The idea behind the code model is that when we want to communicate a message we encode the message into a signal and the hearer decodes the signal and understands the message. Sperber and Wilson’s goal is to show that in the case of metaphorical utterances, and in the case of most statements, the linguistic content of the utterance (simply obtained by decoding) is very limited compared to the intended content. Moreover, the intended content is not obtained simply by decoding the message, and so it is also necessary to make inferences. Firstly, it may be useful to consider how human beings behave in cases of non-verbal communication. In non-verbal communication sensory stimulation can lead to the recognition of an intended meaning even if there is no signal to be decoded. Let us consider an example: imagine that Tom offends May and May does not want to talk to him as a result. When Tom tries to make conversation May can do one of the following things: she can refuse to look at him or she can play with her smartphone. These actions do not encode any message by themselves, however Tom understands that May prefers to perform these actions rather than talking to him. In this case, the intended meaning of May’s actions is not codified, but Tom is able to understand what May wants to communicate or infer to him . The idea is that Tom has some background information (for example he knows what constitutes good relationship between him, May and he knows the character of May, etc.) and from this background information, together with the knowledge of May’s behaviour, he can infer that May does not want to talk to him. even when faced with a coded message (Sperber and Wilson 1995). Imagine the situation between Tom and May once more, in this instance May reacts to Tom’s invitation of conversation (3), she looks at Tom with anger, puts a finger in front of her mouth and emits the sound “Sssh!” In this case this signal transmits a coded message, but the coded message is much less precise than that communicated by May. Because we know that we can whisper a similar sound, for example, to ask to keep a secret. What Sperber and Wilson show with this type of example is that in order for Tom to correctly interpret what May is trying to communicate, he must make an inference, as simply decoding the verbal language is not enough. This does not mean that Sperber and Wilson think that verbal language is superfluous; indeed they believe that the presence of language makes our communication much richer and articulate. They only intend to highlight that to interpret what is communicated it is necessary to make inferences derived from contextual knowledge, as it is not enough to simply decode a message.