Hypothesis Testing and Theory Amending: What corpus research can offer Linguistics? - a Chinese case
Yanan Hu, Dirk Geeraerts, Dirk Speelman · Lirias · 2014
With corpus-related methodology established in linguistic studies, we hereby address the most crucial contribution that corpus research makes to Linguistics, namely hypothesis testing and theory amending. Inspired by Speelman and Geeraerts (2009), our study re-examines the (in)direct causation hypothesis formulated first by Verhagen and Kemmer (1997) on Dutch causatives doen and laten, and analyzed by Stukker (2005), from a Chinese perspective. Starting with the assumption put forward by Ni (2012) that in Chinese shǐ expresses direct causation like doen and ràng expresses indirect causation like laten (among the seven analytic causatives shǐ, lìng, ràng, jiào1, jiào2, gěi and yào), we apply multinomial logistic regression analysis and multiple correspondence analysis to a data set of occurrences retrieved from two corpora, the Sheffield Corpus of Chinese