The Latent Structure of Dictionaries

Abstract

How many words—and which ones—are sufficient to define all other words? When dictionaries are analyzed as directed graphs with links from defining words to defined words, they reveal a latent structure. Recursively removing all words that are reachable by definition but that do not define any further words reduces the dictionary to a Kernel of about 10% of its size. This is still not the smallest number of words that can define all the rest. About 75% of the Kernel turns out to be its Core, a “Strongly Connected Subset” of words with a definitional path to and from any pair of its words and no word's definition depending on a word outside the set. But the Core cannot define all the rest of the dictionary. The 25% of the Kernel surrounding the Core consists of small strongly connected subsets of words: the Satellites. The size of the smallest set of words that can define all the rest—the graph's “minimum feedback vertex set” or MinSet—is about 1% of the dictionary, about 15% of the Kernel, and part‐Core/part‐Satellite. But every dictionary has a huge number of MinSets. The Core words are learned earlier, more frequent, and less concrete than the Satellites, which are in turn learned earlier, more frequent, but more concrete than the rest of the Dictionary. In principle, only one MinSet's words would need to be grounded through the sensorimotor capacity to recognize and categorize their referents. In a dual‐code sensorimotor/symbolic model of the mental lexicon, the symbolic code could do all the rest through recombinatory definition.

Visual cortex recruitment during language processing in blind individuals is explained by Hebbian learning

“…We mimic associative learning between word forms used to speak about objects and their referent objects present in the environment as well as between action words and the performance of their semantically-related actions, as it is well-documented in the literature on language learning 37,64 . Although other forms of semantic learning (e.g., from texts or by definition) also play a role in meaning acquisition, we focus on the direct semantic grounding of words in object and action knowledge, because it is both prominent in early language learning and a precondition for other forms of semantic learning 65,66 . In the sighted model simulations, object-and action-related word acquisition was grounded in sensorimotor information presented to the primary areas of the model: object-related word learning was driven by perisylvian activity in *A1 and *M1 i and concordant visual (*V1) activity patterns; similarly, action-related word learning was driven by semantic activity in the lateral motor area (*M1 L ) along with perisylvian activity (Fig.…”

Connecting concepts in the brain by mapping cortical representations of semantic relations

“…In addition, word2vec learns the semantic space from large corpora in a data-driven manner 21 . This is different from defining the semantic space based on keywords that are hand selected 22 , frequently used 1 , minimally grounded 41 , or neurobiologically relevant 23,42 . Although those word models are seemingly more intuitive, they are arguably subjective and may not be able to describe the complete semantic space.…”

Discovering Psychological Principles by Mining Naturally Occurring Data Sets

“…As a fifth example, Vincent‐Lamarre et al. () analyze information about connections between words and the minimal networks of words sufficient to define all other words, as revealed by large corpora analyses of dictionaries.…”

“…() and Vincent‐Lamarre et al. () also use NODS to distinguish between alternative computational accounts of cognition.…”