When They Change the Way They Speak: Contact-Induced Word Order Shifts in Semitic

Classical Semitic languages, with the exception of Akkadian, all display VSO (Verb- Subject-Object) either as the basic main constituent word order or as one of the most attested main constituent word order (the other being SVO). Other word order features, such as that of Modifier and Head, are mostly typologically harmonious with VO (i.e., Head-Modifier). The greatest deviation from this likely proto word order is found in Akkadian, Ethiopian Semitic and Central Asian Arabic, of which the basic main constituent word order has shifted to SOV, while these Semitic languages were in contact with non-Semitic languages.This dissertation provides analyses of the word order patterns in Ethiopian Semitic and Central Asian Arabic, with text corpora built from scratch for Classical Ethiopic and Bukharan Arabic, representing the two groups. In addition, a mini-corpus was also built for Old Amharic (pre-18th century), considering the fact that Amharic is the only one among the modern Ethiopian Semitic languages with relatively old textual records reflecting its archaic features. The corpus investigations demonstrated the Semitic fronting and resumption mechanisms contributed to the formation of the surface O-V sequence in Classical Ethiopic. Apart from some of the word order features in Old Amharic that are inconsistent with the typical OV typology, some of its O-V sequence appear to echo the fronting mechanism similar to that in Classical Ethiopic. A similar picture emerged from the corpus investigation of Bukharan Arabic: the Semitic fronting and resumption mechanisms are also reflected in its O-V pattern. Furthermore, as it is in the situation of Ethiopian Semitic, it is likely there had been similar analytic constructions that served as the basis for the fronting and resumption. The Semitic case studies are further compared with the situation in the Sinitic family where the stable disharmonic SVO developed into frequent STV in some of its descendant languages via the inherited topic prominent feature and the topic fronting further served as the basis for the development of OV orders in contact settings. Taking the East Asian situation as hints, it is therefore argued that in the contact situations between Semitic VSO and non-Semitic SOV as well as the contact situations between Sinitic SVO and non-Sinitic SOV, the existing Semitic and Sinitic mechanisms created O-V sequences closely resembling the SOV word order in the contact languages. Such a structural congruence provides the bilingual speakers with a convenient middle ground and also a potentially efficient convergence strategy for necessary communication adaptation. In the other Semitic languages where historical word order shifts took place, the existing mechanisms appear also to have been exploited to various degrees, with the similar fronting and resumption mechanisms forming easily identifiable congruence. It is thus emphasized that the Semitic native speakers must have also played an important role, as the speakers in such contact settings selected the congruent structures from the contact feature pool. This study also calls for caution in invoking word order stability based on consistency as a driving force for language change and highlights the potential complexities involved in reconstructing precise scenarios of historical language contact