The use of the English learning advertisements in this book is motivated by considerations in two respects—methodological and social.

In the respect of methodology，the collection of the English learning advertisements can provide a representative sample for study.As Bateman（2009）has suggested，the main task for the immediate future for multimodal discourse analysis must be to undertake detailed empirical analyses with respect to more representative samples of multimodal documents.He argues that the first move in such investigations is to consider larger and more representative selections of documents.

Instead of comparing or describing isolated artefacts as currently often the case，projects can be defined ranging in size from small，possibly group-based pieces of work up to substantial corpus collection efforts.In all cases，the documents collected need to be explicitly situated with respect to some selected set of dimensions drawn from multimodal genre—these may include combinations of target audience，particular type of document（glossy magazine，webpage，newspaper），historical period of production，area of content，communicative function（inform，persuade，etc.），and so on...detailed work of this kind is sorely needed in the area of multimodality at large，since all too often analysis remains“instantial”，i.e.，concerned with the individual case with little chance of reaching motivated generalizations.（Bateman，2009：63）

For the study on coherence construction in multimodal discourses，such a representative collection of samples is especially necessary because the pattern of coherence varies greatly with discourse genre.The analytical framework and methods designed for the analysis of one specific genre may not be directly applicable for another genre.

In the second respect，the English learning advertisement genre is chosen because of its social salience in contemporary China as a special type of English education.