name for example, or is incontext, is likely to be selected. Material is identified or recognised, its relevance, value andimportance weighed and the most relevant is passed upwards for conscious attention.

DEUTSCH AND DEUTSCH'S PERTINENCE MODEL (1963)

DEUTSCH AND DEUTSCH (1963) proposed a more radical departure from Broadbent'sposition in their claim that all inputs are fully analysed before any selection occurs. Thebottleneck or filter is thus placed later in the information processing system, immediately beforea response is made. Selection at that late stage is based on the relative importance of the inputs.

EVALUATION OF THE DEUTSCH AND DEUTSCH (1963) MODEL

FOR THE MODEL

1. Some support for a late selection model is offered by research which shows an unattendedmessage in a dichotic listening task can affect behaviour even though the listener has noconscious awareness of hearing the unattended message. For example, Moray (1969) paired anelectric shock with a word over several trials so that the person became conditioned to produce adetectable change in GSR (Galvanic Skin Response) when the word was spoken. He found thatseveral of his participants produced a change in GSR when the word occurred in an unattendedmessage even though they were not aware of hearing it.

2. McKay (1973) using ambiguous words like 'bark' instructed participants to shadow anambiguous sentence while, in the unattended ear a word was played which could clarify themeaning of the sentence. Later, participants who were quite unaware at a conscious level of theword in their unattended ear, chose meanings for the ambiguous sentence they had shadowedwhich were in line with the unattended word.

Sensese.g. eye,ear

SemanticAnalysis(Unconscious)

ConsciousAttention

Output.

Inputchannels

Top DownFactors

28

3. More recent studies have also shown that under some circumstances unattended material mayreceive some degree of analysis. For example, Wexler (1988) found that a GSR response variednot only according to ear of presentation but also according to the personality of the listener,indicating that processing of unattended material is more complicated than the early modelssuggest.

AGAINST THE MODEL

1. However, the assumption made by Deutsch and Deutsch that all stimuli are analysedcompletely, but that most of the analysed information is lost immediately, seems ratheruneconomical.

2. The research to support the Deutsh and Deutsch Model can also be explained by Treisman'smodel. A word in the unattended ear could have a reduced threshold because of its relevance.Physiological evidence also supports Treisman. When a measure of brain-wave activity knownas the evoked potential is recorded, it typically shows that the initial response to the unattendedmessage is much weaker than the response to the attended message, suggesting attenuatedprocessing of the unattended message.

3. Treisman and Geffen (1967) asked participants to shadow one of two simultaneous messages,and at the same time monitor BOTH messages in order to detect target words. Detection wasindicated by tapping. According to Treisman's theory, detection on the unattended messageshould be less than the shadowed message, whereas Deutsch and Deutsch's Model would predictno difference (as both messages would be fully analysed). As Treisman's Model predicts,detection was significantly higher on the shadowed message.

Perhaps on grounds of economy and explanatory powers of the available experimentaldata, Treisman's is the model most appropriate at present.AUTOMATIC PROCESSING

Researchers interested in attention have suggested a distinction between AUTOMATIC andCONTROLLED processing (Posner ~ Snyder, 197S). The basic idea is that some mental andphysical processes are under an individual's conscious control, while others tend to occurautomatically, without conscious awareness or intention.

Right Ear (Attendedand Shadowed Ear)

“the bark was not likeanything she wasfamiliar with”

Left Ear (UnattendedEar)

Either (a) tree or (b) dog.

29

A frequently quoted example of this is learning to drive a car. When you first learn to drive, suchthings as steering, braking and changing gear, all require a great deal of concentration. Problemsoften arise for the learner driver when they are required to do two or more things at once egbrake and change down gear. Also, as any driving instructor will tell you, learner drivers canbecome so engrossed in such things as changing gear that they fail to attend to what is happeningon the road in front of them! Yet to drive competently frequently requires a driver to do two ormore things virtually simultaneously.

How does the transformation from learner to expert occur? The concepts of AUTOMATIC andCONTROLLED processing have been used to explain this transformation. The basic idea issimple - with practice, skills which initially required a considerable amount of attention becomevirtually automatic. The development of automatic processing has a major advantage in that itreduces the number and amount of things that we have to attend to consciously. Thus the scarceresource of conscious attention is released for other tasks.

However, psychologists such as Gleitman (1981) have pointed out that automatic processing canproduce interference which actually lowers performance on certain tasks. A classic example ofthis is the STROOP EFFECT, named after JR Stroop (1935) who devised a colour namingexperiment. The experiment involved participants naming colours as quickly as possible. In onecondition participants named patches of colour, in a second condition participants had to namethe ink colour in which words were printed, but the words themselves were colour names. Forexarnple, participants would see the word BLUE but it would be written in red ink and their taskwould be to say RED'. Stroop found that participants were much slower at naming the inkcolours when the stimuli were themselves colour words.

