The present study drew on two experimental tasks to collect data, one comprehension task and one production task. The two tasks were implemented in the form of a written questionnaire.
The comprehension task was a grammaticality judgment task. The design of the task is summarized in Table 3-2. As Table 3-2 shows, the grammaticality judgment task was devoted to the three target constructions in English, with focus on two major properties of each construction. One property was concerned with the verb types of the target constructions, and the other property bore on predication by resultatives, particle movement, and word order of datives/double objects.3 Each property was
3 This study examined three constructions at the same time. Since investigation of more than two properties of each construction would bring about a much heavier workload on the subjects, which might reduce the possibility of eliciting intuitional responses, two properties of each construction was selected and examined. One property concerns verb types, for all the three telicity-related constructions are incompatible with any stative verbs—one property in common. The other property pertains to prediction by resultatives, particle movement, and word order of datives/double objects, which were
Table 3-2. The grammaticality judgment task
4Construction Property 9/± No. of Qs Q item
Resultative
characterized by two contrasting subproperties. Since the constructions under examination pertained to telicity, the present study centered on linguistic expressions of completion of telic events. Completion of a telic event is perfective in terms of viewpoint aspect, which is expressed in simple past tense in English.5 Thus, all the test sentences in this task were in past tense. With respect to the number of questions, 2 sentences were chosen for each subproperty of the three constructions. Thus, there
selected for the following considerations. As English resultatives are not easily differentiated from depictives. According to Tenny (1994), the former can be predicated only of the direct object, while the latter can modify the subject or the object. Thus, the restrictive predication is the distinguishing feature of English resultatives. For verb-particles, verb-particles exhibit great resemblance to prepositions with respect to their forms. The characteristic property of the former is particle movement, while the latter are not subject to such a movement (Bolinger 1971). As regards datives/double objects, the constructions are characterized by different but fixed word order of the constructions (Baker 1979).
4 In Table 3-2, the tick ‘9’ signifies that the examples are grammatical in English, and the cross ‘±’
indicates that the examples are ungrammatical.
5 According to Smith’s two-component approach to aspect (1991, 1997), aspect is decomposed into viewpoint and situation aspect. The two levels of aspect will interact with each other in unmarked or marked fashions. Telic situations bear an unmarked relation to perfective viewpoint aspect.
were 24 trial sentences for the constructions (3 constructions × 2 properties × 2 subproperties × 2 sentences＝24). The grammaticality judgment task, plus 6 fillers, contained 30 sentences in total. These sentences were randomized in order in the questionnaire. For a comprehensive shape of the questionnaire, please refer to Appendix A.
As discussed in Section 2.1, English resultative constructions can convert an atelic event into a telic counterpart, if the main verb is felicitous with a resultative predicate, as in The girl wrung the towel dry. Hence, resultative constructions can apply to eligibly dynamic verbs, such as wring and hammer, but not to any stative verbs, like own and love, for it is impossible to turn stative verbs into verbs that do denote a telic event in English. Based on the attribute, trial sentences with dynamic verbs like (1a) were considered grammatical, while examples containing stative verbs were ungrammatical, as in (1b).
(1) a. The girl washed the soap out of her eyes.
b. *Linda liked him touched.
With respect to the property of predication, resultative expressions could be predicated only of an object but are not allowed to function as a matrix subject complement. Thus, trial sentences like (2a) were regarded as well-formed, in which the resultative expression was predicated of the object. In contrast, sentences like (2b) were ungrammatical due to the predication of the subject by the resultative complement.
(2) a. The man shot Irene dead.
b. *George wrote his homework angry.
In the case of English verb-particle constructions, inapplicability to stative verbs and particle movement are characteristic of the constructions. Like resultatives, English verb-particle constructions also have the effect of telicity conversion, converting an atelic event into a delimited event, if the main verb is eligible. By the same token, verb-particle constructions are not congruent with stative verbs, for stative verbs are not felicitous with telic interpretations. Accordingly, test sentences with a felicitously dynamic verb like (3a) were taken to be grammatical, but combination of stative verbs and verb-particle constructions resulted in ill-formed sentences, as in (3b).
(3) a. The man knocked down the boy.
b. *Mary loved up John.
