The effects of homonymy and polysemy on lexical access: an MEG study

Sustained meaning activation for polysemous but not homonymous words: Evidence from EEG

“…The different meanings of a homonym share a common lexical form by historical accident and accordingly lexicographers list different meanings of homonyms as separate entries in dictionaries. Importantly, psychological data is consistent with the linguistic approach to homonymy and it is widely accepted that the different meanings have distinct lexical representations in the brain (Beretta, Fiorentino, & Poeppel, 2005;Pylkkänen, Llinás, & Murphy, 2006). By contrast, the nature of polysemous representations is far more controversial and less well understood.…”

“…On the other hand, the observed lack of activation of homonymous meanings in the current results is well explained by a processing mechanism in which the different meanings compete for activation (see also, Beretta et al, 2005;Rodd et al, 2002). We suggest that in the absence of a biasing context to support one meaning or the other, and the absence of a task requiring retention of either meaning, this competition results in decay of both meanings.…”

“…Across a number of studies it has been shown that processing words with distinct meanings delays word recognition (homonymy disadvantage) as measured by lexical decision times (Beretta et al, 2005;Rodd et al, 2002), eye fixation times (Frazier & Rayner, 1990) and the latency of the M350 brain response (Pylkkänen et al, 2006). These studies mostly interpret their results in terms of competition between the alternative meanings of homonyms 2 .…”

“…These studies mostly interpret their results in terms of competition between the alternative meanings of homonyms 2 . On the other hand, words with multiple senses have been shown to result in faster processing (sense advantage) across a range of measures (Beretta et al, 2005;Rodd et al, 2002), which is typically accounted for by semantically richer representations relative to unambiguous words. Our findings contribute to this body of research by using neural measures to show that whilst multiple related senses of polysemous words (both metonymy and metaphors) act collaboratively to strengthen a representation, alternative unrelated 2 Beretta et al, (2005) suggest that the homonymy disadvantage could also be explained by frequency rather than competition because homonyms were matched with control words on frequency of the surface form, thus the frequency of individual meanings was smaller than that of the controls.…”

“…In contrast, the distinct meanings of homonyms compete for activation or selection, which delays processing. Further evidence for a distinction between homonyms and polysemes comes from lexical decision tasks in which reaction times were faster to polysemous words than to unambiguous words, but slower to homonymous words (Beretta et al, 2005;Klepousniotou & Baum, 2007;Rodd, Gaskell, & Marslen-Wilson, 2002). The processing disadvantage observed for homonymous words is also well explained by an account of word recognition in which the different meanings of the homonym compete for activation or selection, whereas the facilitation observed for processing polysemous senses reflects either co-activation of multiple senses which perhaps form a rich semantic representation, or activation of an underspecified representation where selection of a specific sense is not required.…”