Abstract
According to frame-theory, concepts can be represented as structured frames that contain conceptual attributes (e.g., “color”) and their values (e.g., “red”). A particular color value can be seen as a core conceptual component for (high color-diagnostic; HCD) objects (e.g., bananas) which are strongly associated with a typical color, but less so for (low color-diagnostic; LCD) objects (e.g., bicycles) that exist in many different colors. To investigate whether the availability of a core conceptual component (color) affects lexical access in language production, we conducted two experiments on the naming of visually presented HCD and LCD objects. Experiment 1 showed that, when naming latencies were matched for colored HCD and LCD objects, achromatic HCD objects were named more slowly than achromatic LCD objects. In Experiment 2 we recorded ERPs while participants performed a picture-naming task, in which achromatic target pictures were either preceded by an appropriately colored box (primed condition) or a black and white checkerboard (unprimed condition). We focused on the P2 component, which has been shown to reflect difficulty of lexical access in language production. Results showed that HCD resulted in slower object-naming and a more pronounced P2. Priming also yielded a more positive P2 but did not result in an RT difference. ERP waveforms on the P1, P2 and N300 components showed a priming by color-diagnosticity interaction, the effect of color priming being stronger for HCD objects than for LCD objects. The effect of color-diagnosticity on the P2 component suggests that the slower naming of achromatic HCD objects is (at least in part) due to more difficult lexical retrieval. Hence, the color attribute seems to affect lexical retrieval in HCD words. The interaction between priming and color-diagnosticity indicates that priming with a feature hinders lexical access, especially if the feature is a core feature of the target object.
Highlights
In order to communicate our thoughts to other human beings, it is crucial to be able to refer to concepts shared with the listener
A particular color value can be seen as a core conceptual component for objects which are strongly associated with a typical color, but less so for objects that exist in many different colors
Event-Related Potentials (ERPs) waveforms on the P1, P2 and N300 components showed a priming by color-diagnosticity interaction, the effect of color priming being stronger for high color-diagnostic (HCD) objects than for low color-diagnostic (LCD) objects
Summary
In order to communicate our thoughts to other human beings, it is crucial to be able to refer to concepts shared with the listener. According to decompositional accounts of conceptual representation (e.g., Goldman, 1975; Jackendoff, 1990; Bierwisch and Schreuder, 1992), conceptual features (such as female and parent) are combined to access the concept (such as MOTHER 1). No direct representation of the whole concept (i.e., a single conceptual node for MOTHER) is used to access the lemma. In contrast to decompositional accounts, non-decompositional accounts propose a single abstract representation of the whole concept (e.g., a conceptual node MOTHER), which in itself does not include features of the concept but is connected to conceptual nodes such as FEMALE and PARENT via conceptual relations. In terms of lexical access, non-decompositional accounts assert that each lemma node is connected to and activated by a single conceptual node. A recent attempt to model the cumulative semantic inhibition effect showed no differences when either decompositional or non-decompositional semantics were assumed (Howard et al, 2006)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
