Regionale variatie in de postverbale distributie van presentatief er.

Abstract

Building on an extensive written corpus of formal and informal Belgian and Netherlandic Dutch, this study tackles the complex distribution of non-anaphoric er "there" in adjunct-initial presentative sentences such as Op de hoek van de straat is (er) een winkel At the corner of the street (there) is a shop. The Dutch standard grammar ANS maintains that for this distribution "no strict rules can be given. It can be optional, there may be semantic or stylistic differences, and there is a lot of individual, sometimes also regional variation (1997: 473)." In order to test this view, we confronted two language-structural factors - adjunct type and verbal specificity - and two contextual factors - region and register - in a regression analysis of the use of er in the 1905 adjunct-initial presentative sentences in the corpus. This statistical analysis demonstrates that the ANS is inaccurate and far too pessimistic as far as presentative er's postverbal distribution is concerned. The fact that language-structural and contextual factors are put on a par in the quoted passage inadequately reflects the far greater impact of the structural factors on er's distribution. In addition, the predictive success of the global er-model distilled from the data (the Gamma-index is 84.8 %) is strikingly at odds with the "no strict rules"-pessimism of the ANS. The most important discovery in this paper, however, is that er's distribution in the Belgian and Netherlandic materials is accounted for by proportionally and structurally different models. The practical consequence of this finding is that the ANS should devote separate entries to er's distribution in Belgian and Netherlandic Dutch. On a methodological level, this paper argues strongly in favour of spontaneous, non-elicited language data as the empirical basis for the study of syntactic variation.