sweetleafnotebookMobile - sans fil

12 nov. 2013 (il y a 7 années et 11 mois)

241 vue(s)


To read up on perceptual development, refer to pages 88

101 of Eysenck’s

A2 Level

Ask yourself

How much do you think a newborn baby can see and hear?

What evidence is there that perceptual development is due to

What evidence is there that perceptual development is due to nurture?

What you need to know






Visual constancies:
size constancy;
shape constancy

Does the
improvement in
visual acuity over
time depend on
learning or

What cross
differences are
there in perception?

Is perception innate
or learned?


Depth Perception

Gibson and Walk (1960, see
A2 Level Psychology

page 89) developed the “visual
cliff”, which was a glass
top table, underneath which a check
patterned cloth was
positioned. A “shallow” side was created by positioning the cloth close to the glass
under one half of the table, and positioning the cloth fu
rther away on the other half
created a “deep” side. Infants between the ages of 6½ and 12 months were tested
for depth perception by placing them on the shallow side and seeing if they could be
tempted to cross to the deep side. Most infants could not be t
empted by their
favourite toy nor their mother’s voice, which suggests that depth perception
develops by approximately 6 months, as the babies were reluctant to cross to the
“deep” side.

However, Adolph (2000, see
A2 Level Psychology

page 89) argued that

development of depth perception is more complex than assumed by Gibson and
Walk (1960, see

A2 Level Psychology

page 89). She argues that infants do not acquire
knowledge (e.g. an association between depth information and falling) that
stops th
em from crossing the visual cliff; instead the infants’ knowledge is highly
. According to this model, infants learn how to avoid risky gaps when sitting,
but subsequently have to learn how to avoid such gaps when they are crawling.

Adolph (2000,
A2 Level Psychology
page 89) provided evidence that learning
with the visual cliff is specific to a given posture (e.g. sitting) and new learning is
needed when infants become more mobile. This leads to the conclusion that
learning to avoid gaps is

to a given posture (sitting and crawling) in younger
infants but becomes more

in older ones, because they are able to transfer
depth learning from crawling to walking.

However, research on retinal image size in infants under 2 weeks old (Bow
Broughton, & Moore, 1970, see
A2 Level Psychology

page 90) does provide evidence
of innate perceptual abilities. The two different size objects had the same retinal size
due to being different distances away from the infant, yet the infants were more
istressed by the objects that came closest to them, which shows that they were
able to perceive depth from a very young age.

Visual Constancies

Size constancy is when an object is perceived to be the same size regardless of
distance, and shape constancy
is when an object is perceived to be the same shape
regardless of orientation.

Size Constancy

A size constancy task (Bower, 1966, see
A2 Level Psychology

page 90) was developed
based on initially presenting infants with a 30cm cube placed 1 metre from the
Infants were then presented with the same cube placed 3 metres away and a 90cm
cube placed 3 metres away. Size constancy was shown to some extent as the infants
spent longer looking at the cube that had the same size as the initial cube despite the

it had a much smaller retinal image. But they failed to show complete size
constancy as they attended to cubes placed 1 metre away more then cubes of the
same size placed 3 metres away.

Shape Constancy

The habituation method involves presenting a stimul
us until the infant no longer
attends to it. Once habituation has occurred, a new stimulus is presented, and if the
infant attends to this then discrimination has occurred. Infants aged 3 months were
presented with a square that formed a trapezoidal shape
on the retina. This was
presented until habituation occurred. They were then presented with a real
trapezoid and their interest showed they had habituated to the real shape and not
the retinal image. This is evidence that shape constancy has developed in 3
olds despite inconsistency with the retinal image (Caron, Caron, & Carlson, 1979, see

A2 Level Psychology

page 91). This research shows that the infants habituated to the
shape of the object rather than to the object shape projected onto the re
This strongly suggests that these infants had developed at least partial shape


Limitations of the measures.

The habituation measure indicates
discrimination and the visual preference method indicates which stimuli are
red, but they do not reveal how the stimuli are perceived and
interpreted. Similarly, the visual cliff method shows that the deep side
disturbs infants but this does not reveal how they perceive depth. Thus, the
measures lack precision and, to some extent,

validity as they are not true
indicators of perceptual processes. But this, of course, is due to the difficulty
of investigating “unobservable” cognitive processes and so the measures are
useful, given these difficulties.

