Where to stay in the UK and in Italy: a comparative study of the language of holiday accommodation advertisements

As for my study, I decided to create both a multilingual comparable corpus and a parallel one. They can both be considered as small specialized corpora, because they are not large corpora, yet they are customized to the needs of my project. The former is composed by 55 different texts and 25.410 words, while the latter by 32 texts and 16.748 words. As the size of an individual text is relevant too, I decided to use full texts to compile my corpora, because interesting concepts, terms, patterns and contexts might be located in any section of them. As for the medium, texts employed in my corpora were originally prepared as written texts by experts for non-experts (i.e. by hotels promotion experts to travellers and tourists) and, obviously, their common subject is the tourist promotion of accommodation. With regard to the language, in my comparable corpus there are only original texts in both English and Italian, whereas my parallel corpus is composed of English or Italian original texts - depending on the location of the accommodation - and their translations. The authors of the texts cannot be identified as I found them on the web, however it can be said that my corpora are multi-author (i.e. written by a range of different authors). Multi-author corpora are likely to enable one to get a good picture of what terms and concepts are commonly used in that particular specialized language, thus avoiding to obtain linguistic output that is preferred by a particular writer. For the same reason, the publication date is difficult to determine too, but it could be argued that all texts are up-to-date, as it is important for the hotels to provide tourists with updated descriptions of their features and facilities.