The representation of people in tourism discourse has been recognised as one of problematic aspects of tourism (e.g. Jaworska, 2016; Wu, 2016). Previous research indicates that promotional materials and mass media represent guests as consumers (Ignatova, 2020b; Wu, 2016) relaxing and enjoying free time (Dann, 1996b), exclude hosts or portray them as servants, professional entertainers, cultural markers or part of the scenery (Ignatova, 2020b; Morgan & Pritchard, 1998). There are concerns that current representation maintains the asymmetric power relations between guests and hosts and can lead to social problems (Jaworska, 2016; White, 2007).
There is a lack of studies on the representation of people in user-generated discourse. Moreover, a systematic comparison of tourism discourse about different city destinations might shed light on how the socio-cultural background of the destination might influence such representation. Furthermore, the existing analytical frameworks for the multimodal analysis of the representation of people (Kress & van Leeuwen, 2021; Machin & Mayr, 2012; van Leeuwen, 2008) are not suitable for systematic comparison of relatively large datasets.
The current thesis aims to address this research gap by adapting the existing analytical frameworks and conducting a systematic corpus-assisted multimodal discourse analysis of the representation of people in 900 Tripadvisor travel reviews about three city destinations, Bangkok, Moscow, and London.
The findings indicate that different modes convey different representations. Apart from the patterns previously identified in institutional discourse, there are representations characteristic of online travel reviews (OTRs), namely, the differentiation of the reviewers and their companions from other guests and portrayal of guests as active organisers. The results also demonstrate that the socio-cultural background of destinations can be reflected in such patterns as the representation of martial art trainers as mentors and friends. There are indications that power relations reflected and shaped by OTRs are more nuanced and multidirectional.