The purpose of this research was to find out what Polish immigrants do with the variation that exists in the English language around them. Do they attach social meanings to it? Do they pick it up and copy it? Or do they do something else? To find out, we collected and analysed language data from British-born and Polish-born adolescents living in Edinburgh and London. Polish immigrants included in the study use non-standard features of English but the patterns associated with their use are not the same as those found among UK-born adolescents. That is, Polish adolescents are not copying native speakers wholesale; they are re-interpreting variation in English. This finding has also recently been reported in work on dialect contact and long-term contact. Our project adds weight to the importance of this principle. This project investigates the phenomenon of integration among migrant pupils and aims to find out what happens to immigrants when they come to a new country. One way to examine this is by using sociolinguistic methods to study the language variety migrants acquire once they have settled in a country. To do this, the project investigates the acquisition and sociolinguistic variation of local and non-local non-standard linguistic features among pupils of Polish descent in schools in London and Edinburgh. The large group of newcomers from Poland represents a unique chance to conduct a comparative study in two locales with different local dialects. It will show how immersion in differing contexts of language variation influences migrants' speech. Sociolinguistic methods are used to document how migrant pupils speak, and compare them to a local control group of age-matched teenagers. This reveals where the linguistic features Polish pupils use come from. By studying pupils' language attitudes and the language norms they're aware of, researchers can find out about the motivations behind different patterns of language use. Attitudes may influence what linguistic features pupils use, and how they indicate through language who they are becoming.

Linguistic production data were collected from 16 Polish migrants living in Edinburgh (8 males, 8 females) and 21 Polish adolescents living in London (8 males, 13 females). A comparable corpus of 21 Edinburgh and 24 London-born adolescents attending the same schools as the Polish adolescents was also collected to provide a benchmark for the types of ‘Edinburgh English’ and ‘London English’ to which these Polish adolescents are regularly exposed. Sociolinguistic face-to-face interviews were carried out between all participants and a female researcher from Edinburgh and London respectively. The primary tool used to elicit perception data was the Verbal Guise Technique (VGT). 8 university-educated females were recorded reading a short text about an animal rescue operation that was taken from Newsround (http://news.bbc.co.uk/cbbcnews/hi/uk/default.stm). Efforts were made to match the guise recordings for voice quality and speech rate but the guises all had different accents (representing Edinburgh English, London English Received Pronunciation, Scottish Standard English, Manchester English, Birmingham English, Newcastle English and Polish English). Subjective evaluations to these 8 guises were elicited from the adolescents using a semantic differential scale.