Published online by Cambridge University Press: 31 October 2012
Focus is regularly treated as a cross-linguistically stable category that is merely manifested by different structural means in different languages, such that a common focus feature may be realised through, for example, a morpheme in one language and syntactic movement in another. We demonstrate this conception of focus to be unsustainable on both theoretical and empirical grounds, invoking fundamental argumentation regarding the notions of focus and linguistic category, alongside data from a wide range of languages. Attempts to salvage a cross-linguistic notion of focus through parameterisation, the introduction of additional information-structural primitives such as contrast, or reduction to a single common factor are shown to be equally problematic. We identify the causes of repeated misconceptions about the nature of focus in a number of interrelated theoretical and methodological tendencies in linguistic analysis. We propose to see focus as a heuristic tool and to employ it as a means of identifying structural patterns that languages use to generate a certain number of related pragmatic effects, potentially through quite diverse mechanisms.
We would like to thank the audiences at the Workshop on Linguistic Typology and Language Documentation (ALT 7, Paris 2007), Workshop on Focus at the Syntax–Semantics Interface (Stuttgart 2008), Thematic Session on Information Structure (CIL 18, Seoul 2008), Workshop on Focus Marking Strategies and Focus Interpretation (DGfS 31, Osnabrück 2009) and at the Max Planck Institute for Psycholinguistics (Nijmegen 2010) for their input. We are particularly grateful to Robert Van Valin, Jr. for his comments on the earlier versions of this paper and for his support of our work (though he should not be assumed to agree with everything we say or to bear any responsibility for any of it), and to the Max Planck Institute, Nijmegen, for funding a research visit by the second-named author in 2010. Thanks are also due to Lila Magyari for her help with some of the Hungarian examples and to three anonymous referees of this journal for their helpful suggestions. The first-named author would like to acknowledge the financial support he received from the Max Planck Society and the Volkswagen Stiftung (DobeS Initiative) for his fieldwork in northern Siberia, the results of which are used in this paper.