f the experimental findings of studies such as that of Cherry, yet it fails to explain how people respond to information such as their own name in the unattended channel and what decides that the system should tune to one channel rather than the other. For that matter, Treisman (1960; sited in Eysenk, 2000) pointed out a number of flaws in Broadbents filter theory of attention. In one task her subjects shadowed the message to one ear whilst the unattended message contained random strings of words. However, whilst the task was in progress the channels were reversed, and the shadowed and random-word channels were swapped over. Her subjects were not aware that the meaningful message had been transferred to the channel that had previously been occupied by the random words and they continued to shadow it without interruption.Treisman (1964; sited in Eysenk, 2000) again showed that subjects do analyse information in the unattended channel at the semantic level. Her subjects were bilingual. They were given the task of shadowing a message in one channel whilst the unattended channel contained the same message but in their second language. Treisman found that subjects were aware that the unattended channel was repeating the shadowed message and this must mean that they were making a semantic analysis of the non-shadowed channel. To explain this Treisman proposed a model in which the unattended message is attenuated in much the same way that one would turn down the volume of the radio in order to concentrate on the message coming from the television. In Treismans model, instead of blocking the meaning of the unattended channel the selection device simply attenuate it, causing less overall disturbance to the primary high level analysis of information taking place in the attended channel. However Treismans model does not explain enough about the nature of the attenuator device, for instance how the attenuator controls attention and how the things i...
