Since the publication of Perception and Communication by Broadbent in 1958 there have been a number of further investigations into attention. Broadbent’s influential paper on focused attention was felt by many to be the cornerstone paper in cognitive psychology. Although many of these investigations built on the work of or where undertaken in response to the work of Broadbent, it is still much debated as to whether the questions raised back in 1958 have actually acquired any answers or just become broader. However it is clear that theories concerning attention have progressed since 1958.
Broadbent was impressed by the findings of Cherry (1953), (as cited by French and Colman, 1995)1. The results of Cherry’s investigation showed that when two messages where played through headphones at the same time (one through each ear) subjects found it straightforward to listen to and even repeat (shadow) the message that they were instructed to attend to. The subjects had no idea what the other message had meant and couldn’t repeat it at all but could tell the sex and intensity of the voice they had heard. These findings influenced the theory of attention proposed by Broadbent.
According to Broadbent our information processing system has a limited capacity and therefore needs a filter to prevent an overload of too much information. He suggested that the brain temporarily retains information about all stimuli but that the information fades and disappears if it isn’t selected for further processing through being permitted to pass the filter. This early filtering theory related to Cherry’s work as the filter selects the message to be shadowed and not the other; hence no information about it is processed. However, the more people looked into Broadbent’s work the more problems for the filter theory began to arise.
Treisman (1964) responded to some of these problems that were becoming apparent by proposing an attenuation theory (as cited by French and Colman, 1995)2. This theory suggested that processing of unattended (non-shadowed/repeated) messages is typically attenuated in comparison to the amount of processing of attended messages. This didn’t mean to say that, in accordance with Broadbent, messages were either processed or not but that unattended messages would be processed to a certain level, starting with physical characteristics and then going on to grammatical configuration and meaning later.
So, physical characteristics will be picked up, like sex and voice intensity, but not necessarily the meaning, as was shown by Cherry. Treisman (1969) later presented a more inclusive treatment of the entire field of selective attention (as cited by Kahneman, 1973)3. Treisman suggested that many analysers in parallel were used to process a single input (like the messages played on the headphones), whilst the processing of two inputs is necessarily sequential.
In concluding that divided attention and parallel processing of two inputs was possible so long as they didn’t reach the same analyser, she departed totally from the filter theory and thus progressed on from the foundation work done by Broadbent. Further problems for the filter theory arose from the work of Allport, Antonis and Reynolds (1972), (as cited by Pashler, 1999)4 which showed that where tasks were made similar to one another, for example both messages being given in the same sense modality, the levels of recognition for the unattended message is low.
Allport et al. s subjects where instructed to shadow (repeat) an auditory message whilst at the same time trying to learn either words presented in auditory form or pictures. The results showed the memory of words presented in auditory form to be very low. On the other hand the recognition was much higher, with someone who had had a lot of practice being able to get up to 90% correct performance. This was another progression from Broadbent’s filter theory as it showed that two inputs could be processed at the same time so long as they were presented in different sense modalities from one another.
Work done by Underwood (1974) also showed that practice increased subject’s levels of recognition or detection (as cited by French and Colman, 1995)5. In test involving the detection of digits in the non-shadowed or shadowed message, subjects that had had no practice recorded levels of detection as low as 8% whereas those that had had practice managed levels nearer to 67% of the digits in the same message.
Although this is difficult to explain in terms of going with or against the filter theory it does show a continuation in the study of attention and was still related to the idea of early selection of stimuli. Not all of the progress made after the publication of Perception and Communication was to do with the idea of early selection. Over the years other theories developed, one of these was presented by Deutsch and Deutsch (1963) which (as cited by Kahneman, 1973)6 was an important alternative to filter-attenuation theory.
They argued that all messages reach the same level of analysis and that one is selected just prior to a response being formulated. Unlike the filter theory in which division of attention is impossible, Deutsch and Deutsch implied that detection of relevant signals should be easy even if the observer isn’t attending to the particular message. Although there is still a filter of sorts it is placed later in the processing system. This alternative view is difficult to evaluate but the evidence (as cited by French and Colman)7 is no more than partially correct.
A lack of flexibility was believed by some to be a serious limitation of earlier theories and sought to rectify this through their own studies and works. The inflexibility came from the staunch belief that at one point in the system of processing a selection took place instead of thinking that it could possibly be taking place at a number of different places depending on what was being processed. Findings of an investigation by Johnston and Wilson (1980) (as cited by French and Colman, 1995) showed that full processing took place if it was necessary but not if it isn’t.
Their tests showed that when subjects knew which ear they should expect to hear the target words in the relationship between the target words and non-target words (which where either related to the target word or random) had no effect on the subjects ability to detect the target words; thus suggesting that the non-target words weren’t processed. However when the subjects didn’t know which ear they were going to ear the target word trough the detection rate of the target words was higher if the non-target words were related and not completely random; suggesting that non-target words were processed for meaning.
So if it is necessary there is full processing of all the auditory inputs and if not then it doesn’t occur. Since the publication of Perception and Communication there have been a number of various different investigations, as discussed previously. However the progression of research into attention hasn’t been exclusively focused around the auditory modality, even though that may have been where it began. There have also been number of investigations into visual attention. The main argument here focuses on the idea that visual perception resembles a spotlight.
Everything in the centre is can be seen easily and clearly whereas things outside of the centre are less clear. It is also argued that the spotlight beams are alterable regarding the area they cover. These two argument are represented by Johnston and Dark (1985) and LaBerge (1983) respectively (as cited by French and Colman)8. Tests involving the identification of unclear words were made easier by the presentation of a related word in the central visual field than if the related word were presented outside the central area.
This suggested that more information could be gathered from inside the central visual field than from outside it. A different set of tests by LaBerge involved subjects being shown a five letter word and having to focus on all of, or just the middle part of, the word. The results suggested that the subjects were able to change the area covered by the spotlight by widening the beam, so to speak, thus increasing the central visual field. Since Broadbent’s 1958 work the study of attention has also branched off to look more specifically at individual areas of attention.
Things like divided attention, how good we are at doing two things at once, what causes absent-mindedness, when people fail to attend to what they are doing, vigilance tasks, attention and its effects on memory are just a few of the areas now looked at. So it can be seen that there has been a great deal of progress since the publication of Perception and Communication into the research of attention. The questions originally posed by Broadbent have been shown to be much broader and although there may seem to be no specific answers as such, much has been examined and discovered about attention.