2018 Friday Poster 6578

Friday, November 2, 2018 | Poster Session I, Metcalf Small | 3pm

Conjunction and numeral as cues to generate alternatives to satisfy scalar presupposition of “only”: Evidence from Japanese learners of English
A. Sugawara

Background: Sentences with a focus-sensitive particle “only” can generate surface ambiguous sentences such as in (1), with possible interpretations in (1a-b). The ambiguity is yielded since “only” could locate its associate F at multiple positions. How is the locating process governed? One idea concerns the nature of the scalar presupposition of “only” (Beaver&Clark ’08, a.o.) – that is, the prejacent is a relatively low-ranked alternative among Alt(S). For (1a), suppose a set of alternatives

{water, food, feast}. Since the alternative {water} is low-ranked, (1a) satisfies the scalar presupposition and is a valid sentence. In contrast, in the sentence (2), the alternative {feast} is not low-ranked, so the only possible interpretation is (2b). This suggests that “easily scalable” constituents could attract F. Another idea on how to locate F concerns the Alt(S) generation algorithm. Building on the analysis proposed by Fox&Katzir (2011), it is expected that a conjunction could also attract F (details omitted here due to the space limitation; but see (3)).

Predictions: If the two ideas above are on the right track, the truth condition of ambiguous sentences will be manipulated by (i) placing an “easily scalable” constituent such as numeral, and (ii) placing a conjunction, on one of the object positions.

Experiment: The participants recruited were 57 Japanese college students, who have studied English for at least 6 years at school. The counterpart of “only” in Japanese is a bound morpheme (“-sika … nai” and “-dake”), and thus when they process English ambiguous sentences (with “only” and F far apart), they cannot use their transferred knowledge from L1.

The participants read stories with stimulus sentences to judge if they match the contexts (True/False). The target sentences vary in three factors as in (4a-c). From the responses, we can deduce which object the participants associated “only” with. E.g., if one answers False to (4a), (s)he associated “only” with the 1st object, while if True to (4a), (s)he associated it with the 2nd object. The contexts vary in two ways as in (4-5). Theese 6 conditions in total was Latin-squared.

The results are summarized in Table (6). As expected, when a numeral is placed in one of the object positions, the interpretation is much affected so that the interpretations show that F is attracted by the numeral.

Table (8) shows the detailed results of the same data, with responses divided by the context types. Notice that Context A ends with the conjunction in the 1st object, while Context B ends with the conjunction in the 2nd object. If the position of the conjunction affects/primes the ease with the Alt(S) generation, the interpretation with F in the 1st object is facilitated under Context A, while the interpretation with F in the 2nd object is facilitated under Context B (indicated with squares in (8)). The results indeed suggest that this is the case.

In sum, the predictions (i-ii) are borne out. This study shows that L2ers also process the ambiguous sentences with “only” with a complex strategy that is shared with native speakers.

References

Beaver, David and Brady Clark. 2008. Sense and sensitivity: How focus determines meaning. Oxford: Blackwell.

Fox, Danny and Roni Katzir. 2011. On the characterization of alternatives. Natural Language Semantics 19 (1): 87-107.