In this dissertation, I investigate word order and information structure in the light of recent developments within the minimalist program. I specifically pursue a principled explanation of word order within the biolinguistic perspective. In that sense, I entertain the thesis that all properties of the faculty of language contribute to a computationally efficient satisfaction of interface conditions. The language examined is Romeyka, the only Asia Minor Greek variety still spoken in the area historically known as Asia Minor (present-day Anatolia, Turkey). The objective of this study is therefore twofold: (a) descriptively, to examine word order variation in Romeyka and (b) theoretically, to investigate whether such word order variation could be a language specific property or, rather, could be accommodated in a minimalist system. Descriptively, I aim (a) to determine the pragmatically unmarked and marked word orders in Romeyka, (b) to examine their typological classification and (c) to investigate their evolution. Theoretically, this dissertation is fundamentally about the role that order plays in the efficient computation of interface conditions, mainly in regard to the syntax and semantics interface. Generative Grammar is the study of linguistic capacity as a component of human cognition. As such, Generative Grammar has made significant progress in identifying some of the computational mechanisms that distinguish man from animals; the basic tenet is that only humans appear to possess a mental (universal) grammar that permits the composition of infinitely many meaningful expressions from a finite stock of discrete units. The basic compositional operation of grammar is said to be Merge. Merge can create a set K of two linguistic objects x and y. Set K can be merged with another object z, or with another set of objects L and so on. Merge imposes a hierarchical structure, i.e. x and y are elements of K, but not vice versa. Merge is assumed not to impose order, i.e. {x, y} = {y, x}. As such, order is structure-dependent, i.e. no syntactic operation can make reference to it. It has also been claimed that hypothetical languages, in which syntactic operations are defined in linear terms, such that Merge creates an ordered pair <x, y>, are outside of the spectrum of variation defined by universal grammar. The question I am asking is whether the order of the constituents of a clause plays a role (a) in the computation from narrow syntax to the semantics interface and (b) in the semantic component. I pursue an approach where the constituents of a clause do play such a role and ask what the implications are for the syntax and semantics interface. The findings of the dissertation show that order plays a role in the semantic component and in the computation from narrow syntax to semantic interface in Romeyka.