Identifying the phonological backbone in the mental lexicon

Area of Science:

• Linguistics
• Network Science
• Psycholinguistics

Background:

• Previous research applied network science to identify important nodes in English phonological word-form networks.
• The concept of a

Purpose of the Study:

• To extract a subnetwork of

Main Methods:

• Applied a network simplification process, known as the backbone, to remove redundant edges.
• Extracted a subnetwork (backbone) from the original phonological word-form network.
• Compared psycholinguistic and network measures of words within and outside the backbone.

Main Results:

• The backbone procedure removed 68.5% of edges, yielding a giant component of 6,211 words.
• Words in the backbone were more frequent, shorter, phonologically similar to more neighbors, and closer to other words.
• Words in the backbone may constitute a

Conclusions:

• The identified backbone words may represent a