Underwood (1974, cited in Eysenck, 2000) found that it is assumed incorrectly that the unattended message is always rejected early. Participants were asked to detect digits from either the shadowed or non-shadowed message; naive participants only detected 8% on the non-shadowed message, but Moray, an experienced researcher, detected 67% of them. This research provides evidence that if a person knows they are going to be asked about the non-shadowed message, then they are able to process information from it. Gray and Wedderburn (1960, cited in Parkin, 2000) found that both the attended and unattended channels could be processed if there is a meaning of the items and so a link can be made. Moray (1959) noticed that although most information in the unattended channel is ignored if the information is of high enough importance, for instance hearing your own name, participants could discriminate and orient to the information. These findings led many researchers to believe that all information must be processed to allow some information to breakthrough.
Deutsch and Deutsch (1963, cited in Gazzaniga, 2000) suggested the late selection model which is the concept that selection for processing and representation in conscious awareness occurs after both ignored and attended inputs are processed equally by the perception system and reach a stage of semantic analysis and encoding. There is little experimental evidence for late selection theory, but there is some by MacKay (1973, cited in Parkin, 2000). Participants shadowed a lexically ambiguous message, the content of the non-shadowed message could bias the meaning of the word "bank" (meaning either river bank or money bank). MacKay found a reliable biasing effect, recognition favoured the interpretation of the shadowed message suggested by the unattended message. However further research from MacKey, did not conclude that unattended messages are fully analysed.
Treisman and Riley (1969, cited in Parkin, 2000) suggested that if analysis of both channels is full then detection of letters within messages should be equally accurate on both channels, but they found that detection was twice as likely on the shadowed message, therefore arguing against the theory of late selection. Treisman (1964, cited in Eysenck, 2000) proposed a model to explain the phenomena of breakthrough. She suggested an early selection attenuation theory in which the filter reduces the analysis of unattended information. There are two important aspects of the theory: (1) stimulus analysis proceeds systematically through a hierarchy, first on physical cues, then syllabic pattern, specific words, analysis of individual words, grammatical structure and finally meaning; (2) the thresholds of all stimuli consistent with current expectations are lowered, partially processed stimuli on the unattended channel sometimes exceeds the threshold of conscious awareness, this accounts for breakthrough. Neurophysiological evidence from Woldroff et al (1993) supports the concept of early selection theories. Participants had to listen selectively to sequences of rapidly presented tones in one ear while ignoring tones of a different pitch in the other. Event-related potentials were recorded from both stimuli. There were greater ERPs to attended stimuli (20-50 milliseconds after stimulus onset). Therefore there is more processing of attended stimuli than unattended auditory stimuli starting from the initial activation of the auditory cortex. Focused auditory attention exerts a selective control over early processing in the auditory cortex of the supratemporal plane. The main problem with the attenuation theory is that the concept of attenuation is poorly specified (Parkin, 2000). Alternative theories suggest there are two attention systems. Posner (Eysenck, 2000) suggested we have one endogenous attention system that is controlled by intentions and is involved when central cues are presented, and an exogenous system that automatically shifts attention and is used when peripheral cues are presented.
There are a number of problems with the validity into research on auditory attention. What we attend to in the real world is often determined by our goals and emotional states, however in experimental conditions participants are told what they should attend to. Therefore the results are not valid because people cannot choose what to attend to. Shadowing is used in many of the experiments, however shadowing very rarely occurs in real life and does not replicate how people attend in the "cocktail party" situation where people are engaged in conversation; they do not only have to listen (and repeat) what they are attending to whilst ignoring other stimuli, but must respond to the conversation. The research ignores the fact that people can attend to either the external environment or the internal environment. In the real world we attend to 3D objects and people and make decisions based on them, in experimental conditions participants are told what to respond to which is usually 2D displays and make arbitrary responses. This means that the findings from auditory attention may not reflect real life and so the results may not be valid or real and the process may be more complex than any of the theories suggested (Tipper, 1992, cited in Eysenck, 2000).
In conclusion, the validity of both early and late selection models of attention is poor. The early selection theory by Broadbent had been disproved by a number of psychologists, the theory does not explain why sometimes information from an unattended channel can come into consciousness and therefore could not explain breakthrough. There are implicit and non-conscious effects of semantic processing of the ignored channel, therefore the early selection models are invalid. It also seems very unlikely that all information is fully processed (late selection model) because there is research to show that information from an attended channel is processed more accurately than an unattended channel. It seems likely that Treisman's attenuation theory is more valid, however there is still confusion over what attenuation is and how it works. The methods used in experimental research into auditory attention are unlikely to provide valid results because they does not replicate how people attend in real life situations.
Bibliography
Cave, K. R. and Bichot, N. P. (1999) Visuospatial attention: Beyond a spotlight model. Psychonomic Bulletin and Review, 6, 204-223.
Cherry, E. C. (1953) Some Experiments on the Recognition of Speech, with One and with Two Ears. The Journal of the Acoustical Society of America, 25, 975-979.
Eysenck, M. W. and Keane, M. T. (2000) Cognitive Psychology: A Handbook. Hove: Psychology Press.
Gazzaniga, M. S., Ivry, R. B. and Mangun, G. R. (1998) Cognitive Neuroscience: The Biology of the Mind. London: W. W. Norton & Company Ltd.
Moray, N. (1959) Attention in dichotic listening: Affective cues and the influence of instructions. Quarterly Journal of Experimental Psychology, 11, 56-60.
Parkin, A. J. (2000) Essential Cognitive Psychology. Hove: Psychology Press.
Reber, A. S. and Reber, E. (2001) The Penguin Dictionary of Psychology. London: Penguin Books Ltd.
Tipper, S. P. (1985) The Negative Priming Effect: Inhibitory Priming by Ignored Objects. The Quarterly Journal of Experimental Psychology, 37A, 571-590.
Woldorff, M. G., Gallen, C. C., Hampson, S. A., Hillyard, S. A., Pantev, C., Sobel, D. and Bloom, F. E. (1993) Modulation of early sensory processing in human auditory cortex during auditory selective attention. Proceedings of the National Academy of Sciences, 90, 8722-8726.