Mundane realism and ecological va

The stimuli shown to children
and the experimental set
up led to a lack of mundane realism. For example,
the stark contrast between the “deep” and the “shallow” side of the visual cliff
is rare in real
life environments, and the mother would be unl
ikely to tempt
the infant into a dangerous environment. Thus, external validity is limited.

Nature or nurture.

It would make evolutionary sense for perceptual
abilities to be biologically driven. However, the research evidence does not
clarify this as rese
arch is often based on infants a few months old who will
have learned some of the perceptual skills, but maturational factors may also
be involved.

The babies in the “visual cliff” study experienced discomfort. Do the
ends justify the means?


Infant Studies


Infants have very poor visual acuity. We can assess visual acuity by
presenting a display of alternating black and white lines, and then making the
nes progressively narrower until they can no longer be separated in vision.
In order for newborns to detect the separation of the lines, they need to be
about 30 times wider than is the case for adults (Braddick & Atkinson, 1983,
A2 Level Psychology

ge 92).

Another visual limitation in newborns results from the fact that they don’t
show accommodation. Their eyes have a fixed focal length for the first 3
months. This means that they can only focus clearly on objects at a given
distance from them (abou
t 8 inches).

Newborns lack colour vision as this does not begin to develop until 2 months.

Newborns are deficient in binocular vision, which involves using the
disparity or difference in the images projected on to the retinas of the two
eyes to assist dep
th perception. Teller (1997, see
A2 Level Psychology

92) reviewed evidence suggesting that binocular vision is first found in
infants between the ages of 3 and 6 months.

Banks, Aslin, and Letson (1975, see
A2 Level Psychology

page 93) studied
adults w
ho had had a problem with binocular vision because of having a
squint in childhood that was subsequently corrected. They found that the
later this was corrected, aged 4 plus, the less binocular disparity they had as
adults. These findings suggest that ther
e is a critical or sensitive period for
the development of binocular disparity during the early years of life in older
children and adults. However, visual acuity in adults with amblyopia
(impaired vision due to disuse of an eye with no obvious damage to i
t) can be
improved if they receive prolonged practice on a difficult visual task (Levi,
2005, see
A2 Level Psychology

page 93), indicating that learning can be

There is evidence infants are innately programmed to recognise human faces.
Morton and Johnson (1991, see
A2 Level Psychology

page 93) argued that
human infants are born with a mechanism containing information about the
structure of human faces. This mecha
nism is known as CONSPEC, because the
information about faces it contains relates to conspecifics (members of the
same species).

Johnson et al. (1991, see
A2 Level Psychology

page 93) found that newborns
in the first hour after birth showed more visual tra
cking of realistic faces
than of scrambled but symmetrical faces in which facial features were moved
to unusual locations within the face. This supports the influence of nature
over nurture in face perception.

However, Turati et al. (2002, see
A2 Level Psy
page 93) claim there is
nothing special about faces. Instead, newborns simply have a preference for
stimuli having more patterning in their upper than in their lower part. This
contradicts an innate bias for faces.

In a replication of research into

scrambled versus natural faces Simion et al.
(2006, see
A2 Level Psychology

page 94) also found that newborns prefer
scrambled faces with more elements in the top area to natural faces.
However, 3
olds preferred facial stimuli to scrambled faces. Th
leads to the conclusion that cognitive mechanisms specialised for faces
develop over the first few weeks or months of life.


Scientific support.

There is strong evidence that it does take several months
for some of the basic mechanisms involved in vision to function effectively.

Difficult to research infants.

It is of course difficult to research infants
because we cannot ask them what they see! T
his leads to a great deal of
controversy over whether they do actually have an innate bias for faces or
not. We have to rely on indirect evidence based on their behaviour, their gaze
patterns, and so on, from which not all conclusions may be valid.

cultural Studies


Segall, Campbell, and Herskovits (1963, see
A2 Level Psychology

page 94)
found that the Müller

Lyer illusion was dependent on experience of a
“carpentered environment”, which is an environment made up of mainly
l and horizontal lines and so mainly made up of rectangles. A rural
Zulu community did not show the illusion because they had no experience of
such an environment.

