Affix Productivity of Mem-, Memper- and Memper-kan: A Corpus Linguistics Approach

(Produktiviti Imbuhan Mem-, Memper- dan Memper-kan: Pendekatan Linguistik Korpus)

Abstract

This study measures and compares the morphological productivity of the prefixes mem- and memper- and the circumfix memper-kan in Malay. The data used in this study are comprised of two corpora, that is Sports News Online corpus that contains nearly 2.5 million words and an UKM-DBP Corpus which totals five million words. This study also examined whether the size of the corpus has any effect on the morphology productivity of the affixes studied. Wordsmith Tools 4.0 was used to generate and sort the affixes mem-, memper- and memper-kan together with other affixes and root words. Using Baayen's productivity measure, the productivity of the affixes was measured. The results show that the productivity value of the affix memper- have a higher productivity index of P=0.009 and P=0.030 compared to the affix mem- which has a lower productivity index of P=0.002 and P=0.001 for both corpora respectively. Adding the per- and -kan morpheme from mem- affix to memper- and memper-kan affix does not fully affected productivity, instead is influenced by hapaxes. The affix which has the highest morpheme, that is memper-kan, only ranked second with a value of P=0.004 and P=0.008 for both corpora concerned. However, the size of the corpus affected the productivity value. The productivity of morphological value memper- in the UKM-DBP corpus (P=0.030) has a relatively large range of difference compared to the Sports News Online corpus (P=0.009), that is P = 0.021.

Keywords: productivity measure, prefix, circumfix, Malay morphology, sport corpus data, general corpus data, linguistics, computer