Dissertation: Pseudocoordination, serial verb constructions and multi-verb predicates: The relationship between form and structure
Pseudocoordination is exemplified by phrases like go and get or try and do in English. In short, pseudocoordination refers to the use of coordinator ‘and’ where the construction seems to behave unlike typical coordination. This often relates to the historical development of subordination out of coordination, although frequently the resulting constructions appear to be intermediate between coordination and subordination, with properties of both but not fully compatible with either analysis. For this reason syntactic analysis of pseudocoordination is challenging and important.
The phenomenon has received relatively little attention cross-linguistically, yet careful investigation reveals relevant examples in many language families throughout the world; it may not be unusual or unexpected as one might initially assume. Further, when it has been discussed many names have been used (pseudocoordination, subcoordination, fake coordination, serialization, verb-verb agreement, and so forth) with little cross-linguistic comparison.
Pseudocoordination has received the most attention in the Scandinavian Germanic languages, especially Swedish and Norwegian, where it is a salient grammatical feature (cf. Wiklund 2007; Lødrup 2002; and references therein). It has also been discussed some for English (Ross 1967; Carden & Pesetsky 1977; de Vos 2005; Ross, in press; and others). Beyond these languages, descriptions are sparse but can be found for some languages, and for other languages descriptive grammars reveal similar constructions. One exception to the lack of broadly cross-linguistic work is Coseriu (1966), who identifies a large number of languages, primarily in the Indo-European family, with a construction of the form take and, which he originally proposed as an areal feature, shared through borrowing.
My ongoing research has revealed similar constructions in many unrelated families that cannot be due to borrowing yet show remarkable similarities (as well as variation). See Ross (2016) for a cross-linguistic survey. Pseudocoordination is found in many languages including throughout Indo-European, as well as in Finno-Ugric, Semitic, Khoisan, Atlantic languages in the Niger-Congo family, Oceanic and Formosan languages in Austronesian, and elsewhere; at the same time, other languages such as Japanese, Korean and Turkish display an apparently inverse pattern of pseudosubordination where originally subordinate forms act as coordinate structures thereby displaying many of the same properties as pseudocoordination (cf. Yuasa & Sadock 2002). More broadly there are also syntactic similarities in other constructions such as serialization and switch-reference. This typology will be examined in detail in my dissertation. For discussion of related typological features, see Typology.
I have also investigated the diachronic and formal syntactic properties of pseudocoordination. In a way, pseudocoordination may be best defined diachronically as a transitional state between coordination and subordination; it is clear that an originally coordinated structure can undergo reanalysis and grammaticalize as subordinate (cf. Ross, 2013). The resulting constructions often still display some properties of coordination, such as the requirement of parallel morphology on both verbs, making formal syntactic analysis challenging, especially in certain cases such as English try and where this diachronic residue adds complexity to the grammatical system (Ross, 2014).
As a consequence for grammatical theory, there is no one-to-one (isomorphic) correspondence between form and meaning. I believe that the surface form is due to linearization and represents a realization of an underlying grammatical structure of a very different form; for example, there is often no reason, beyond form, to separate pseudocoordination from serial verb constructions, complex predicates with converbs, and other similar constructions – these are language-specific strategies for linearizing the same underlying structures. Therefore, syntactic analysis should be divided into two separate subfields: the mapping of form to underlying structure (Syntax in the sense of the study of word order and sentence forms), and the properties of those structures at the level of Semantics. This is the focus of my dissertation.