THEORIES OF PERCEPTUAL ORGANISATION
To read up on perceptual organisation, refer to pages 73–88 of Eysenck’s A2 Level
Psychology.
Ask yourself
 Look out of the window (take the opportunity—you are not usually
encouraged to do this whilst revising!). Consider the immense complexity of
all the lines and colours you see. How do we effortlessly translate the simple
two-dimensional images that are received by the retina into a complex threedimensional coloured world, in which we can perceive distance, depth, and
movement?
 When we interpret a scene and perceive depth, do we simply use the many
cues that are provided in the environment (such as parallel lines seeming to
converge as they go into the distance), or do we need to actively interpret the
information?
What you need to know
OUR VIEW OF
THE WORLD—IS
IT BUILT IN OR
BUILT UP?

Bottom-up
(determine
d by
external
stimuli) or
top-down
(determine
d by
knowledge
and
expectation
s)
processes?
GIBSON’S THEORY
OF DIRECT
PERCEPTION—A
BOTTOM-UP
APPROACH







The optic array
The optic flow
Perception and
action
Invariants
Texture
gradients
Resonance
Affordances
CONSTRUCTIVIST
THEORIES—
GREGORY’S TOPDOWN/INDIRECT
THEORY



Expectations
Motivational
factors
Visual illusions
SYNTHESIS—
NEISSER’S
CYCLIC

The
addition
of nonoptical
informati
on
OUR VIEW OF THE WORLD—IS IT BUILT IN OR BUILT UP?
Theories of visual perception attempt to explain how the two-dimensional images
that impinge on our retinas are translated into complex three-dimensional images
and how, when the images received on the retina are constantly changing, we are
able to maintain a stable perceptual world that enables us to perceive depth, shape,
and colour with impressive accuracy.
Bottom-up or Top-down Processes?
When we look at any scene, we have two sources of information:


The actual sensory input that we receive from the scene itself.
The non-visual information stored in our brain about what we expect the
world to look like.
In line with these two sources of information there are two sets of processes that
operate during our perception of the world:
 Bottom-up processes involve information travelling “up” from the stimulus
to the brain via the sense organs. These processes are therefore dependent
on input from outside and are also referred to as data-driven processes
because the information (raw data) from the sense organs “drives”
perception.
 Top-down processes involve the brain “sending down” stored information,
which then affects the perception of information already in the system,
thereby enabling us to make sense of the input received. Top-down
processes are driven by a person’s knowledge and expectations; for this
reason they are also referred to as cognition-driven processes.
Some theories of perception assume that perception is indirect in the sense that
information has to be processed at a higher level and that we construct our own
perceived world. These are known as constructivist theories and emphasise the
importance of top-down processing. Other theories assume that we do not need to
use higher level processes in order to perceive our world because data from the
environment provides us with sufficient information to perceive the world in a
direct manner and they emphasise the importance of bottom-up processes. Such
theories are called direct theories.
REMEMBER
Constructivist = top down
Direct = bottom up
We will look at the direct theory of Gibson and the constructivist theories of Gregory
and Neisser.
GIBSON’S THEORY OF DIRECT PERCEPTION—A BOTTOM-UP APPROACH
The main theme of Gibson’s theory (1950, 1966, 1979, see A2 Level Psychology
pages 74–80) is that the sensations received by the visual system are so highly
organised and rich in information that we do not need to interpret them to make
them meaningful; we can perceive them directly. This means that there are no
intermediate stages between light reaching the retina and an animal’s response to it.
According to Gibson, all animals use information from the visual system to move
around their environment without needing to make use of internal hypotheses or
thought processes—perception is automatic. Because of the emphasis on
environmental stimuli, Gibson’s theory is also known as the theory of ecological
optics.
The Optic Array
The most important source of information is the optic array, which contains all the
visual information available at the retina. Gibson argued that as we move around the
environment, we effortlessly detect information from the whole of the optic array.
Since this optic array provides a great deal of unambiguous information, little or no
information processing is required. This information comes in many forms,
including optic flow patterns, movement, and texture gradients.
The Optic Flow
Gibson was particularly interested in the problems pilots experience when landing.
Imagine you are a pilot coming into land. The point you focus on, called the “pole”, is
static, while the remainder of your field of vision is moving. The further from the
pole an object or scene is, the quicker it appears to move. All of this information is
known as the optic flow pattern, and it provides unmistakable information on
direction, speed, and altitude.
Perception and Action
In the normal course of our daily lives, we may not be flying aeroplanes, but we are
moving around our environment. Gibson consistently argued that there are very
close links between perception and action. We do not perceive our world from a
static viewpoint but by moving around and interacting in it, and this supports the
idea that perception is entirely dependent on action.
