Language Change versus Grammar Change: What diachronic data reveal about the distinction between core grammar and periphery

The paper focusses on two questions. A) How can we explain that language change often proceeds in a slow, gradual fashion, extending over long periods of time in which the old and the new pattern coexist side by side? This fact conflicts with the idea that grammar change via parameter resettings should give rise to abrupt changes. Alternatively, this paper proposes that the gradualness results from a change in language use which eventually may lead up to a change in grammar. More precisely, it is proposed that the grammar provides a limited amount of optionality in the form of stylistic or peripheral rules that can be exploited by speakers for their communicative purposes. These rules may affect word order and prosodic phrasing to derive information-structurally marked forms, which, over time, may loose their stylistic force and become reanalysed as (obligatory) rules of the core grammar. B) What is the nature of word order change? In a framework subscribing to the UBH, changes in word order cannot be relegated to changes in the head complement parameter. Based on the above conception of language change, this paper proposes a novel account of the well known OV-VO change in the history of English, in which word orders resulting from stylistic Light Predicate Raising are reanalyzed as the result of obligatory VP-intraposition. Furthermore, it is claimed that a similar change (from OV+VO to OV) took place in the history in German, in which a stylistic rule that moved focussed DPs into a preverbal position led to a change in the syntax/phonology interface, affecting the possible unmarked word orders in the language