Basima Mahmood (Göttingen)

Language change and (ir)regularization

The nature of the mental representation and processing of morphologically complex words has constituted one of the major points of controversy in linguistics and psycholinguistic research over the past two decades. In this respect, two different types of approaches can be distinguished in this respect: single and dual mechanism approaches of morphological processing. Single mechanism approaches posit no fundamental distinction between regular and irregular inflections, and contend that both are built via a single mechanism - either a rule system or an associative system. Followers of rule-based models assume that both regular and irregular forms are generated by rule (Chomsky and Halle's 1968; Halle and Mohanan 1985). By contrast, followers of associative models assert that all inflected words are stored within a single associative system (Rumelhart and McClelland 1986; Bybee 1996). Both single mechanism approaches (rules or storage) do not predict a clear correlation between (ir)regularity and word frequency, since there is no firm distinction between regular and irregular forms. The dual-mechanism model defends the necessity of two separate mechanisms for linguistic processing, an associative memory and a rule system, which account for the processing of irregular and regular word forms, respectively. From the dual mechanism perspective, there is a positive correlation between word frequency and irregularity: the more frequent words are, the more likely they are irregular. Nevertheless, regular forms are rule-produced and are consequently frequency-insensitive. Diachronically, it has been attested that the number of irregular verbs in English has steadily declined over time. Irregular verbs with low frequency are regularized more often than the ones with high frequency (Fries 1940; Lieberman et al. 2007; Michel, et al 2011). Yet, this one-sided view that looks at morphological change in the direction of regularization has been objected by some linguists. Changes in the other direction, the direction of irregularization, have been observed as well (Peters 2004; Nübling 2000 and Fertig 2013).

The current study is a contribution to this debate by exploring the possibility of a relationship between word frequency and (ir)regularity in the English verbal system by running a corpus study. The aim of this study is to investigate whether only the process of regularization applies (favoring the dual mechanism approach) or the process of irregularization can also be attested (favoring single mechanism approaches). The results of the data analysis suggest a relationship between word frequency and irregularity: IVs are high-frequency, whereas RVs are not. In addition, a trend towards regularization is attested while irregularization takes place rarely. These results are preliminarily accounted for by the dual mechanism approach, whereas they speak against the hypothesis suggested by the single mechanism approaches.