Besides (in-)compatibility with verbs, the grammaticality judgment task also took particle movement into consideration. English verb particles can appear on either side of a direct object NP. Based on the property, test sentences like (4a) and (4b) were regarded as grammatical since aside and out were a verb particle in English and were subject to the particle movement.
(4) a. Teresa put aside the magazine.
b. Janet blew the candles out.
In English dative/double object constructions, all the verbs are ditransitive and dynamic. Combination of any stative verbs and the constructions led to ungrammatical sentences, as in (5b).
(5) a. The man told her daughter a story.
b. *Jean knew Richard a ball.
As for their word order, it is fixed in both versions.6 When a direct object precedes an indirect object, the case marker to is obligatory for the latter. However, when a direct object follows an indirect one, the case marker is absorbed. Thus, trial sentences as in (6a) were grammatical, whereas retention of the case marker to in sentences like (6b) gave rise to an ill-formed sentence.
(6) a. The man gave a toy to his son.
b. *Charles sent to Miss Li a letter.
In addition to examining the subjects’ comprehension of the three constructions, this study attempted to elicit production data from the subjects, which could be examined from a quantitative as well as a qualitative perspective and served as a source of information concerning the relationships of the three constructions. A free translation task was designed to obtain production data of the target constructions, i.e.
translation from Chinese sentences into complete and corresponding English counterparts.
Table 3-3 shows a rough picture of the translation task. Similar to the grammaticality judgment task, the free translation task was targeted at the three telicity-related constructions in English and two main properties of each construction.
The properties investigated in the translation task were duplication of those probed in the grammaticality judgment task: that is, applicability to dynamic verbs and
6 Dative constructions can alternate with double object constructions in various contexts. However, dative alternation is licensed by different constraints in English and Chinese (cf. Baker 1979, Pinker 1989, Liu 2006). In the present study, dative shift was not a property at issue. To factor out the effect of different constraints on dative alternation, all the ditransitive verbs selected in the grammaticality judgment task and the free translation task were dativizable verbs, i.e. ditransitive verbs that can apply both to dative constructions and to double object constructions in English.
Table 3-3. The free translation task
Construction Property No. of Qs Q item
Resultative Dynamic verb 2 Q2, Q9
Predication of the object 2 Q6, Q14
Verb-particle Dynamic verb 2 Q1, Q12
Particle movement 2 Q4, Q15
Dative/
double object
Dynamic verb 2 Q3, Q8
Word order 2 Q7, Q13
Filler 3 Q5, Q10, Q11
Total 15
predication of the object in resultative constructions, compatibility with dynamic verbs and particle movement in verb-particle constructions, and congruity with dynamic verbs and characteristic word order in dative/double object constructions.
Nonetheless, the ways the major properties were examined in the two tasks were slightly different in the following two respects. First, the grammaticality judgment task contained grammatical and ungrammatical English sentences, whereas the translation task focused on grammatical English sentences only. The second point consisted in the property of particle movement. As a verb particle is allowed to appear on either side of an object NP, it is likely that skewed occurrences of verb particles in their distribution might be elicited, which does not indicate the failure of the acquisition of the property. To avoid the possibility, sentences involving a pronoun were chosen in the translation task to elicit data of particle movement, since a verb particle conventionally appears after a pronoun, as the trial sentence in (7) illustrates.
(7) The woman found her key on the ground and picked it up.
In the translation task, as Table 3-3 shows, two sentences were designed for each property of the three constructions. Thus, there were 12 trial sentences for the target
constructions (3 constructions × 2 properties × 2 sentences＝12) plus 3 fillers in total. As completion of a telic event is expressed in simple past tense in English, in the translation task, the perfective aspect marker -le in Chinese was used to elicit perfective past-tensed English translation (cf. Li and Thompson 1982).7 An example of the translation task is shown below.
(8) Ta ti-kai-le na-shan men.
he kick-open-ASP that-CL door He kicked the door open.
In the questionnaire, Chinese sentences, i.e. the Romanization in the first line in (8), were presented in the form of Chinese characters, and the glosses in the second line were not provided. The English sentence underscored in the third line was an expected answer to the Chinese test sentence. For a complete version of the questionnaire, please refer to Appendix B.