How words are used in naming different categories of objects

Abstract

Aims: The research is a part of a cross-linguistic study on evolution of semantic systems (Majid, Jordan, & Dunn, 2011). The main aim is to investigate how the meaning of words vary over space and change over time, and how semantic categories diverge across languages. We present the data compiled in Serbian language, and explore the meanings of words used in describing different semantic categories: kinds of objects (containers), attribute of objects (color), parts of objects (body parts), and relations between objects (spatial relations). The focus is on the variety of words used for different kinds of meaning. Method: Four series of pictures were presented to 20 adult participants (half male) who were asked to name each stimulus (elicitation task). Colors: A series of color chips (developed by Majid &Levinson, 2007). Spatial relations: A series of pictures containing an orange figure object against a black ground object (Bowerman & Pederson, 1992). The task was to describe the relation between the figure and ground. Body Parts: Pictures of a human body or face with a red dot marking the body part (Jordan, Dunn, & Majid, 2009). Containers: A series of photographs of household containers (Ameel, Storms, Malt, & Sloman, 2005). The testing (four experimental tasks) lasted approximately one hour per participant. The answers were audio recorded and transcribed. Coding and analysis: Participants’ full responses in all four tasks were coded into the main response - a core term in a uniform citation form (e.g. in the Color Naming task dark green was coded as green, in the Spatial Relations task The cup is on the table was coded as on, in the Body Parts task upper shoulder was coded as shoulder, in the Containers tea cup was coded as cup). The analysis aimed at counting the frequency of different core terms for each picture and finding the stimuli that provoke uniform responses, and those that trigger a large variety of them. More detailed inspection aimed at searching for objective properties of stimuli in relation to the variety of answers. Results: The stimuli for each of four experimental tasks were ordered by the variety of answers. Among the Containers, the largest variety in responses was triggered by the objects with aberrant form, size and/or material. In the Body Parts, it was triggered by dislocation of the red spot from the central and/or privileged position of a particular body part. In Spatial Relations, the larger variety was produced for functionally unrelated objects (e.g. a rubber hose on a tree-stump). The Color naming task provoked the largest number of uniform responses, and in contrast to previous tasks, it was difficult to define what objective property of color (hue, value, chroma), or a combination of properties, could be specified as a trigger for the observed variety of responses. The results were discussed from the pragmatic point of view and the zones of possible cross-linguistic variety were hypothesized.