When an observer moves, the entire optic array is transformed (see figure on page
74 of A2 Level Psychology). It is this changing scenery that provides crucial
information in the perception of the world. Gibson was critical of other theorists in
the field of perception for concentrating on how we perceive a static object (or even
glimmer of light) when we stare at it. He pointed out that this is simply not how we
perceive in ordinary everyday life.
Invariants
As an animal moves around in its natural environment, it identifies those aspects of
it that remain the same—that are invariant. These invariants provide essential
information about the layout of the environment. This makes the layout of objects in
space unambiguous, so they can be picked up directly by the perceiver. For a pilot,
the “pole” is an invariant.
Texture Gradients
Another invariant is texture gradient in the natural world. When we look at almost
any scene, the texture of things changes as they get further away (think of a sandy
beach, or brickwork); in essence, the texture appears finer in the distance than in
the foreground. Since texture gradient is a characteristic of almost any surface in the
natural world around us, this is yet another cue to perception. Again, it is an aspect
of the environment that is visible to the observer, not one that needs to be
constructed in the brain.
Resonance
Gibson believed that we are able to make use of the rich source of environmental
information because our nervous system has been specially attuned by evolution to
receive it. He drew an analogy to a radio in the same way that this resonates with the
information contained in the electromagnetic radiation so we are equipped to make
direct use of information from the environment without the need for constant
judgements and hypothesis testing, as implied by constructivist theories. Like a
radio, we operate as a single unit and if one part goes wrong, the system does not
work. In a similar way, Gibson argued, the nervous system works as a single unit
when perceiving.
Affordances
According to Gibson, we can only understand an animal’s perceptual system in the
context within which it has evolved, particularly in terms of the optic arrays to
which it has been exposed. Within this context, we attach meaning to the type of
visual information we receive. Constructivist theorists believe that these meanings
are attached by means of “top-down” processes, as a result of our memory and
expectations based on past experience. However, Gibson disagreed. He argued that
all the potential uses of an object, called affordances, are perceived directly. We
would, for example, “afford” a ladder the properties of being able to assist ascent or
descent, by considering the gaps between successive rungs in relation to the length
of our legs. The importance of action is again of great importance here. They
emphasise that our perception of affordances (our perception of the possibilities for
action offered by the environment) can only be accurately assessed if we move
around. The notion of affordances is therefore important in this theory because it
emphasises that all the information needed to make sense of the environment is
directly present in the visual input and that perception depends on action.
RESEARCH EVIDENCE FOR GIBSON’S DIRECT PERCEPTION THEORY
 Warren and Hannon (1988, see A2 Level Psychology page 76) produced films
consisting of patterns of moving dots simulating the optic flow that would be
produced if someone were moving in a given direction. As predicted by
Gibson, observers used the optic-flow information to make accurate
judgements of the direction in which they were heading.
 According to Milner and Goodale (1995, 1998, see A2 Level Psychology page
78), we have two visual perceptual systems, which often interact with each
other and so aren’t entirely separate. The vision-for-perception system is
used to decide that the animal in front of us is a cat or a buffalo or to admire a
beautiful painting. The vision-for-action system is used for visually guided
action. It resembles what Gibson had in mind when he argued that
perception and action are very closely linked and so supports Gibson’s
theory.
 Haart et al. (1999 see A2 Level Psychology pages 79–80) found the Müller–
Lyer illusion is only obtained when the vision-for-perception system is used.
There was now no illusion effect at all when the vision-for-action system was
used. This finding has also been obtained with other visual illusions. This
shows that the vision-for-action system generally operates in the way
proposed by Gibson.
RESEARCH EVIDENCE AGAINST GIBSON’S DIRECT PERCEPTION THEORY
 Van den Berg and Brenner (1994, see A2 Level Psychology page 77) provide
evidence that optic flow doesn’t sufficiently account for depth perception. We
only need one eye to use optic-flow information. However, judgements about
the direction in which observers were moving were more accurate when
they used two eyes rather than only one. This happened because of what is
known as binocular disparity—the slight difference in images on the retina of
each eye allows observers to obtain additional information about the relative
depths of objects.
 Rushton et al. (1998, see A2 Level Psychology pages 77–78) provide evidence
that perceived visual direction is more important than optic-flow
information. In their study they produced a conflict between perceived visual
direction (the direction in which observers seemed to be heading) and opticflow information. The conflict was achieved by the observers wearing prisms
that provided misleading information about visual direction but accurate
information about optic flow. If optic-flow information were all-important,
the prisms should have had no effect on observers’ direction of walking—
they should have walked straight towards the target. In fact, however,
observers walked along a curved path as would be expected if they were
focusing on perceived visual direction. This finding is bad news for Gibson’s
theory because it suggests that optic-flow information is less important than
he assumed.
