Dative Constructions

Table 3-27 shows the four groups’ interpretations of dative sentences where a universal QNP preceded an existential QNP:

Table 3-27: Subjects’ Interpretations of Dative Sentences with a Universal QNP Preceding an Existential QNP (in percentages)

Distributive Collective Both

Reading Group

Item E1 E2 NE NC E1 E2 NE NC E1 E2 NE NC

N1 6.7 13.3 0 6.7 93.3 83.3 88.9 93.3 0 3.3 11.1 0 N2 20.0 13.3 0 13.3 66.7 80.0 83.3 80.0 13.3 6.7 16.7 6.7 Average 13.35 13.3 0 10.0 80.0 81.65 86.1 86.65 6.65 5.0 13.9 3.35

Different from their interpretations of simple actives, simple passives, and double object constructions, Groups 1 and 2 were in favor of the following QNP to have a wide scope interpretation when they interpreted N1 and N2, shown in Table 3-27. The preference for the collective reading in response to a sentence with a universal QNP preceding an existential QNP violates the linearity principle, which claims that the preceding QNP has a wide scope interpretation. The percentage of the collective reading is comparatively higher than the other readings for Groups 1 and 2.

According to the Ch-square test, Groups 1 and 2 had similar interpretations of N1 and N2 and their interpretations were also similar to the English control group (χ² <

5.991, p > 0.05).

Compared with the other three groups, the Chinese majors had similar interpretations of Chinese QNPs to their interpretations of English QNPs (χ² < 5.991, p > 0.05).

Our subjects’ responses to the dative sentences with an existential QNP preceding a universal QNP are shown in Table 3-28:

Table 3-28: Subjects’ Interpretations of Dative Sentences with an Existential QNP Preceding a Universal QNP (in percentages)

Distributive Collective Both

Reading Group

Item E1 E2 NE NC E1 E2 NE NC E1 E2 NE NC

N1 53.3 60.0 77.8 66.7 30.0 36.7 0 26.7 16.7 3.3 22.2 6.7 N2 46.7 70.0 72.2 46.7 43.3 20.0 0 26.7 10.0 10.0 27.8 26.7 Average 50.0 65.0 75.0 56.7 36.65 28.35 0 26.7 13.35 6.65 25.0 16.7

Similar to their interpretations of dative constructions with a universal QNP preceding an existential QNP, Groups 1 and 2 preferred the following QNP to have a wide scope interpretation when they interpreted the dative constructions where an existential QNP preceded a universal QNP. As shown in Table 3-28, Group 2 preferred the distributive reading more than Group 1 in response to N1 and N2;

nevertheless, the difference was not significant. Compared with our English control group, the experimental groups had higher percentages of the collective reading.

According to the Chi-square test and posterior comparisons, Groups 1 and 2 responded significantly differently from the control group in the collective reading (E1 vs. NE in N1: χ² = 6.661, ψ of the collective reading = 0.3 ± 0.205; E1 vs. NE in N2: χ² = 11.240, ψ of the collective reading = 0.433 ± 0.223; E2 vs. NE in N1:

χ² = 10.987, ψ of the collective reading = 0.367 ± 0.216).

Similar to Groups 1 and 2, the Chinese control group liked the collective reading more than the English control group in interpreting N1 (NC vs. NE in N1: χ² = 6.246, ψ of the collective reading = 0.267 ± 0.260).

In general, in response to dative constructions with QNPs, the experimental groups and the Chinese controls preferred the following QNP to have a wide scope interpretation, which was different from their interpretations of QNPs in simple actives, simple passives, and double object constructions. When the experimental

groups interpreted dative constructions with an existential QNP preceding a universal QNP, they liked the collective reading more than native speakers of English, which might be relevant to the influence of the linearity principle.

