Language Change and (Ir)regularization
The Synchronic and Diachronic analyses of (Ir)regularization
by Basima Al-Hussein
Date of Examination:2018-04-10
Date of issue:2018-04-24
Advisor:Prof. Dr. Hedde Zeijlstra
Referee:Prof. Dr. Hedde Zeijlstra
Referee:Dr. Panizza Daniele
Files in this item
Name:Language Change and (Ir)regularization- Basi...pdf
Size:2.92Mb
Format:PDF
Abstract
English
For the last two decades, three main approaches have dealt with the nature of regular versus irregular aspects of language processing in human mind. According to connectionism, all inflected forms are processed in the associative memory (Rumelhart and McClelland 1986; Plunkett and Marchman 1993; Elman 1999). So, both regulars and irregulars are predicted to display frequency effects as a result of storage. However, rule-based approaches predict that all inflected forms are generated by rules and hence they are predicted to exhibit no frequency effects as an outcome of the computation. In both approaches, regularization processes (e.g. using binged instead of brought in the past tense) and irregularization processes (e.g. brang instead of brought) are predicted to be at the similar rates. This bidirectional prediction of verbal changes by approaches of rules and storage is in conflict with the prediction by the dual mechanism approach that defends the necessity of two separate mechanisms for language processing: storage and for irregular processing and rules for regular processing. Irregular verbs can be produced correctly if they are memorized and retrieved successfully before the rule-governed route creates forms of regularization. Nevertheless, the irregularization rate is predicted to be rare compared to the regularization rate. Hence, from the dual mechanism perspective, verbal changes mostly occur unidirectionally, towards regularization only. To date, many studies of language processing cannot offer fully results that undoubtedly approve the predictions of either approach. The current study will attempt to make a contribution to this debate by investigating recent linguistic developments and movements in English verbal system in Contemporary English. I run a corpus study based on data from the multilingual environment of the internet where language change is expected to be faster than in any monolingual environment. In this study, I aim to explore whether verbal developments and changes are towards regularization only favoring the dual mechanism approach or towards both regularization and irregularization favoring single mechanism approaches. The results of the selected data in the current study suggest that on the synchronic level there is a trend towards regularization, while irregularization processes occur rarely. These findings are compatible with the dual mechanism approach, whereas they speak against the hypotheses suggested by the single mechanism approaches. The results of the diachronic analyses of regularization show that the regularization rate is slightly increasing in the time spans (old span: 0.68% versus new span: 0.85%). However, the verbal changes in the direction of regularization are not statistically different in the two spans. Similarly, the results of the diachronic analysis of irregularization indicate that the verbal changes towards irregularization are very infrequent and have the diachronic tendency to be constant over time. This means that the results of the diachronic analyses of (ir)regularization are incompatible with the tenets of single and dual mechanism approaches. From the dual mechanism perspective, verbal changes are predicted to be nonconstant and unidirectional towards regularization only. From single mechanism perspectives, verbal changes are predicted to be bidirectional towards both regularization and irregularization.
Keywords: regularization, irregularization, language change, language acquisition, salience of change