 However, Gibson’s approach is threatened by the existence of a large number
of visual illusions in which perception is very inaccurate. In the Müller–Lyer
illusion, participants are asked to compare the lengths of two vertical lines
(see figure in A2 Level Psychology page 78). Most people think one of the
vertical lines is longer than the other, when in fact they are the same length.
In the Ebbinghaus illusion (see figure in A2 Level Psychology page 78), the
central circle surrounded by smaller circles looks larger than a central circle
of the same size surrounded by larger circles. In the Ponzo illusion (see figure
in A2 Level Psychology page 78), rectangle A looks larger than rectangle B
although they are actually the same length. These illusions would not occur if
perception was solely bottom-up as Gibson suggests.
 Menzel’s (1978, see A2 Level Psychology page 80) research shows Gibson’s
key omission: he ignored cognition. Chimpanzees were carried around a
field, and were shown the locations of 20 pieces of food. When each
chimpanzee was then released, it moved around the field picking up the food
efficiently. There could be no relevant information in the light reaching the
chimpanzees (because they were now moving independently rather than
being carried), and so they must have made use of stored information in
long-term memory to guide their search, which Gibson did not account for.
EVALUATION OF GIBSON’S DIRECT PERCEPTION THEORY
 It draws attention to the rich source of information available to the
perceiver. Gibson pointed to the amazingly rich source of stimulation that is
available to a moving perceiver and to the fact that the light reflected from
objects must be organised in some way. Previously, theorists had
concentrated only on the image on the retina, not the way in which this
changed as the observer moved. Gibson pointed out that we have far more
information than that at our disposal.
 It accounts well for everyday perceptions. Gibson’s theory accounts well
for the considerable accuracy of our everyday perception. To those who
argue that many visual illusions show that our perception is not accurate, his
reply was that such data are obtained because people are presented with
ambiguous, brief, or disjointed stimuli in very artificial conditions that have
little relationship to perception in everyday conditions. He had a point, as
research shows no visual illusions with the perception-for-action system.
 It takes account of evolution. Considering that all animals have evolved in
response to the demands of their particular environment, it is highly
probable that they do possess some in-built systems for receiving external
information that is important to their survival.
 An underestimation of the difficulties of perceiving the world. Even
though Gibson has shown that various perceptions (such as the perception of
depth) could be accomplished by external cues alone, that does not mean
that they are accomplished in this way. Gibson underestimated the difficulty
involved in detecting and using the aspects of the environment in order to
perceive the world correctly. The processes involved in identifying invariants
in the environment, in discovering affordances, and in producing resonance,
are all much more complex than was assumed by Gibson. None of these
achievements of visual perception happens as rapidly and automatically as
Gibson claimed.
 Gibson exaggerated the role of optic-flow information. This is not the
only factor that we use to move successfully around the environment. Such
information is sometimes used, but it is much less important than assumed
by Gibson, as there are other factors, including visual direction and binocular
disparity.
 The need for cognitive processes in perception of objects. Few cells or
neural networks have been found that pick up the very complex features of
the world. It may be that we have the innate capability to directly perceive
something very basic and essential for survival, such as depth, but we may
not have cells that respond to a particular object, such as a goose. Some
cognitive processes must need to operate when we identify such a creature, as
supported by Menzel’s (1978) research on chimpanzees, see above.
 Ignored the additional meaning attached to objects. Gibson provides a
valuable account of seeing (especially when related to action), but he did not
consider the additional meaning that is attached to what is seen.
CONSTRUCTIVIST THEORIES—GREGORY’S TOP-DOWN/INDIRECT THEORY
Gregory (1970, 1972, 1973, see A2 Level Psychology pages 80–85) is a present-day
constructivist theorist who argues that perception is an active and constructive
process in which top-down processing is an essential element. He believes that the
data (information) we derive from the images that impinge on the retinas are
insufficient for us to build up an accurate picture of the world. He refers to the data
obtained as “fragmentary scraps”. Gregory believes that various cognitive processes
are involved in perception. Schemas based on expectations are amongst the most
important of these.
