Abstract
The present study examined how the network science measure known as closeness centrality (which measures the average distance between a node and all other nodes in the network) influences lexical processing. In the mental lexicon, a word such as CAN has high closeness centrality, because it is close to many other words in the lexicon. Whereas, a word such as CURE has low closeness centrality because it is far from other words in the lexicon. In an auditory lexical decision task (Experiment 1) participants responded more quickly to words with high closeness centrality. In Experiment 2 an auditory lexical decision task was again used, but with a wider range of stimulus characteristics. Although, there was no main effect of closeness centrality in Experiment 2, an interaction between closeness centrality and frequency of occurrence was observed on reaction times. The results are explained in terms of partial activation gradually strengthening over time word-forms that are centrally located in the phonological network.
Highlights
Complex networks are increasingly being used to better understand various aspects of human cognition (e.g., Steyvers and Tenenbaum, 2005; Hills et al, 2009)
The results from Experiment 1 showed that words with high closeness centrality were responded to more quickly and accurately than words with low closeness centrality, providing evidence that closeness centrality influences spoken word recognition
We suggest that the processing advantage observed for words with high closeness centrality may stem from the advantageous position they occupy in the lexical network, allowing for partial activation to accrue benefits over time
Summary
Complex networks are increasingly being used to better understand various aspects of human cognition (e.g., Steyvers and Tenenbaum, 2005; Hills et al, 2009). In the present study nodes represent phonological word-forms in that part of memory known as the mental lexicon, and links between nodes indicate that the words are phonologically similar to each other (as in the network created in Vitevitch, 2008). One measure of the structure of the network is called the clustering coefficient, or C, which measures how many nodes connected to a target node are connected to each other. C is a measure of the extent to which phonological neighbors of a word are phonological neighbors of each other. The word BADGE has the neighbors BAG, BAD, and BAT, which are neighbors of each other. The word LOG has a low C value because its neighbors tend to not be neighbors of each other
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
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
