The Processing and Mental Representation of Polysemes
Andreas Brocher, Jean-Pierre Koenig, and Gail Mauner1
1
State University of New York at Buffalo
contact: [email protected]
While Homonyms have two semantically unrelated senses (e.g. bank—financial institution; side
of river) polysemes have two semantically related senses (e.g. cold—illness; temperature).
Research on homonyms has focused on lexical access and has revealed that homonym retrieval is
affected by both the relative frequency of their meanings and preceding context (Duffy et al.
1988; Tabossi & Zardon 1993). In the absence of context, the more frequent, dominant
interpretation is retrieved, while biasing context toward the less frequent meaning leads to
competition among meanings thereby delaying lexical access. In contrast, research on polysemes
has focused primarily on whether both senses are stored separately (Klein & Murphy 2001) or in
a single lexical entry (Beretta et al. 2005; Pylkkänen et al. 2006).
We examined the lexical retrieval of polysemes. We hypothesized (1) that in the absence of
biasing context, only shared features of a polyseme’s meanings will initially be activated and (2)
that sense retrieval should be faster when switching between related, as opposed to unrelated
senses, because it only involves switching a subset of semantic features (the unshared features).
Experimental trials consisted of a context word, an ambiguous homonym or polyseme prime
word, and a target word. The dominant and subordinate readings of each prime word were
instantiated and counterbalanced across context and target words. Target words preceded by a
non-word served as baseline controls. Participants made a lexical decision to context, prime and
target words.
context word
ambiguous prime
target word
homonymy
dominant target
subordinate target
FARM
PEN
TEST
TEST
PEN
FARM
polysemy
dominant target
subordinate target
COLOR
SHADE
SUMMER
SUMMER
SHADE
COLOR
Two important results emerged. First, in the absence of context, dominant targets were accessed
faster than baseline targets, but only for homonyms. Thus, dominance effects are apparently
restricted to homonyms. Second, switching from subordinate to dominant senses facilitated
target retrieval equally strongly for homonyms and polysemes. However, switching from
dominant to subordinate senses only boosted target retrieval for polysemes.
The first result suggests that the semantic relatedness of polysemes prevents the processor from
fully committing to one sense when no context is available. The second result suggests that the
availability of shared semantic features eases access of the less frequent sense when the
competing sense has already been retrieved. Results from the context condition further indicate
that an inconsistent context word can facilitate access of a competing target word.
We interpret our results within a featural representation of meaning (McRae et al. 1997) and a
model of lexical retrieval in which both senses of an ambiguous word benefit from cascading
activation between the word’s orthographic and lexical representation (Hino & Lupker 1996).
This explains why inconsistent contexts can facilitate lexical access of competitors. In this
model, intralexical competition additionally guides meaning/sense selection and is greater for
interpretations that do not share any meaning features than for words that do. Sharing meaning
features helps quickly overcome competition. Finally, dominance is a predictor of both cascading
activation strength and competition strength, as the more frequent reading builds up activation
more quickly than the less frequent reading and more readily suppresses its competitor than vice
versa.
References:
Beretta, A., Fiorentino, R, and Poeppel, D. (2005). The effects of homonymy and polysemy on lexical access: and
MEG study. Cognitive Brain Research, 24, 57-65.
Duffy, S. A., Morris, R. K., and Rayner, K. (1988). Lexical ambiguity and fixation times in reading. Journal of
Memory and Language, 27, 429-446.
Hino, Y. and Lupker, S. J. (1996). Effects of polysemy in lexical decision and naming: An alternative to lexical
access accounts. Journal of Experimental Psychology: Human Perceptions and Performance, 22, 1331-1356.
Klein, D. E. and Murphy, G. L. (2001). The representation of polysemous words. Journal of Memory and Language,
45, 259–282.
McRae, K., De Sa, V. R., and Seidenberg, M. S. (1997). On the nature and scope of featural representation of word
meaning. Journal of Experimental Psychology: General, 126, 99-130.
Pylkkänen, L., Llinas, R., and Murphy, G. L. (2006). The representation of polysemy: MEG evidence. Journal of
Cognitive Neuroscience, 18, 97-109.
Tabossi, P. and Zardon, F. (1993). Processing ambiguous words in context. Journal of Memory and Language, 32,
359-372.