| dc.description.abstract | Previous research conducted primarily in English demonstrated that polysemous words (multiple related senses) are processed faster than unambiguous words, while homonymous words (multiple unrelated meanings) are processed slower(Rodd, Gaskell, & Marslen-Wilson, 2002). Research in Serbian replicated those effects, although in separate studies (for polysemes: Filipović Đurđević, 2007; for homonyms:Filipović Đurđević, 2015). The aim of this study wasto replicate those previous findings in Serbian, in the same design as in studies in English, i.e. to test whether the effects remain if the two ambiguity types are presented in the same experimental list. The first study tested the effect of ambiguity type (homonymous –H, polysemous –P, and unambiguous –U) on reaction times and no effect was found. Considering that the main difference was that all three groups were presented in the same experimental list, we decided to investigate whether experimental context was the factor that modulated the effect. In order to test this hypothesis, we employed a different design and explicitly manipulated the presentation of different ambiguity types. Instead of randomised stimulus presentation, we presented blocks ofjust one group of words at the time. The design was modified to control for the order blocks consisting only of homonyms (H-block) and polysemes (P-block), and whether the block consisting of unambiguous words (U-block) was before the other two or afterthem. The second study showed that there is an interaction between the order of H and P blocks order and ambiguity type. Namely, in HP order, the usual homonym disadvantage was present, while there were no processing time differences between U and P blocks,whereas in the PH order, we observed only the polysemy advantage. Position of the U-block was not a significant factor. These results were surprising, so a third study was conducted. Its goal was to replicate this interaction and add another two experimental situations, with U-block between the other two blocks. The interaction from the second study was only partially replicated. Results revealed only a polysemy advantage in PH order, with no influence of U-block position. Overall, all three studies go in line with the conclusions drawn in literature
32review by Eddington & Tokowicz (2015). This paper states that the ambiguity effects are inconsistent, depending on factors other than ambiguity type. Our results also show that there is at least one additional criterion for ambiguity effects to appear. Other research (Armstrong & Plaut, 2016)suggested that another factor influencing the appearance of semantic ambiguity effects is semantic dynamics. Theseand many other findings show that number and relatedness of senses/meaning isnot sufficient enough information to precisely describe mechanisms of ambiguity processing and sense/meaning representation | sr |
