Distributional semantic models (DSMs) are a primary method for distilling semantic information from corpora. However, a key question remains: What types of semantic relations among words do DSMs detect? Prior work typically has addressed this question using limited human data that are restricted to semantic similarity and/or general semantic relatedness. We tested eight DSMs that are popular in current cognitive and psycholinguistic research (positive pointwise mutual information; global vectors; and three variations each of Skip-gram and continuous bag of words (CBOW) (...) using word, context, and mean embeddings) on a theoretically motivated, rich set of semantic relations involving words from multiple syntactic classes and spanning the abstract–concrete continuum (19 sets of ratings). We found that, overall, the DSMs are best at capturing overall semantic similarity and also can capture verb–noun thematic role relations and noun–noun event-based relations that play important roles in sentence comprehension. Interestingly, Skip-gram and CBOW performed the best in terms of capturing similarity, whereas GloVe dominated the thematic role and event-based relations. We discuss the theoretical and practical implications of our results, make recommendations for users of these models, and demonstrate significant differences in model performance on event-based relations. (shrink)

Language is more than a source of information for accessing higher-order conceptual knowledge. Indeed, language may determine how people perceive and interpret visual stimuli. Visual processing in linguistic contexts, for instance, mirrors language processing and happens incrementally, rather than through variously-oriented fixations over a particular scene. The consequences of this atypical visual processing are yet to be determined. Here, we investigated the integration of visual and linguistic input during a reasoning task. Participants listened to sentences containing conjunctions or disjunctions and (...) looked at visual scenes containing two pictures that either matched or mismatched the nouns. Degree of match between nouns and pictures and between their expected and actual spatial positions affected fixations as well as judgments. We conclude that language induces incremental processing of visual scenes, which in turn becomes susceptible to reasoning errors during the language-meaning verification process. (shrink)