Constructivists like Gregory suggest that our perception of the world is strongly
affected by our existing schemas. You have come across schemas in several parts of
the AS-level course (e.g. when studying memory). Schemas are like categories into
which we incorporate the information from our experiences. One example of a
schema in the world of visual perception is that of faces. Our “face” schema tells us
that however different one face is from another, they have certain common features,
such as a nose that protrudes, two eyes, a mouth, and so on. One of Gregory’s
favourite demonstrations of the role of schemas based on expectation is the “hollow
face illusion” in which people view a hollowed mask of a face (see photo in A2 Level
Psychology page 81). Although there is sufficient information to see the mask as
hollow, we still perceive a normal face. Our schema of a face overrides the evidence
of our own eyes and we accept an incorrect hypothesis.
According to Gregory, perception is a series of hypothesis-testing exercises. As we
encounter certain life experiences, we “call up” the most appropriate, the most
relevant, or the most recent schemas and use these to provide the context in which
to interpret our present experience. Most of the time these hypotheses are
accurate—they fit our expectations. Sometimes, however, the schema does not fit
the occasion and perception is made difficult or inaccurate.
Gregory’s Explanation of Visual Illusions
Gregory argued that we treat two-dimensional illusions as if they were threedimensional, even though we know they are only two-dimensional. Gregory
suggests the misapplied size–constancy theory explains several illusions.
Gregory suggests that when we look at the Müller–Lyer illusion and the “fins” out of
context, there are no cues to help us perceive their length so we draw on existing
hypotheses and expectations. These tell us that the figure with the fins diverging
represents an inside corner and the one with the fins converging represents an
outside corner. Since it is part of our expectation (according to Gregory) that outside
corners are usually further away than inside ones, when the vertical line is the same
length, the “outside” corner one is perceived as longer since it is perceived as
further away.
In a sense, we perceive a two-dimensional figure as if it were a three-dimensional
one. This may well be true since if these fins are presented as luminous figures in a
dark room, they do appear three-dimensional.
RESEARCH EVIDENCE FOR GREGORY’S TOP-DOWN/INDIRECT THEORY
 Evidence that expectations influence our everyday perception comes from
situations in which expectations produce errors in perception. For example,
proofreading often involves numerous mistakes. Daneman and Stainton
(1993, see A2 Level Psychology page 82) found that errors were more
common when readers were proofreading an essay handwritten by them
rather than an essay written by someone else. Their extreme familiarity with
their own writing meant that they focused less on possible errors in their
own essay.
 Bruner and Goodman (1947, see A2 Level Psychology page 82) asked rich and
poor children to estimate the size of coins. The poor children overestimated
the size of every coin more than did the rich children, which may reflect the
greater value placed on money by the poor children.

Ashley, Harper, and Runyon (1951, see A2 Level Psychology pages 82–83)
introduced an ingenious modification to the experimental design used by
Bruner and Goodman (1947). They hypnotised adult participants into
believing they were rich or poor. The size estimates of coins were
consistently larger when the participants were in the “poor” state.
RESEARCH EVIDENCE AGAINST GREGORY’S TOP-DOWN/INDIRECT THEORY
 Bruner and Goodman (1947) may not support expectation because there is a
simpler explanation. The rich children would have had more familiarity with
coins, and this may have made them more accurate in their size estimates.
 DeLucia and Hochberg (1991, see A2 Level Psychology page 84) reported
findings that are very hard (or impossible) to explain using Gregory’s theory.
They presented observers with three-dimensional 2-foot-high fins placed in a
line on the floor in an arrangement like that shown in the Müller–Lyer. The
usual illusion effect was obtained even though it was obvious that all the fins
were the same distance from the observers. This finding can’t be explained
by Gregory’s misapplied size–constancy theory.
EVALUATION OF GREGORY’S TOP-DOWN/INDIRECT THEORY
 Strong evidence that top-down processing influences perception.
Gregory and other indirect theorists have performed the valuable service of
showing that top-down processes shouldn’t be ignored and there is strong
evidence for this.
 Considers motivational and emotional factors. Gregory’s theory is wideranging in that it assumes that motivational and emotional factors can
influence perception, and so it is more holistic than other theories.
 Explains inaccuracies in perception. Misapplied size–constancy does
explain inaccuracies in perception. The artificiality of such evidence is a
weakness however; proofreading errors show that expectations do play a
role in real life.
 It does not apply to many other variations of the illusion. The most
obvious problem with the theory is that there are numerous variations on
the Müller–Lyer that produce just as strong an illusion, if not stronger, and
yet cannot be explained in terms of misapplying the rules of size constancy.
The illusion still works if the lines are horizontal, if they are absent, or if the
fins are replaced by squares, circles, or other shapes. None of these figures
could be said to represent the corner of a building.
 Artificiality of research may limit bottom-up processing and exaggerate
top-down-processing. The research involves the very rapid presentation of
visual stimuli. Brief presentation reduces the impact of bottom-up processes,
thus allowing more scope for top-down processes to operate. This is very
artificial because in real life we have longer than a fraction of a second to
look at an object.
