NINJAL特別講義 レナード・タルミー教授 (ニューヨーク州立大学) 「標的設定のシステム : 直示と照応の統合」

窪薗 晴夫 (国立国語研究所 理論・対照研究領域 教授)
文法研究班 「動詞の意味構造」
松本 曜 (国立国語研究所 理論・対照研究領域 教授)
平成31年3月22日 (金) 13:30~15:00
平成31年3月23日 (土) 13:30~15:00
国立国語研究所 講堂 (東京都立川市緑町10-2)


These talks propose that a single cognitive system underlies the two domains of linguistic reference traditionally termed anaphora and deixis. In anaphora, the referent is an element of the current discourse itself, whereas in deixis, the referent is outside the discourse in its spatiotemporal surroundings. This difference between the lexical and the physical has traditionally led to distinct theoretical treatments of such referents. Our proposal, on the contrary, is that language engages a single linguistic/cognitive system -- "targeting" -- to single out a referent whether it is speech-internal or speech-external. This system can be outlined as follows.

As a speaker communicates with a hearer, her attention can come to be on something in the environment -- her "target" -- that she wants to refer to at a certain point in her discourse. This target can be located near or far in either the speech-external (deictic) or the speech-internal (anaphoric) environment. She thus needs the hearer to know what her intended target is and to have his attention on it jointly with her own at the relevant point in her discourse. The problem, though, is how to bring this about. She cannot somehow directly reach into the hearer's cognition, take hold of his attention, and place it on her selected target at the intended moment.

Language solves this problem through targeting. First, at the intended point in her discourse, the speaker places a "trigger" — one out of a specialized set of mostly closed-class lexical forms. English triggers include: this/these, that/those, here, there, yonder, now, then, therefore, thus, so, such, yay, the, personal pronouns, relative pronouns, and tense markers.

Next, on hearing the trigger, the hearer undertakes a particular 3-stage procedure. In the first stage, he seeks all available "cues" to the target. Such cues belong to ten distinct categories, representing ten different sources of information. In the second stage, he combines these cues so as to narrow down to the one intended target. And in the third stage, he maps the concept of the target he has found back onto the original trigger for integration with the sentence's overall reference.

The ten cue categories, divided into five groups of two each, are (lexical:) core and co-form cues; (bodily:) gestural and corporal cues; (collateral:) targetive and hearer-focus cues; (background:) environmental and epistemic cues; and (temporal:) chronal and perichronal cues.


  • 3月22日 (金) 13:30~15:00
    “Gestures as cues to a target”
  • 3月23日 (土) 13:30~15:00
    “The targeting system of language”