Speakers’ attitudes are considered a fundamental barometer for the vitality of a language (e.g., UNESCO, 2003). This, together with findings that implicit attitudes are generally stronger predictors of habitual and spontaneous behaviour (e.g., Perugini, 2005), raises two core questions: (1) which types
of attitudes and thus which attitude measurements are better predictors of language usage? (2) to what extent do different language policies feed different types of speakers’ attitudes? We explored these questions by measuring rates of spontaneous language usage and comparing them with attitudinal results from two methods that vary in degrees of implicitness: the Matched Guise Technique (Lambert et al., 1960) and the Implicit Association Test (Greenwald, McGhee & Schwartz, 1998) across two bilingual communities whose regional/minority languages receive radically different degrees of sociopolitical recognition: Lombard–Italian (Italy) and Welsh–English (UK). Results from 163 participants
aged between 24–36 years show that usage rates correlate with MGT status scores for Lombard but not for Welsh. The reverse holds for IAT scores, correlating with usage rates for Welsh but not Lombard.
We propose that these findings can be understood in view of the different socio-political support associated with the two languages: while strong support for Welsh led to its use becoming habitual and thus able to be captured by implicit attitude measurements, the use of Lombard has been discouraged for decades, and therefore younger speakers who choose to use it are making a more deliberate, conscious decision, resulting in behaviour that corelates with the less implicit measurements of the MGT. These results have important implications for the study of language attitudes, particularly for the measurement of attitudes as a proxy for language vitality. Specifically, they suggest that the degree to which an attitudinal measurement can predict linguistic behaviour depends partly on the social and political circumstances of the language at issue.