8.2.3　Implications for methodology：combining textual and cognitive methods

This book employs a toolkit which is interdisciplinary and comprehensive.The research methods include semiotic analysis（Chapters 4-6），genre analysis（Section 3.2）and empirical methods such as eye-tracking technology（Section 7.1）and memory recognition test（Section 7.2），and qualitative research methods such as the survey and interview in the evaluation investigation（Section 7.3）.

Firstly，the biggest contribution in methodology should be the combination of the semiotic analysis with empirical methods for the reader's cognitive reaction.As discussed in Chapter 3，they have their own strengths and weaknesses and will beneficially complement each other.For example，the semiotic analysis can provide detailed and rich description of the textual patterns and thus help achieve a profound understanding of the issues underlying the surface of the texts.However，due to the deductive and heuristic nature of semiotic analysis，the assumptions derived cannot be empirically proved.On the other hand，the eye-tracking technology provides not only behavioral end products of our cognitive processes but also clues to the comprehension process through which a coherent understanding of the discourse is achieved.The beneficial combination of the two research approaches ensures that the theoretical assumptions are drawn on the basis of a thorough analysis of textual features on the one hand，and their validity and operational patterns can be supported empirically on the other.

Secondly，the use of a relatively large，principled collection of natural text as the basis of framework building and analyses is another advantage of this book.As Bateman（2009）proposes，the approach of corpus-based linguistics should be beneficial for the study of multimodal discourses.Multimodal corpora are also needed in order to carry out investigations that are empirically sound and through which carry the multimodality research forward.This book collects and uses 97 advertisements，including flyers，leaflets and brochures，from 14 English language schools or companies in Beijing with principled criteria.This representative collection of data ensures that the framework building and conclusions drawn from analyses and empirical research in this book are reliable.

Thirdly，the usefulness of eye-tracking technology in the field of discourse analysis has not been fully realized yet.So far this technology has been used in two main research areas—discourse processing and visual cognition.The former almost exclusively uses purely verbal texts as materials whereas the studies in the latter group are mainly conducted with visual type of data which have little to do with linguistics.There has been few researches with eye-tracking technology that focus on multimodal discourse analysis which bears much the same origins with linguistics and at the same time draws many insights from disciplines of visual genres，such as images and graphs.In this sense，this book is a great effort.The eye-tracking experiment in it yields significant outcomes on the processing and comprehension of multimodal discourse analysis and produces huge implications for the understanding of the dynamic nature of the coherence construction.

In summary，the combination of multiple research methods has yielded rewarding outcomes and useful implications for discourse analysis，especially multimodal discourse analysis.It is worthwhile for more attention and exploration in future research in related field.