• Examining the Case for Unidirectionality in Grammaticalization

Grammaticalization is a phenomenon common in many of the world’s languages where linguistic elements (e.g., words) become more grammatical over time. It is often assumed to be a unidirectional process. Such an assumption has far-reaching implications both for linguistic theory and its applications (e.g., historical reconstruction). This essay offers a critical analysis of this widely held principle of strict unidirectionality. Specifically, three common arguments defending unidirectionality of grammaticalization are evaluated. Firstly, using examples of the development of the postverbal 1PL pronoun in Irish and the free connective ga in Japanese, this essay demonstrates that there are valid and cogent examples of degrammaticalization. Secondly, using Norde’s (2012) parameters of degrammaticalization, this essay also shows that degrammaticalization is not simply ‘noise in the system’ but can be systematically classified and explained in a typology. Finally, this essay evaluates the argument that no good motivation exists for degrammaticalization processes (famously posited by Haspelmath, 1999). Contrary to Haspelmath (1999), this essay argues that arriving at the conclusion of strict unidirectionality is premature. Instead, grammaticalization and degrammaticalization may be understood as a clockwork of different processes that are motivated by different pressures. Such an approach concedes the asymmetries in occurrences of grammaticalization and degrammaticalization while at the same time proposing that unidirectionality in grammaticalization should at least be weakened to a statistical universal.