3.4.2.5 Subject Control Constructions

Table 3-29 shows the subjects’ responses to subject control sentences with a universal QNP preceding an existential QNP:

Table 3-29: Subjects’ Interpretations of Subject Control Sentences with a Universal QNP Preceding an Existential QNP (in percentages)

Distributive Collective Both

Reading Group

Item E1 E2 NE NC E1 E2 NE NC E1 E2 NE NC

N1 60.0 63.3 27.8 66.7 20.0 20.0 16.7 20.0 20.0 16.7 55.6 13.3 N2 76.7 83.3 50.0 80.0 13.3 3.3 5.6 20.0 10.0 13.3 44.4 0 Average 68.35 73.3 38.9 73.35 16.65 11.65 11.15 20.0 15.0 15.0 50.0 6.65

In interpreting N1 and N2, Groups 1 and 2 liked the distributive reading. Group 2 preferred the distributive reading more. The Chi-square test indicated that Groups 1 and 2 had similar interpretations of N1 and N2. However, compared with our English control group, Groups 1 and 2 had significantly higher percentages of the distributive reading and lower percentages of the ambiguous reading in response to N1 and N2.

According to the Chi-square test and posterior comparisons, Groups 1 and 2 significantly differed from the English native controls in the distributive and ambiguous readings when they interpreted QNPs in N1 (E1 vs. NE: χ² = 6.771, ψ of the distributive reading = 0.322 ± 0.297, ψ of the ambiguous reading = 0.356 ± 0.285;

E2 vs. NE: χ² = 8.356, ψ of the distributive reading = 0.355 ± 0.337, ψ of the ambiguous reading = 0.389 ± 0.243). In response to N2, Group 1 differed from the English native controls in the ambiguous reading and Group 2 preferred the distributive reading more than the English control group (E1 vs. NE: χ² = 7.678, ψ

of the ambiguous reading = 0.344 ± 0.317; E2 vs. NE: χ² = 6.254, ψ of the distributive reading = 0.333 ± 0.332). The above analyses tell us that Group 1’s and Group 2’s interpretations were similar and that they exhibited less preference for the ambiguous reading than the English control group.

As for the Chinese controls’ interpretations of QNPs were similar to those of the experimental groups. However, they preferred the distributive reading more and the ambiguous reading less than the English controls (NC vs. NE in N1: χ² = 6.783, ψ of the distributive reading = 0.389 ± 0.384, ψ of the ambiguous reading = 0.423 ± 0.409; NC vs. NE in N2: χ² = 9.232, ψ of the ambiguous reading = 0.444 ± 0.327).

Presented in Table 3-30 are the subjects’ responses to subject control sentences with an existential QNP preceding a universal QNP:

Table 3-30: Subjects’ Interpretations of Subject Control Sentences with an Existential QNP Preceding a Universal QNP (in percentages)

Distributive Collective Both

Reading Group

Item E1 E2 NE NC E1 E2 NE NC E1 E2 NE NC

N1 0 3.3 0 0 96.7 96.7 83.3 93.3 3.3 0 16.7 6.7 N2 10.0 0 0 13.3 86.7 96.7 61.1 80.0 3.3 3.3 38.9 6.7 Average 5.0 1.65 0 6.65 91.7 96.7 72.2 86.65 3.3 1.65 27.8 6.7

Table 3-30 shows that in response to N1 the two experimental groups had the same interpretations and that in response to N2 Group 2 preferred the collective reading more. The Chi-square test indicated that they had similar interpretations of N1 and N2. The Chi-square also showed that the two experimental groups had similar interpretations to the English control group in interpreting N1, but in response to N2, Group 1 was different from the English native controls in the ambiguous reading (E1 vs. NE: χ² = 11.286, ψ of the ambiguous reading = 0.356 ± 0.293) and Group 2 was more in favor of the collective and ambiguous readings than the English native

controls (E2 vs. NE: χ² = 10.240, ψ of the collective reading = 0.356 ± 0.293, ψ of the ambiguous reading = 0.356 ± 0.293).

The Chinese native controls did not exhibit significant differences from Groups 1 and 2. Nevertheless, they were less in favor of the ambiguous reading than the English controls in response to N2 (NC vs. NE: χ² = 6.323, ψ of the ambiguous reading = 0.322 ± 0.321).

Generally speaking, similar to the interpretations of simple actives, simple passives, and double object constructions, in response to subject control constructions the experimental groups’ and the Chinese majors’ interpretations supported the linearity principle. The two experimental groups and the Chinese majors liked the preceding QNP to have wide scope interpretations more and liked the ambiguous reading less than native speakers of English.

