Conceptual and Perceptual Dimensions of Word Meaning Are Recovered Rapidly and in Parallel during Reading

Abstract

A single word (the noun " elephant") encapsulates a complex multidimensional meaning, including both perceptual (" big," " gray," " trumpeting") and conceptual (" mammal," " can be found in India") features. Opposing theories make different predictions as to whether different features (also conceivable as dimensions of the semantic space) are stored in similar neural regions and recovered with similar temporal dynamics during word reading. In this magnetoencephalography study, we tracked the brain activity of healthy human participants while reading single words varying orthogonally across three semantic dimensions: two perceptual ones (i.e., the average implied real-world size and the average strength of association with a prototypical sound) and a conceptual one (i.e., the semantic category). The results indicate that perceptual and conceptual representations are supported by partially segregated neural networks: Although visual and auditory dimensions are encoded in the phase coherence of low-frequency oscillations of occipital and superior temporal regions, respectively, semantic features are encoded in the power of low-frequency oscillations of anterior temporal and inferior parietal areas. However, despite the differences, these representations appear to emerge at the same latency: around 200 msec after stimulus onset. Taken together, these findings suggest that perceptual and conceptual dimensions of the semantic space are recovered automatically, rapidly, and in parallel during word reading.

Enriching Language Models with Visually-grounded Word Vectors and the Lancaster Sensorimotor Norms

“…However, this only works if an agent that has learned the language has the knowledge of horses, black, white, stripes, and vertical concepts-i.e., via direct experience, not just through linguistic exposure or encyclopedic definitions. These claims are further backed up by neuroscience research that showed that neural assemblies encode concrete content words (i.e., words that denote visual objects) and verbs (i.e., words that denote actions) are learned and represented in different brain regions (Pulvermüller, 1999;Borghesani et al, 2019). Rogers et al (2020) provides a recent primer and overview of research that has attempted to uncover strengths and weaknesses of BERT and related language models (so-called BERTology).…”

Multiple spatial codes for navigating 2-D semantic spaces

“…We further also showed, here, that these dimensions are represented also separately, in those brain regions that represent the perceptual features when the objects are physically presented, using a distance code adapted to the lower 1-dimensional space that define each perceptual feature. Interestingly, Borghesani et al 2018…”

Building up Open-Ended Empirical Modules in Translating: A Case Study of “Logical Meaning Extensions”

“…Word meaning is a frequently-discussed concept (e.g., Austin, 1962;Clark & Gerrig, 1983;Chandler, 1995;Galántai, 2002;Fawcett, 2007;Vahid Dastjerdi, 2011;Vitello & Rodd, 2015;Armstrong & Plaut, 2016;Jurko, 2017;Borghesani, 2019;Ke, 2019;Butler, 2020;Chen, 2020;Rodd, 2020) and perception of lexical meaning and choices of words with justified motivations is always a major and important task in the study of translating practice of moving the meaning of Source Language (SL) to Target Language (TL). But just as Baker (1992) noted, "it is rarely possible to analyze a word, … into distinct components of meaning as language is much more complex to allow that", the complexity of making justifiable choices of words in bilingual translation involves different factors influencing language usages (cf., Duffy et al, 1988;Grace, 1998;Bolger et al, 2008;Yeibo, 2011;Chang, 2018 etc.).…”