In addition to the trend of N1-preference or N2-preference of our study, the second research question attempts to delineate the effects of animacy on the subjects’
performances on each subtype. Table 4-4 shows the subtypes:
Table 4-6 Subtype Design for Testing the Animacy Effect
When the N1 was (1) [+animate] The N1-preference reading The N2-preference reading (2) [-animate] The N1-preference reading The N2-preference reading When the N2 was (3) [+animate] The N1-preference reading The N2-preference reading (4) [-animate] The N1-preference reading The N2-preference reading
4.2.1 Overall Findings
We now examine the results type by type to see the overall pattern from each subtype:
(1) N1 was [+animate]:
Table 4-5 shows the frequency counts of all subjects’ performance when the modifier position was controlled as [+animate]:
52
Table 4-7 N1-preference Reading Frequency Counts When N1 was [+animate]
T1-1: Animate-Animate T1-2: Animate-inanimate N1-preference Frequency count % Frequency count %
Age 4 53 47.3% 58 51.8%
Table 4-8 N2-preference Reading Frequency Counts When N1 was [+animate]
T1-1: Animate-Animate T1-2: Animate-inanimate N2-preference Frequency count % Frequency count %
Age 4 59 52.7% 54 48.2%
Type 1-1 and Type 1-2 both belong to Type 1, which features the [+animate] modifier.
Therefore, we listed their frequency counts together here. These counts could not supply an obvious growing pattern, so we will transfer these counts into percentage
53
format, as in Figures 4-6 and 4-7:
Figure 4-6 N1-preference Reading for Type 1-1 and Type 1-2
Figure 4-7 N2-preference Reading for Type 1-1 and Type 1-2
When it was a N1-preference reading, Type 1-2 with the inanimate head was preferred by all age groups (51.8%, 63.3%, 65.4%, 66.9%, 67.4%, and 72.4% for Ages 4~8 and
54
the control group.) Among them, only the control group reached 70%, and Age 4 got the lowest percentage. When it was a N2-preference reading, Type 1-1 with the animate head was preferred by all age groups (52.7%, 48.4%, 47.1%, 44.1%, 41%, and 37.5% for Ages 4~8 and the control group.) Almost all the percentage, however, were below 50%. Among them, the native controls got the lowest percentage.
(2) N1 was [-animate]:
Table 4-6 shows the frequency counts of all subjects’ performance when the modifier position was controlled as [-animate].
Table 4-9 N1-preference Reading When the N1 was [-animate]
T2-1: Inanimate-Animate T2-2: Inanimate-Inanimate N1-preference Frequency count % Frequency count %
Age 4 55 49.1% 54 48.2%
Age 5 68 53.1% 66 51.6%
Age 6 79 58.1% 75 55.1%
Age 7 85 62.5% 88 64.7%
Age 8 92 63.9% 99 68.7%
Control 102 67.1% 108 71.1%
Total 481 490
55
Table 4-10 N2-preference Reading When the N1 was [-animate]
T2-1: Inanimate-Animate T2-2: Inanimate-Inanimate N2-preference Frequency count % Frequency count %
Age 4 57 50.9% 58 51.8%
Age 5 60 46.9% 62 48.4%
Age 6 57 41.9% 61 44.9%
Age 7 51 37.5% 48 35.3%
Age 8 52 36.1% 45 31.3%
Control 50 32.9% 44 28.9%
Total 327 318
Type 2-1 and Type 2-2 both feature the [-animate] modifier. Therefore, their frequency counts were listed together here. It was not easy to show an obvious growing pattern we would like to see, so we would transfer these counts into percentage format, as in Figures 4-8 and 4-9:
56
Figure 4-8 N1-preference Reading for Type 2-1 and Type 2-2
Figure 4-9 N2-preference Reading for Type 2-1 and Type 2-2
When it was a N1-preference reading, Type 2-2 with the inanimate head was preferred only by Ages 7, 8, and the control group, that is, the older age groups. When it was a N2-preference reading, Type 2-1 with the animate head was preferred only by Age 7,
57
8, and the control, that is, the older age groups. Therefore, here we could see a major difference between the younger three age groups and the older three age groups, and the turning point seemed to be Age 7.
(3) N2 was [+animate]:
Table 4-7 shows the frequency counts of all subjects’ performances when the head was [+animate].
Table 4-11 N1-preference Reading When N2 was [+animate]
T1-1: Animate-Animate T2-1: Inanimate-Animate N1-preference Frequency count % Frequency count %
Age 4 53 47.3% 55 49.1%
Age 5 66 51.6% 68 53.1%
Age 6 72 52.9% 79 58.1%
Age 7 76 55.9% 85 62.5%
Age 8 85 59% 92 63.9%
Control 95 62.5% 102 67.1%
447 481
58
Table 4-12 N2-preference Reading When N2 was [+animate]
T1-1: Animate-Animate T2-1: Inanimate-Animate N2-preference Frequency count % Frequency count %
Age 4 59 52.7% 57 50.9%
Age 5 62 48.4% 60 46.9%
Age 6 64 47.1% 57 41.9%
Age 7 60 44.1% 51 37.5%
Age 8 59 41% 52 36.1%
Control 57 37.5% 50 32.9%
Total 361 327
Type 1-1 and Type 2-1 both feature the [+animate] head. Therefore, we listed their frequency counts together here. However, these counts could not supply an obvious growing trend, so we transferred these counts into percentage format, as in Figures 4-10 and 4-11:
59
Figure 4-10 N1-preference Reading for Type 1-1 and Type 2-1
Figure 4-11 N2-preference Reading for Type 1-1 and Type 2-1
When it was a N1-preference reading, Type 2-1 with the inanimate modifier was preferred by all age groups. When it was a N2-preference reading, Type 1-1 with the animate modifier was preferred by all age groups. All the age groups performed
60
similarly.