 The information received is not as fragmented and inadequate as
Gregory states. Gregory points to the inadequacy and ambiguity of visual
information. For example, that an object in the distance can produce an
identical image on the retina as a small object in the foreground, so we
cannot judge the size of the retinal image alone. Hence, how do we perceive
depth if not on the basis of additional cognitive processes? However, as
Gibson pointed out, we have a wealth of other information in the optic array
available in addition to relative size: the texture gradient of the surroundings,
the height in the horizontal field, and the relative size of other objects, to
name but a few. The optic array we have at our disposal is far more complex
than Gregory proposed.
SYNTHESIS—NEISSER’S CYCLIC THEORY
Neisser (1976, see A2 Level Psychology page 85–87) has advanced a model that
attempts to integrate the constructivist and the direct approaches; visual perception
depends on both bottom-up and top-down processes. He considers perception to be
a cyclical activity, a continuous process that has no definite starting or finishing
point.
According to this model, people use their previous experience of the world to build
up schemas, as discussed under Gregory’s theory. These schemas lead us to
anticipate the information that is likely to be available to us (to take a simple
example, we would not expect to encounter a candle that was burning at the
bottom).
So what happens in ordinary perception? As we move around in our environment
we sample the perceptual world. This is “bottom-up” processing. If the stimulus is
significant we pay attention to it. This sets in motion a search for expected features
based on our existing schemas. This search involves movement and action, just as in
Gibson’s direct theory. If the expected features match the schemas (as they usually
do) we accept the hypothesis as correct. If not, we use another schema.
Thus, the perceptual cycle described by Neisser includes elements of bottom-up and
top-down processing. Bottom-up processing is represented by the sampling of
available environmental information that can modify the current schema. Top-down
processing is represented by the notion that schemas influence the course of the
information processing involved in perception.
The combination of bottom-up and top-down processing can be seen by the
following example of smiling provided by Neisser. When we see someone smile, the
light from the smiling face is picked up directly (bottom-up processing). However,
the context in which it occurs provides us with non-optical information such as the
mood the person is in: perhaps happy or just being polite. In order to ascertain this,
we use the top-down processing involved in the perceptual cycles.
EVALUATION OF NEISSER’S CYCLIC THEORY
 It is a plausible theory that provides a balance between top-down and
bottom-up processing. The emphasis on the interdependence of top-down



and bottom-up processing means that both sets of processes are allowed for
and one set is not emphasised to the exclusion of the other, as in the previous
two theories.
There is empirical evidence in favour of the theory. In accordance with
the theory, there is evidence that we perceive and remember schemarelevant objects better than schema-irrelevant ones. For example Biederman,
Glass, and Stacy (1973, see A2 Level Psychology page 86–87) demonstrated
that people recalled far more objects when they were in a familiar setting
(such as objects in a living room) than when just arranged in a pile. Being
able to use a familiar schema makes recall easier.
It is very vague. The theory is not very explicit in parts. For example, Neisser
has hypothesised an interaction between schemas and sensory cues, but has
said nothing about the neural activities (the processes going on in the
nervous system) that underlie this interaction.
It lacks detail. It is not clear how schemas are stored or how they interact
with incoming sensory information. How, for example, do incoming stimuli
modify the existing schemas?
SO WHAT DOES THIS MEAN?
We have seen that the emphasis of the direct theorists is on bottom-up processes,
whilst that of the indirect theorists is on top-down ones. In fact, it is likely that both
play a part in perception but not always to the same extent in different situations.
Eysenck (1993) suggests that visual perception is primarily determined by bottomup processes when the optic array is rich, but by top-down processes when it
becomes impoverished, a view that is consistent with Neisser. After all, it seems
intuitively obvious that we make more guesses (and more mistakes) when viewing
at dusk than in clear daylight.
There are, however, several questions to which neither the bottom-up nor the topdown approaches can supply a satisfactory answer. Exactly which elements of the
outside world do we use as the basis for construction and interpretation? Since we
do not have the neural mechanisms in our heads to pick up, for example, a whole
house or a dustbin, exactly what features of these stimuli do we use when we
perceive them?
The models we have considered in this section have provided a sound foundation
for the understanding of perception, but in order to understand the complex
processes that operate in the brain when sensation is translated into perception, we
now need more complex computational models.
OVER TO YOU
1. Describe and evaluate Gregory’s top-down/indirect theory of visual
perception. (25 marks)
2. Describe and evaluate Gibson’s bottom-up/direct theory of visual perception.
(25 marks)