2 Visual attention

2.1 Introduction

I introduced Section 1 by suggesting that the auditory system had a special problem: unlike the visual system, it needed processes which would permit a listener to attend to a specific set of sounds without being confused by the overlap of other, irrelevant noises. The implication of that line of argument was that vision had no need of any such system. However, although we do not see simultaneously everything that surrounds us, we can certainly see more than one thing at a time. Earlier, I wrote of attending to the sound of the computer in front of me, or of the birds to one side. I can do much the same visually. While keeping my eyes directed to the computer screen, I can either attend to the text I am typing or, out of the corner of my eye, I can be aware of the window and detect a bird when it flies past. If our eyes can receive a wide range of information in parallel, does that give the brain an attentional problem analogous to that of disentangling sounds? If visual information is handled in much the same way as auditory information seems to be, then we might expect the various items in the field of view to activate representations in memory simultaneously. That should lead to effects equivalent to those found in listening experiments; in other words, it might be possible to show that we are influenced by items which we did not know we had seen. We shall examine evidence of this shortly, but I shall first draw your attention to another area of similarity between hearing and seeing.

I pointed out at the start of Section 1.2 that, whereas we often have to follow one speech stream while ignoring others, we do not normally have to disentangle overlapping handwriting. However, it is worth bearing in mind that visual objects do overlap and hide parts of each other, and the brain certainly has the problem of establishing which components of the image on the retina ‘go together’ to form an object.

As with hearing, a variety of cues is available to help in directing visual attention. Taking my window again as an example, I can either look at the glass and see a smear (I really must get round to washing the window!), or I can look through that, to the magpie sitting chattering in the apple tree. In this kind of situation we use distance to help separate objects, in much the same way as we use direction in hearing. However, we can deploy our attention in a more sophisticated way than simply on the basis of distance, as can be demonstrated by another aircraft-related example.

Military jets are often flown very fast and close to the ground (to avoid radar detection), requiring the pilot to attend intently to the outside view. At the same time, there are various pieces of information, traditionally displayed on instruments within the cockpit, which the pilot must check frequently. To avoid the pilot having to look down into the cockpit, the ‘head-up display’ (HUD) was developed. This comprises a piece of glass, just in front of the pilot, in which all the vital information is reflected. The pilot can read the reflection, or look through it to the outside world, just as one can look at reflections in a shop window, or look through to the goods on display. With a simple reflection, the pilot would still have to change focus, like me looking at the smear or the bird. However, modern HUDs use an optical system which makes the information reflected in the display appear to be as far away as the outside scene. This saves valuable re-focusing time. Nevertheless, although the numerals in the HUD now appear to be located at the same distance as, say, a runway, pilots still have the sensation of focusing on one or the other; if they are reading their altitude they are relatively unaware of the scene on which it is superimposed. This suggests (as we shall see in more detail later) that visual attention can be linked to specific objects rather than to general regions of space, very much as auditory attention can follow a particular speaker's voice, or the sense of a sentence.