Hypocoristics: a derivational problem

This study is an investigatory research on the two major schools of linguistics, formal and functional. The study looks at earlier versions of Generative Theory as the representative of formal linguistics and contrasts it to Skousen’s computational model which is taken as the representative of functional linguistics. The way each of the theories are described and evaluated are by considering how each of them can be used in analysing hypocoristic data. A description of hypocoristics for 165 names collected from Kuwaiti Arabic speakers were the base for the analysis. The data was given a general description at first to show how they can be accounted for in the two theories. The first approach that was used was a rule-based approach used previously with Jordanian Arabic Hypocoristics which use Semitic root and Pattern Morphology. The second rule-based approach was also a rule-based approach the employed phonological processes to account for the derivation. The two were considered part of formal theories of analysis. The functional analysis which uses a computational model that employs phonological features defined over statistically driven frequencies was used to model the data. An evaluation of the model with low success rates lead to the change of the model and present an alternative hybrid model that utilises both rules and analogy. The model was inspired by a rule-based theory which was not fleshed out and analogy was used to flesh it out and place it with a usage-based theory of language. Finally, the thesis ended with an open evaluative stand requiring further research on computational models from a computational perspective rather than a linguistics view.