3.4.2.6 Object Control Constructions

Shown in Table 3-31 are the four groups’ interpretations of object control constructions with a universal QNP preceding an existential QNP:

Table 3-31: Subjects’ Interpretations of Object Control Sentences with a Universal QNP Preceding an Existential QNP (in percentages)

Distributive Collective Both

Reading Group

Item E1 E2 NE NC E1 E2 NE NC E1 E2 NE NC

N1 53.3 60.0 11.1 46.7 33.3 26.7 38.9 40.0 13.3 13.3 50.0 13.3 N2 46.7 50.0 16.7 40.0 36.7 33.3 38.9 46.7 16.7 16.7 44.4 13.3 Average 50.0 55.5 13.9 43.35 35.0 30.0 38.9 43.35 15.0 15.0 47.2 13.3

In interpreting N1 and N2, Groups 1 and 2 preferred the distributive reading.

Interestingly, there were also about thirty percent of subjects of each group choosing the collective reading when they interpreted N1 and N2, which implied that these

sentences had ambiguous property for the experimental groups. The Chi-square test indicated that Groups 1 and 2 had similar interpretations of N1 and N2. The Chi-square test also showed that the two experimental groups did not differ in the readings of N2 from the English native controls, but differed significantly from the control group in interpreting N1 (E1 vs. NE: χ² = 11.031; E2 vs. NE: χ² = 12.576).

According to posterior comparisons, in response to N1 Groups 1 and 2 differed from the English controls in the distributive and ambiguous readings (E1 vs. NE: ψ of the distributive reading = 0.422 ± 0.288, ψ of the ambiguous reading = 0.367 ± 0.326; E2 vs. NE: ψ of the distributive reading = 0.489 ± 0.284, ψ of the ambiguous reading = 0.367 ± 0.326). The experimental groups preferred the preceding QNP to have a wide scope interpretation more and the ambiguous reading less than the English control group.

The Chinese majors’ interpretations of QNPs were similar to those of Groups 1 and 2. Compared with the English controls, they liked the distributive reading more in interpreting N1 (NC vs. NE: χ² = 7.095, ψ of the distributive reading = 0.356 ± 0.350).

Presented in Table 3-32 are the subjects’ interpretations of object control constructions with an existential QNP preceding a universal QNP:

Table 3-32: Subjects’ Interpretations of Object Control Sentences with an Existential QNP Preceding a Universal QNP (in percentages)

Distributive Collective Both

Reading Group

Item E1 E2 NE NC E1 E2 NE NC E1 E2 NE NC

N1 20.0 3.3 0 0 66.7 66.7 55.6 86.7 13.3 30.0 44.4 13.3 N2 10.0 6.7 0 0 90.0 90.0 61.1 100 0 3.3 38.9 0 Average 15.0 5.0 0 0 78.35 78.35 58.35 93.35 6.65 16.65 41.65 6.65

Table 3-32 shows that in interpreting N1 and N2, Groups 1 and 2 preferred the collective reading, as predicted by the linearity principle. In response to N1, Group 1 preferred the distributive reading more and liked the ambiguous reading less than Group 2. According to the Chi-square test, the percentage differences between Group 1 and Group 2 was not significant in response to N1 and N2. The Chi-square test also indicated that the experimental groups did not differ from the English native controls in interpreting N1 but differed significantly in the ambiguous reading in interpreting N2 (E1 vs. NE: χ² = 14.653, ψ of the ambiguous reading = 0.389± 0.282; E2 vs. NE:

χ² = 10.919, ψ of the ambiguous reading = 0.356 ± 0.293).

Similar to the experimental groups’ interpretations of QNPs in the object control constructions, the Chinese majors differed from the English controls in interpreting N2 (NC vs. NE: χ² = 7.404, ψ of the collective reading = 0.389 ± 0.281, ψ of the ambiguous reading = 0.389 ± 0.321).

To sum up, similar to their interpretations of simple actives, simple passives, double object constructions, and subject control constructions, Group 1’s, Group 2’s, and the Chinese control group’s preference for the preceding QNP to have a wide scope interpretation in object control constructions supported the linearity principle.