Evidence for cross-script abstract identities in learners of Japanese kana

Teresa Schubert*, Roderick Gawthrop, Sachiko Kinoshita

*Corresponding author for this work

Research output: Contribution to journalArticle

1 Citation (Scopus)


The presence of abstract letter identity representations in the Roman alphabet has been well documented. These representations are invariant to letter case (upper vs. lower) and visual appearance. For example, “a” and “A” are represented by the same abstract identity. Recent research has begun to consider whether the processing of non-Roman orthographies also involves abstract orthographic representations. In the present study, we sought evidence for abstract identities in Japanese kana, which consist of two scripts, hiragana and katakana. Abstract identities would be invariant to the script used as well as to the degree of visual similarity. We adapted the cross-case masked-priming letter match task used in previous research on Roman letters, by presenting cross-script kana pairs and testing adult beginning -to- intermediate Japanese second-language (L2) learners (first-language English readers). We found robust cross-script priming effects, which were equal in magnitude for visually similar (e.g., り/リ) and dissimilar (e.g., あ/ア) kana pairs. This pattern was found despite participants’ imperfect explicit knowledge of the kana names, particularly for katakana. We also replicated prior findings from Roman abstract letter identities in the same participants. Ours is the first study reporting abstract kana identity priming (in adult L2 learners). Furthermore, these representations were acquired relatively early in our adult L2 learners.

Original languageEnglish
Pages (from-to)1010–1021
Number of pages12
JournalMemory and Cognition
Issue number6
Early online date7 May 2018
Publication statusPublished - Aug 2018


  • abstract letter identities
  • Japanese kana
  • masked-priming same–different match task
  • letter naming; L2 reading acquisition

Fingerprint Dive into the research topics of 'Evidence for cross-script abstract identities in learners of Japanese kana'. Together they form a unique fingerprint.

  • Cite this