1. Introduction 1
2. Early work on attention 10
3. Selective report and interference effects in visual attention 26
4. The nature of visual attention 48
5. Combining the attributes of objects and visual search 69
6. Selection for action 92
7. Task combination and divided attention 108
8. Automaticity, skill, and expertise 122
9. Intentional control and willed behaviour 142
10. The problems of consciousness 165
11. Epilogue 185
References 188
Author index 205
Subject index 215
Introduction
What is attention?
Any reader who turns to a book with the word attention in the title, might be forgiven for
or thinking that the author would have a clear idea or precise definition of what attention
actually is. Unfortunately, attention is a concept that psychologists have been particularly
reluctant to define. Despite William James's (1890) oft-quoted remark that "Everyone
knows what attention is", it would be closer to the truth to say that "Nobody knows what
attention is" or at least not all psychologists agree. The problem is that attention is not a
single concept, but the name for a variety of psychological phenomena.
We can easily see some of the many varieties of attention in the common usage of the
word when we apply the same word to different situations and experiences. Let's take an
everyday example. While we are out walking in a wood, I tell you that I have just seen an
unusual variety of butterfly land on the back of the leaf in a nearby tree. I point out the
tree and whereabout the leaf is. and tell you to pay attention to it. Following my
instruction you are able to select one tree from many and then attend to a particular leaf,
rather than the tree itself, so presumably you and I share some common understanding of
what attention is. You continue to look carefully, hoping you will see the butterfly when
it moves out from behind the leaf. Now. you will try to keep your attention on that leaf so
as not to miss the butterfly when it appears. In addition, you will have some expectation
of what the butterfly will look like and how it may behave and you'll be monitoring for
these features. This expectation and anticipation will activate what psychologists call topdown
processes which will enable you to be more ready to respond if a butterfly appears
rather than some dissimilar animal—say, a caterpillar. However, if while you are
selectively focusing attention on the leaf an apple suddenly falls out of another part of the
tree, you will be distracted. In other words, your attention will be automatically captured
by the apple. In order to continue observing the leaf, you must re-engage your attention to
where it was before. After a time you detect the beautiful butterfly as it flutters round the
leaf: it sits a minute and you watch it as it flies away.
In this example we have a variety of attentional phenomena that psychologists need to
understand, and if possible explain, in well-defined scientific terms. We shall see that no
single term is appropriate to explain all the phenomena of attention and control even in
this visual task. Let's look at what you were asked to do. First of all. I asked you to attend
to a leaf. In order to do this simple task, there had to be some kind of setting up of your
cognitive system that enabled leaf rather than tree to become the current object of
processing: and one particular leaf was selected over others on the basis of its spatial
location. Once you are focusing on the leaf you are expecting butterfly-type shapes to
emerge and may occasionally think you have detected the butterfly if an adjacent leaf
flutters in the breeze. Here the perceptual input triggers, bottom-up, one of the attributes
of butterfly (fluttering) that has been primed by your expectations, and for a moment you
are misled. The idea of mental set is an old one. Many experiments on attention use a
selective set paradigm, where the subject prepares to respond to a particular set of stimuli.
The notion of selection brings with it the complementary notion of ignoring some stimuli
at the expense of those that are selected for attentional processing. What makes selection
easy or difficult is an important research area and has exercised psychologists for
decades. Here we immediately run into the first problem: is attention, the internal setting
of the system to detect or respond to a particular set of stimuli (in our examplebutterflies),
the same as the attention that you pay or allocate to the stimulus once it is
detected? It seems intuitively unlikely. Which of these kinds of attention is captured by
the unexpected falling apple? We already have one word for two different aspects of the
task. A second issue arises when the apple falls from the tree and you are momentarily
distracted. We said your attention was automatically drawn to the apple, so, although you
were intending to attend to the leaf and focusing on its spatial location, there appears to
be an interrupt process that automatically detects novel, possibly important.
environmental changes outside the current focus of attention and draws attention to them.