One explanation for the Stroop Effect is that we automatically process the meaning of words.Thus when a participant sees the word BLUE but is supposed to respond to the ink colour andsay RED, the name of the word is automatically processed. This interferes with the participants'ability to process and name the ink colour (RED), thus delaying their response. In particular ithas been suggested that the Stroop effect produces a 'mental race' between the 2 processesinvolved in naming colours, the reading response wins the race and slows the colour naming.The difficulty experiences in naming the ink colour of the colour words is therefore theconsequence of an overlearned skill, and cannot be brought under conscious control.

DIVIDED ATTENTION

Can we do two things at once?

TASK DIFFICULTY

An obvious factor determining how well we can perform two tasks together is their level ofdifficulty. However, a task that is difficult for one person may be straightforward for another (egwhen we first learn to drive). We should also consider the difficulty of each of the tasksseparately.

PRACTICE

30

While experienced drivers can converse and drive at the same time, learner drivers havedifficulty doing the 2 tasks. Spelke et al l976 demonstrated the value of practice with 2 subjects(Diane and John) who were given approximately 90 hours of training on a variety of tasks. Thestudents were first of all asked to read short stories for comprehension while writing down wordsat dictation. To begin with their reading speed and their handwriting during dictation bothsuffered substantially. After 30 hours of practice, however, their reading speed andcomprehension had both improved up to the levels they displayed when not taking dictation andtheir handwriting was also better quality.

TASK SIMILARITY

It may well be that the inability to report much about the non-shadowed message in theshadowing situation is due to the great similarity between the 2 inputs - both English prosepassages presented in an auditory fashion.

Allport et al 1972 found when 2 shadowing tasks were dissimilar - for example, the standardshadowing task and the task of learning pictorial information - 90% of the pictures wererecognised.

The extent to which 2 tasks can be performed successfully together seems to depend on a numberof factors:

3. Two dissimilar, highly practised and simple tasks can typically be performed well together,whereas

4. 2. two similar, novel and complicated tasks cannot.

Dual task experiments imply that some well-learnt skills are virtually automatic. Once a decisionhas been made to drive somewhere, the actual driving of the car goes into 'autopilot'. And yetsome form of unconscious monitoring of environmental requirements must be going on to enableus to deal with sudden emergencies.

KAHNEMAN'S CAPACITY THEORY OF ATTENTION

Kahneman (1973) proposed that there is a certain amount of ATTENTIONAL CAPAClTYavailable which has to be allocated among the various demands made on it. On the capacity side,when someone is aroused and alert, they have more attentional resources available than whenthey are lethargic. On the demand side, the attention demanded by a particular activity is definedin terms of MENTAL EFFORT; the more skilled an individual the less mental effort is required,and so less attention needs to be allocated to that activity. If a person is both motivated (whichincreases attentional capacity) and skilled (which decreases the amount of attention needed), heor she will have some attentional capacity left over.

People can attend to more than one thing at a time as long as the total mental effort required doesnot exceed the total capacity available. In Kahneman's model allocation of attentional resourcesdepends on a CENTRAL ALLOCATION POLICY for dividing available attention betweencompeting demands.

Once a task has become automatic it requires little mental effort and therefore we can attend tomore than one automatic task at any one time, e.g. driving and talking.

31

Kahneman's Capacity Model of Attention

(1) Attention is a central dynamic process rather than the result of automatic filtering ofperceptual input.

(2) Attention is largely top-down process as opposed to the Filter Models which suggest abottom-up process.

(3) The focus of interest is the way the central allocation policy is operated so as to shareappropriate amounts of attention between skilled automatic tasks and more dimcult tasks whichrequire a lot of mental effort.

(4) Rather than a one-way flow of information from input through to responses, attentioninvolves constant perceptual evaluation of the demands required to produce appropriateresponses.

EVALUATION OF KAHNEMAN'S MODEL

1. Cheng (1985) points out that when tasks have been learnt we change the way we process andorganise them, but this is not necessarily 'automaticity'. For example, if asked to add ten two'syou could add 2 and 2 to make 4, add 4 and 2 to make 6, add 6 and 2 to make 8, add 8 and 2 tomake 10 etc. Indeed young children when first learning arithmetic would do just this. When wehave more arithmetical knowledge and realise that adding ten two's is the same as multiplying2 x 10, the solution can be produced in one step. The answer is quicker because we haveprocessed the information differently, using different operations, not because we have added tentwo's 'automatically'.

2. A MAJOR PROBLEM with Kahneman's theory is that is does not explain how the allocationsystem decides on policies for allocating attentional resources tasks.

The need for a homunculus to make decisions is a weakness of a psychological theory, as whatmakes the homunculus make decisions - another little person inside him(her) ????