A comparison of Canadian Cree Indians who lived in cities and those who
lived in tepees show
ed that when asked to decide if two lines were in
parallel, the tepee dwellers were successful no matter what the angle of the
lines, but the city dwellers were better when the lines were horizontal or
vertical. This shows the importance of past experience

to visual perception
and so supports the role of learning in perceptual development (Annis &
Frost, 1973, see
A2 Level Psychology
page 95).

Allport and Pettigrew (1957, see
A2 Level Psychology

page 95) demonstrated
the importance of experience in their st
udy of a visual illusion. They created
a “window”, which was trapezoid in shape and fitted with horizontal and
vertical bars. It was then rotated in a circle, which gave the illusion that a
rectangle was moving backwards and forwards. However, only partici
who lived in cultures with rectangular windows experienced the illusion.
Those without such environmental experience, e.g. Zulus living in rural areas,
did not see the illusion because they were not predisposed to look for
rectangular shaped windows.

Gregor and McPherson (1965, see
A2 Level Psychology

page 95) provide
evidence that an absence of a carpentered environment does not necessarily
lead to an inability to see visual illusions. They compared two groups of
Australian Aborigines. One group live
d in a carpentered environment
whereas the other group lived in the open air and had very basic housing.
The two groups didn’t differ on either the Müller

Lyer illusion or the

vertical illusion. Cross
cultural differences in visual illusions may

depend more on training and education than on a carpentered environment.

Hudson (1960, see
A2 Level Psychology

page 96) investigated perception of
simple two
dimensional drawings. He found that non
Western societies had
difficulty perceiving three
onal scenes in such drawings, whereas
Western participants had no difficulty with depth perception.

Deregowski, Muldrow, and Muldrow (1972, see
A2 Level Psychology

page 96)
found that members of the Me’en in Ethiopia didn’t seem to recognise
drawings of an
imals on paper. This might suggest that they had poor ability
to make sense of two
dimensional representations. However, when the
Me’en were shown animals drawn on cloth (a material familiar to them), they
were generally able to recognise them.



Small samples are a common weakness of cross
research and as the target population is usually restricted the findings are
not very representative.

Ethnocentric tasks.

The two
dimensional picture perception is a culturally
ic task as these were based on Western drawing styles and used
Western materials. Deregowski, Muldrow, and Muldrow found that the Me’en
culture in Ethiopia did not respond to drawings on paper but did to drawings
on cloth, as this was a material more famil
iar to them. Thus, we must be
cautious in drawing conclusions on cultural differences in perception as the
findings may be biased by ethnocentrism. The “carpentered environment” is
also ethnocentric, as experience of an environment made up of mainly
al and horizontal lines depends on culture.

Difficult to interpret findings.

Different cultures vary in many ways and so
it is difficult to interpret the source of the cultural differences in perception.
This makes it hard to draw conclusions, particularly given the problems of

Linguistic barriers may decrease valid

The cross
cultural variations
may be due to language differences, which mean that perception is not
reported in the same way, but this does not mean that they are not perceived
in the same way. Thus, differences may be a consequence of language rather

than actual perception and so validity would be low, as the research would
not have measured what it set out to.

Weaknesses of self

report is vulnerable to researcher bias and
participant reactivity. For example, researcher expectancy or cuei
ng may
have yielded demand characteristics and so led the participants, which
means internal validity is questionable.

Mundane realism and ecological validity.

Most cross
cultural research has
focused on two
dimensional visual illusions, which are often We
e.g. the “carpentered environment”, and are not relevant to everyday
perception. This means the cultural differences found lack mundane realism.
This is a significant weakness as research could instead have focused on
different visual stimuli
, which may demonstrate cultural differences with
more relevance to real life. Instead the research may tell us very little about
cultural differences in everyday perception.




Evidence for the role of nature in perceptual development is that some
aspects, e.g. aspects of face perception; aspects of size and shape constancy;
dominance of global over local features, seem to be present at birth or very
shortly thereaf
ter. This suggests these are innate visual capacities or very
rapid learning.

Some aspects of visual perception emerge a few weeks after birth, e.g. visual
acuity and colour vision, and so seem to involve maturation, which means the
behaviours are instinct
ive but only emerge at a particular point in
development. Other aspects seem to develop in a critical or sensitive period,
e.g. binocular disparity. This suggests the involvement of genetic factors and

The substantial improvements in infants’ p
erceptual abilities in a fairly short
time period from birth points to the importance of maturation, i.e. that
perceptual abilities are genetically predisposed but require a certain amount
of physical development before they are realised. For example, infa
nts’ visual
acuity is about one
fortieth that of adults, and typically reaches the adult
level around the age of 12 months. It is likely that the large increase in
neurons in the visual cortex occurring during the first 6 months of life plays a
major role,

and so this physical maturation is needed before the predisposed
ability can fully develop.