An automatic process is one that is defined as not requiring attention although, of course.
if we are not certain how to define attention, this makes the definition of automatic
processes problematic. Note now, another problem: I said that you have to return
attention to the leaf you were watching. What does this mean? Somehow, the temporary
activation causing the apple to attract attention can be voluntarily overridden by the
previously active goal of leaf observation. You have remembered what you were doing
and attention can then be directed, by some internal process or mechanism, back to the
original task. To say that you do this direction voluntarily tells us nothing: we might as
easily appeal to the little — man — in — the — head. or homunculus on which many theories seem
to rely.
To continue with the example: if you have to sustain attention on the leaf, monitoring
for the butterfly for more than a few minutes, it may become increasingly difficult to stop
your attention from wandering. You have difficulty concentrating: there seems to be
effort involved in keeping to the task at hand. Finally the butterfly appears: you detect it,
in its spatial location, but as soon as it flies away, you follow it, as if your attention is not
now directed to the location that the butterfly occupied but to the object of the butterfly
itself. The question of whether visual attention is spatially based or object based is
another issue that has recently begun to be widely researched.
Of course, visual attention is intimately related to where we are looking and to eye
movements. Perhaps there is nothing much to explain here: we just attend to what we are
looking at. However, we all know that we can "look out of the comer of the eye". If,
while you fixate your gaze on this* you are able to tell me quite a lot about the spatial
arrangement of the text on the page and what the colours of the walls are, it demonstrates
that it is not the case that where we direct our eyes and where we direct attention are one
and the same. Let's leave the example of looking for butterflies and consider other
modalities.
In the case of vision there appears to be an obvious limit on how much information we
can take in, at least from different spatial locations, simply because it is not possible to
look in two directions at once—although, of course, there is the question of how we
select from rom among spatially coincident information. Similarly, we cannot move in
two directions or reach for different places at the same time. Auditory attention also
seems to be limited. Even when there are several different different streams of sound
emanating from different different locations around us—the traffic outside, the hum of
the computer on the desk, the conversation in the room next door—we do not appear to
be able to listen to them all at once.
We all know that we can selectively listen to the intriguing conversation at the next
table in the restaurant even though there is another conversation continuing at our own
table. This is an example of selective auditory attention, and a version of the "cocktail
party" problem. Listening to a conversation in noise is clearly easier if we know
something about the content. Some words may be masked by other noises, but our topdown
expectations enable us to fill in the gaps: we say that there is redundancy in
language, meaning that there is more information present than is strictly necessary. We
make use of this redundancy in difficult conditions. If the conversation was of a technical
nature, on some topic about which we knew very little, there would be much less topdown
expectation and the conversation would be more difficult to follow. Although, we
may be intent on the conversation at the next table, a novel or important sound will
capture our attention, rather like the visual example of the apple falling out of the tree.
However, as occurs in vision, we are not easily able to monitor both sources of
information at once: if we are distracted, we must return our attention back to the
conversation.
Now we have another question: is the attention that we use in vision the same as that
we use in audition? Whilst it is difficult to do two visual or two auditory tasks
concurrently it is not necessarily difficult to combine an auditory and a visual task. What
about other modalities? While most research has been involved with vision and hearing,
we can of course attend to smells, tastes, sensations and proprioceptive information. To
date we know little about these areas and they will not concern us here. However, the
question of why some tasks do interfere with each other while others seem capable of
independence, and how we can share or divide attention, may crucially depend on the
modality of input and output as well as on the kind of information processing that is
required in the two tasks.
An important question in attentional research is why some tasks or kinds of processing
require attention but others do not. While you were looking for butterflies, we may have
been walking and talking at the same time. It is possible to continue eating dinner in the
restaurant at the same time as listening to a conversation. Walking, talking and eating
seem to proceed without attention—until the ground becomes uneven, a verbal problem
is posed or your peas fall off your fork. At these moments, you might find one task has to
stop while attention is allocated to the other. Consider learning a skill, such as juggling.
To begin with, we seem to need all our attention (ask your self which kind of attention
this might be) to throw and catch two balls. The prospect of ever being able to operate on
three at once seems rather distant! However, with practice, using two balls becomes easy: