Word order optionality in the postfinite domain in North and Continental West Germanic has attracted much attention in generative syntax in the past three decades, in particular the phenomena known as ‘Object Shift’ and ‘Scrambling’. However, the precise factors conditioning this word order optionality remain unclear, and there is to date no account which can explain the full cross-Germanic variation on display.
This project will shed new light on this area via novel historical data from Icelandic (North Germanic) and Low German (Continental West Germanic), which are well placed to fill gaps in the literature on the topic and are accessible for this type of study via parsed corpora (Icelandic Parsed Historical Corpus; Corpus of Historical Low German).
In order to achieve this, a new annotation scheme which encodes the relevant information-structural and semantic properties will be developed as a generally applicable enhancement for historical Penn-style treebanks and will be employed for the two corpora in use.
Additionally, the project will develop a novel way to model gradience in word order flexibility within the architecture of Lexical-Functional Grammar in order to facilitate comparison between the Low German and Icelandic data and with further Germanic varieties.
Broader questions which the project plans to address:
- What insights can be gained from comparing historical North and West Germanic varieties?
- How can we better capture uncertainty in corpus annotation and how can we harness this for theoretical insights?
- How can microvariation in word order be modelled in terms of gradient degrees of configurationality?
- How does information structure interact with subordination and coordination and how can this acccount for word order variation across complex sentences?