Other aspects of visual perception take months to develop, e.g. depth
perception, and so this suggests the role of learning (
). The significant
al differences also support the role of learning as they suggest
that experience (i.e. learning) of the environment shapes perceptual

Evidence for the role of learning is provided by distortion studies. G.M.
Stratton (1896, see
A2 Level Psych

page 98) tested the effects of wearing
a lens on one eye that turned the world upside down (he kept his other eye
covered). At first everything looked unreal, but within 5 days he reported he
could write and walk around with relative ease. He took th
e lens off after 8
days and because everything was now a reversal of what he had grown used
to it also required some readjustment time. Thus shows the visual system is
not completely fixed because it was able to adapt to such a changed
environment. However
, the fact that Stratton’s visual system was fairly intact
after 8 days of distortion can be used to support an innate basis as clearly
any learning did not have a lasting effect.

Further evidence as to whether perception is innate or learned is provided
by readjustment studies, which involve individuals born blind because they
have cataracts who then have those cataracts removed. These provide useful
insights because their visual system has had no learning experience and so
any abilities they have initial
ly must be due to nature. To a great extent
nurture has been controlled! Gregory and Wallace (1963, see
A2 Level

page 99) documented the case history of SB. SB did not show
depth perception, nor could he recognise visual illusions, and in fact
ontinued to use his hands to “see” as touch was the sense he was most
accustomed to using.

Maurer, Lewis, and Mondloch (2005, see
A2 Level Psychology
page 99)
discussed the findings from people who had cataracts removed when they
were only 2 or 3 months o
ld. Many of their visual abilities were normal but
there was impairment in the more complex aspects that take the longest to
develop. They were poor at recognising faces from different viewpoints, at
discriminating between faces that only differed in terms

of the spacing
among facial features, and at detecting small differences between shapes.
These long
lasting deficits support the importance of learning from early
visual experience. However, the fact that more simple visual abilities (e.g.
between simple shapes, discriminating a face from a
scrambled image, and discriminating between faces that differed in various
ways from each other) were not impaired suggests that visual experience is
not necessary for their development.



It would be reductionist (oversimplified) to assume
perceptual abilities are due to either nature or nurture. Most perceptual
abilities depend on both nature and nurture, with the relative contributions
of each varying from one perceptual ability
to another. Our current
understanding of maturation suggests that the brain requires input to
mature. Thus, it is nature and nurture.

Difficult to interpret adult readjustment studies.

The evidence is difficult
to interpret because the absence of perceptua
l development does not
necessarily reject the influence of nature. The brain areas responsible for
visual processing may have degenerated or been used to facilitate the use of
touch. The use of touch may interfere with the development of visual
There is also the problem of motivation. For example, SB preferred
to rely on his sense of touch because he found vision difficult and frustrating.


The cross
cultural studies provide evidence of visual development
being specific to environment an
d this is of course adaptive because it makes
evolutionary sense for an individual to adapt to the requirements of the
environment as this will increase his/her chances of survival.

Perception requires understanding.
Some basic visual abilities are mainly
innate such as visual acuity and colour perception. However, understanding
what has been perceived is only possible with learning from the
environment, and so shows that meaningful perception depends very much
on nurture.


The infant studies support the influence of nature as these suggest that many
perceptual abilities are innate. But the cross
cultural studies show that perceptual
organisation is not universal and this cultural variation points to the role of nurture.
In c
onclusion we are
(nature) programmed to respond to
(nurture) and so an interactionist approach is needed to fully understand
perceptual development. It is impossible to separate out nature and nurture as the
innate schemas are influence
d by experience and we are predisposed to seek out
and adapt to experience. We are born with neurophysiological systems that need to
interact with the environment for maturation to take place.



Describe and evaluate one study of the devel
opment of perceptual abilities.
(10 marks)


Describe and evaluate one explanation of perceptual development.


Discuss the nature

nurture debate as it applies to perceptual development.