Search

We investigate the onomasiological question of which words speakers actually use and produce when trying to convey an intended meaning. This is not limited to selecting the best-fitting available existing word, but also includes word formation, the coinage of novel words. In the first two experiments, we introduce the taboo game paradigm in which participants were instructed to produce a single-word substitution for different words so that others can later identify them. Using distributional semantic models with the capability to produce quantitative representations for existing and novel word responses, we find that (a) responses tend to be semantically close to the targets and (b) existing words were represented closer than novel words, but (c) even novel compounds were often closer than the targets' free associates. In a final third experiment, we find that other participants are more likely to guess the correct original word (a) for responses closer to the original targets, and (b) for novel compound responses as compared to existing word responses. This shows that the production of both existing and novel words can be accurately captured in a unified computational framework of the semantic mechanisms driving word choice. • We investigate which words exactly speakers use to express an intended meaning. • In the taboo game paradigm, speakers substitute a target with a single other word. • Participants are allowed or required to also produce novel words. • Existing and novel word responses are well captured in a single distributional framework. • Responses not well captured by this framework are also less informative for other speakers. [ABSTRACT FROM AUTHOR]

This paper examines changes in the accent class affiliation of c. 1900 words from Middle Korean into the modern South Kyengsang dialect. The data present the profile of a canonical analogical change: words are attracted to larger lexical classes and words of lower token frequency are more likely to change their affiliation. Several properties of the syllable onset and coda as well as syllable weight are shown to bias a word to particular accent classes. A novel word experiment suggests that speakers have tacit knowledge of some of these phonological biases but not others. The paper considers whether these biases can explain the default accent assigned to English loanwords and whether they can be modeled with weighted constraints in a Maxent grammar. [ABSTRACT FROM AUTHOR]

Abstract: The present study investigated the role of spelling in phonological variant processing. Participants learned the auditory forms of potential reduced variants of novel French words (e.g., /plur/) and their associations with pictures of novel objects over 4days. After the fourth day of training, the spelling of each novel word was presented once. Half the words were spelled with an orthographic representation of the schwa (i.e., “e”), half were not. In the subsequent naming tasks, participants produced more schwa variants for novel words whose spelling contained an “e”. In addition, reduced variants with an “e” in spelling and an onset cluster attested word-internally in non-schwa words were produced with longer latencies than the same items whose spelling did not contain an “e”. Finally, in a recognition task where participants had to decide whether a given spoken item was part of the experimental stimuli trained the previous days, participants were more likely to say yes to a schwa variant when the spelling for the given word corresponded to this variant. These results show that a single exposure to spelling following extensive phonological learning can change the way speakers and listeners store and process words with phonological variants both in production and recognition tasks. [Copyright &y& Elsevier]