Theoretical Implications

Abstract

In this final chapter, I discuss the theoretical implications of the conclusions reached in the previous chapters. First, now that the facts about pronoun placement have been exposed in full detail, I turn my attention to the question of their status as clitics, a term that has been used by other researchers to describe them. This question is considered within the broader context of the necessity of the tripartite classification of grammatical elements into words, clitics and affixes, and it is determined that the weak object pronouns of Later Medieval Greek are best characterized by the term ‘atypical affix’. Determining the status of the pronouns has direct implications on the hypothesis of ‘unidirectionality’, which is a seminal component of the theory of ‘grammaticalization’, and, by extension, on whether we treat this theory as a separate mechanism of linguistic change or as the aggregate of other processes. This issue is discussed in the second section of the chapter. Another hypothesis that is directly impacted by the results of this study is Kroch’s (1989) claim that in order for disparate parts of a grammar to undergo the same process of change, they must share abstract, not surface characteristics. This claim is antithetical to the analysis of weak object pronoun placement that was given in Chapter 7. The two opposing views are discussed in the third section. The final section of the chapter discusses what is perhaps the most surprising result of this study, namely, that there is no generalization, or even a set of generalizations, that can account for the behavior of weak object pronouns in LMG. Although such a claim may go against the traditional view of what the objectives of linguistic analysis should be, there have been several instances in which researchers have had to propose grammars with less-than-perfect generalizations. These case-studies are discussed in the final section and their conclusions are used to illuminate the issue at hand.