ABSTRACT
This lexical decision with eye-tracking study investigated how Japanese trimorphemic compounds (e.g. 体温計 “clinical thermometer”) are recognised. The questions answered were, in the course of decomposing and composing Japanese trimorphemic compounds, (1) whether recognition processes are tuned for a specific branching direction, (2) whether the morphological processing proceeds in a bottom-up combinatorial manner, and (3) whether the three constituents of trimorphemic compounds are equally important and processed serially. Mixed-effects regression analyses of response times and fixation durations revealed that a left-branching advantage appears in a late time frame and that, although there was early processing of the whole compound from the first fixation, a character frequency effect was also observed. Furthermore, the first and the third, but not the second, constituent frequencies contributed to compound recognition. This bathtub-like effect was further supported by corpus-based evidence: the conditional probability for the second constituent is incomparably high.
Acknowledgements
The response time and fixation duration data in a long format, accompanying participant and item properties, are downloadable from the first author’s website (http://kojimiwa.com/publication.html). The descriptions of the variables and data trimming procedures are found in this paper. The frequency data obtained from Balanced Corpus of Contemporary Written Japanese (BCCWJ) are published with permission of the National Institute for Japanese Language and Linguistics. We thank Denis Drieghe and an anonymous reviewer for their comments on the earlier version of this paper.
Disclosure statement
No potential conflict of interest was reported by the authors.
ORCiD
Koji Miwaa http://orcid.org/0000-0002-0890-9265