(4) N2 was [-animate]:
Table 4-8 shows the frequency counts of all subjects’ performance when the head position was [-animate].
Table 4-13 N1-preference Reading When N2 was [-animate]
T1-2: Animate-Inanimate T2-2: Inanimate-Inanimate N1-preference Frequency count % Frequency count %
Age 4 58 51.8% 54 48.2%
Age 5 81 63.3% 66 51.6%
Age 6 89 65.4% 75 55.1%
Age 7 91 66.9% 88 64.7%
Age 8 97 67.4% 99 68.7%
Control 110 72.4% 108 71.1%
Total 526 490
61
Table 4-14 N2-preference Reading When N2 was [-animate]
T1-2: Animate-Inanimate T2-2: Inanimate-Inanimate N2-preference Frequency count % Frequency count %
Age 4 54 48.2% 58 51.8%
Age 5 47 36.7% 62 48.4%
Age 6 47 34.6% 61 44.9%
Age 7 45 33.1% 48 35.3%
Age 8 47 32.6% 45 31.3%
Control 42 27.6% 44 28.9%
Total 282 318
Type 1-2 and Type 2-2 both feature the [-animate] head. Therefore, their frequency counts were listed together here. Again, it was not easy to show an obvious growing pattern we would like to see, so we would transfer these counts into percentage format, as in Figures 4-12 and 4-13:
62
Figure 4-12 N1-preference Reading for Type 1-2 and Type 2-2
Figure 4-13 N2-preference Reading for Type 1-2 and Type 2-2
When it was a N1-preference reading, Type 1-2 with the animate modifier was preferred except for Age 8. When it’s a N2-preference reading, Type 2-2 with the inanimate modifier was preferred except for Age 8. Since Age 8 performed slightly
63
different from other groups, there might be a turning point at Age 8 in the growing pattern.
4.2.2 General Discussion
Generally speaking, when the modifier and head are controlled as [+animate], the effect is stronger than the time when they are controlled as [-animate]. When the modifier or head is controlled as [+animate], all the age groups have the same tendency of N1-preference or N2-preference reading. In contrast to that, when the modifier or head is controlled as [-animate], the tendency is not that clear.
When the modifier is controlled as [-animate], only half of the six age groups have the same tendency. When the head is controlled as [-animate], Age 8 children behave differently from other age groups. All in all, these imply that animacy, [+animate], has a stronger effect than inanimacy.
This makes sense because animacy, always associated with actions and vivid lives, provides a more concrete or more familiar concept than inanimacy. Even though we are exposed to a variety of contexts, "the most eye-catching among them are organisms and objects: people, animals, and all kinds of everyday artifacts" (Ungerer
& Schmid 1996: 1). That is, these categories are impressive for most of us. It follows that they are important in cognition. Moreover, Marchand (1969:44) holds that
"compounds chiefly denote living beings (persons or animals), but there are also plant names and a few words denoting various things.
Huang (2008) compared the three groups of compounds with similar constituents.
The data displays a noteworthy point: The proportion of hybrid interpretations for animal-animal compounds is much higher than that for artifact-artifact and plant-plant compounds. The majority of hybrid interpretations fall on animal-animal compounds;
the number of hybrid interpretations for artifact-artifact compounds comes in second;
64
least hybrid interpretations are found in plant-plant compounds. Therefore, the preferable use of hybrid comprehending strategies may not be generalized to all similar pairs, but needs to be narrowed down to specific ones, particularly animal-animal compounds.
Chang (2008) considered the issue of animacy of the structural subject (the patient role) and object (the agent role). Animacy has been widely discussed in the literature in that it exerts certain influences on the acceptance for passives (Bock 1986, Dewart 1979, Harris 1978, Kuo 1995, Lempert 1984, 1989, 1990, Tseng 1997).
Generally speaking, it is believed that the prototypical subject should be conceptualized as a category, which always comprises cognitively salient properties or components such as animate, cause, and dynamism (cf. Bates and MacWhinney 1982, Hinton 1981, Lempert 1989). Accordingly, an animate subject in passives should be more qualified than an inanimate subject. On the other hand, when we look into the thematic role, the prototype of an agent is also believed to be animate (Givon 1984). Under this proposition, it is suggested that a passive with an animate agent should be the unmarked form (cf. Kuo 1995). In Chang’s (2008) study, she discussed the relation between the passive and subject/object animacy, and her study confirmed the literature that animacy did play a role in children’s language acquisition (cf. Bock 1986, Dewart 1979, Harris 1978, Kuo 1995, Lempert 1984, 1989, 1990, Tseng 1997), which is an idea matching what we found for section 4.2.