Analysing linguistic atlas data: the (socio-)linguistic context of h-dropping

This presentation will seek to illustrate how linguistic atlas data can be employed to obtain a better understanding of the mechanisms of linguistic variation and change. For this purpose, I will take a closer look at ‘H-dropping’ – a feature commonly found in various European languages and also widely used in varieties of British English. H-dropping refers to the non-realization of /h/ in initial position in stressed syllables before vowels, as for example, in hand on heart ['ænd ɒn 'ɑːt] or my head [mɪ 'ɛd]. It is one of the best-known nonstandard features in British English, extremely widespread, but also heavily stigmatised and commonly regarded as ‘uneducated’, ‘sloppy’, ‘lazy’, etc. It prominently appears in descriptions of urban accents in Britain (cf. Foulkes/Docherty 1999) and according to Wells (1982: 254), it is “the single most powerful pronunciation shibboleth in England”. H-dropping has frequently been analysed in sociolinguistic studies of British English and it can indeed be regarded as a typical feature of working-class speech. Moreover, H-dropping is often cited as one of the features that differentiate ‘Estuary English’ from Cockney, with speakers of the former variety avoiding ‘to drop their aitches’. The term ‘Estuary English’ is used as a label for an intermediate variety between the most localised form of London speech (Cockney) and a standard form of pronunciation in the Greater London area. After briefly discussing the history of H-dropping, the main emphasis will be put on geolinguistic aspects of the feature. 25 items from the Survey of English Dialects were analysed from a qualitative and quantitative point of view. Modern, urban dialects will also be examined. In spite of its sociolinguistic significance, there is relatively little information on the actual phonological process of H-dropping. My research results indicate that H-dropping is not necessarily such a straightforward, binary feature as is suggested by some textbooks or by single-item maps in linguistic atlases. At least in some varieties, the data reveal a more complex picture with variable realisations, including the use of semivowels. It is particularly relevant to analyse the geographical distribution of these realisations and to consider the relationship between rural and urban varieties. Finally, H-dropping can serve as a further example that geolinguistic data frequently provide interesting insights into the variation and history of a language and can advance our knowledge of (socio-) linguistic change.