In the preceding chapter I concentrated on the resources available within the language which enabled writers to construct texts that were cohesive and coherent. I argued that cohesion was largely a textual property whereas coherence was something that had to be assumed by writers. Nevertheless, writers could construct messages with the reasonable expectation that readers would identify and interpret the linguistic signs in the ways they had intended. For example, where two sentences co-exist readers would judge that they were meant to be related in some way even if that relationship were not overtly signalled. Lexical choice and the coherence relations between sentences would tend to establish what the text was about and how writers were approaching this ‘aboutness’. However, texts typically are intended to convey messages within particular social and physical spaces, therefore a fuller theory was needed to show how writers could signal these kinds of relationships. I considered register theory in some detail because it seemed most adequate to this task. However, I noted a lack of clarity in the use of key terms, particularly mode. Nevertheless, when applied to any particular text a registerial analysis does seem to capture the motivations behind different writers’ selections from the lexicogrammatical range of English. However, because each text is a unique intervention into a continuing communicative activity, register theory can really only show how individual texts create a social space for themselves. If we want to account for the ways in which texts are perceived to be of the same kind in that they are fulfilling similar social purposes, we need a higher level of analysis.