Gibson, Edward. 2000. The dependency locality theory: A distance-based theory of linguistic complexity. Image, language, brain, 95–126.
Gieseler, Jens. 1996. Vom Nutzen und richtigen Gebrauch der frühen Zeitungen. Zur sogenannten Pressedebatte des 17. Jahrhunderts. In Fritz, Gerd & Erich Strassner (eds.), Die Sprache der ersten deutschen Wochenzeitungen im 17. Jahrhundert. Berlin, New York: Max Niemeyer Verlag, 1996.
Hawkins, John A. 2004. Efficiency and complexity in grammars. Oxford: Oxford University Press.
Holler, Anke. 2005. Weiterführende Relativsätze. Empirische und theoretische Aspekte. Berlin: Akademie Verlag.
Konieczny, Lars. 2000. Locality and parsing complexity. Journal of psycholinguistic research, 29, 627–645.
Lötscher, Andreas. 1972. Some problems concerning Standard German relative clauses. In The Chicago which hunt. Papers from the relative clause festival. A paravolume to papers from the eigth regional meeting. Chicago Linguistic Society, 47–58.
Shannon, Thomas. 1992. Toward an adequate characterization of relative clause extraposition in modern German. In Rauch, Irmengard, Gerald F. Carr & Robert L. Kyes (eds.), On Germanic linguistics. Issues and methods. Berlin, New York: de Gruyter: 253–281.
Strunk, Jan. 2014. A statistical model of competing motivations affecting relative clause extraposition in German. In MacWhinney, Brian, Andrej Malchukov & Edith Moravcsik (eds.). Competing motivations in grammar and usage. Oxford University Press 2014, 1:88–106.
Uszkoreit, Hans, Thorsten Brants, Denys Duchier, Brigitte Krenn, Lars Konieczny, Stephan Oepen & Skut Wojciech. 1998. Studien zur performanzorientierten Linguistik. Aspekte der Relativsatzextraposition im Deutschen. Kognitionswissenschaft 7:65.

This study argues that the varying position of relative clauses (RC) in the history of German can be explained by processing constraints. Although there has been much research regarding which factors influence the RC position, most studies in this field are based on Present-day data. In the present diachronic corpus study, I investigate how the preferred word order in German newspapers changes over time and which factors play the most important role by using variable importance measures.
In general, German relative clauses can occur intraposed or extraposed. In the first case, the nominal head and RC are adjacently placed in the midfield (1). In contrast, they appear in different syntactic positions in the extraposition pattern: the head noun in the midfield and the RC in the postfield (2).

(1) Anna soll [[den Kaktus]NP, [den Fred ihr geschenkt hat]RC], regelmäßig gießen.
(2) Anna soll [den Kaktus]NP regelmäßig gießen, [den Fred ihr geschenkt hat]RC.
‘Anna should regularly water the cactus that Fred gave her.’

The choice for a particular word order is due to various factors. These include e.g. the restrictiveness of the RC (Lötscher, 1972, Holler, 2005) and the definiteness of the antecedent (Shannon, 1992), but also dependency length (DL) measures. This can be explained by processing constraints. Hawkins (2004) for example states that dependencies between a final verb and its preceding arguments can be significantly shortened by extracting the RC from the matrix clause. Gibson (2000) takes a similar approach concerning sentence comprehension. However, presenting competing motivations, extraposition lengthens the dependency between antecedent and RC. The likelihood of extraposition should therefore increase with increasing RC lengths and decrease with longer distances to the antecedent. This has been confirmed in corpus studies (e.g., Uszkoreit et al., 1998) as well as in experimental studies (e.g., Konieczny, 2000).
Based on that, Strunk (2014) tests the influence of different factors synchronically in German newspaper texts. He concludes that the distance between antecedent and RC as well as the RC length play the most important role in the selection of the RC position. Considering the fact that my diachronic data show a steady increase in extraposition patterns, this study can provide new insights into the relevance of DL factors throughout the history of German. Therefore, I investigate the placement of heavy nominal and prepositional phrases followed by an RC modifier in newspaper texts from 1600 until today. Using conditional inference trees and random forests, I look at nine factors: RC length, distance to antecedent, restrictiveness, definiteness of antecedent, antecedent’s part of speech, part of the sentence, embedding degree, time period, and the newspaper’s publication region. My results support Strunk’s study insofar as DL measures are the most important factors for the RC position in contemporary German. However, a less clear picture emerges from the early data, indicating that DL seems to become more important only with time. I argue that this can be explained by the accessibility of newspapers. The first newspapers were accessible only to a minority who could read. For private individuals, the benefits as well as the right to read were disputed (Gieseler, 1996). Only since the expansion of the reading society in the 19th century, processing constraints seem to be considered to make the content more accessible to the reader.