Docs / Studies / Studies.GundelHedbergZacharski1993

Cognitive Status and the Form of Referring Expressions in Discourse

Gundel, J. K., Hedberg, N. & Zacharski, R. (1993). Language, 69(2), 274–307

A referring form signals a cognitive status of its referent, and the six statuses of the Givenness Hierarchy are nested, so a form fits any referent at or above its status. Across Chinese, English, Japanese, Russian and Spanish, corresponding forms require the same status, except the demonstrative determiners. In discourse the forms meet their necessary conditions and occur at every status that meets them, and the maxim of Quantity governs which: an in-focus referent mostly takes the most restrictive form, a familiar full noun phrase the least.

Implementation notes

  • The forms Table 1 aligns across languages are Form.Kinds. A printed form of Tables 2–6 has the kinds it may be: one, except a Spanish personal pronoun, which was not coded for stress, and which requires the weaker of the two statuses.
  • A demonstrative's deixis is the participant sets of its anchors, read off its fragment entry (Chinese nèi is nà, and the masculine singular stands for each paradigm). Spanish ese, usable at any distance from the speaker, covers every participant set without the speaker, and Japanese so- the listener's alone, its further reading slightly removed from both participants unrecorded. Harbour takes Modern Japanese ko–so–a to have become distance-based, while Terenghi lists it as person-oriented and treats Spanish ese as also denoting intermediate distance from the speaker.
  • Japanese kare, listed under activated, is a stressed pronoun here, since the paper takes Japanese to have no unstressed personal pronoun.

Errata

  • The indefinite article is listed among the forms whose required status differs across languages, but Table 1 and the paper's own discussion of the article give it the same status, type identifiable, in Chinese, English and Spanish; only the demonstrative determiners differ.
  • The share of in-focus referents coded by the most restrictive form is said to range from 71% (Russian) to 87% (English), but Japanese, with 87 of 125, codes a smaller share than Russian.
  • The share of familiar full noun phrases that are bare in Japanese is printed as 59%, but Table 4 gives 63 of 108, under 58.5%.

References

Table 1

L70–90

The corpus

L164–198
L200–207