Language sciences poster session and lunch

Language scientists - come and share your research with colleagues from other departments over lunch (kindly provided by the Pro-Vice-Chancellor for Research, Professor Lynn Gladden).

There is an open call for posters from within the University of Cambridge and affiliated research bodies. A maximum of 6 will be accepted on a first-come, first-served basis. The deadline for submissions is 18 February.

Researchers from arts and humanities disciplines where posters are not the usual medium of presentation are encouraged to participate by using alternatives such as A3 photocopies - please get in touch to discuss.

Poster abstracts

Multidimensional scaling analysis of Mandarin Chinese and Thai shape classifiers

Woramon Prawatmuang, Faculty of Asian and Middle Eastern Studies

A classifier is a type of functional word present in some Asian languages, such as Mandarin Chinese and Thai. Among various types of classifiers, there are shape classifiers, those that select nouns with certain perceptual attributes, such as being round, long or flat.

Multidimensional scaling (MDS) is a technique used by some researchers to illustrate the relationship among shape classifiers in Chinese. It is argued that the closer the classifiers are in the MDS model, the closer they are in terms of semantic attributes, and the more likely they can be used with the same nouns.

In this study, it is shown that the technique is also suitable for the Thai data, as the model shows clustering of classifiers with the same attributes, similar to those in the Chinese model. The MDS models in Chinese and Thai both indicate that the relationship between shape classifiers in these languages is highly based on whether the classifiers share a perceptual attribute. That is, if two classifiers both have a "round", "long" or "flat" attribute, their relationship is likely to be tighter than classifiers with different attributes. Moreover, the models also reveal a fine distinction of classifiers in the "long" group, where those with an additional "soft or bendable" attribute have a loose relationship with other "long" classifiers. This phenomenon is found in both Chinese and Thai data.