Background

Motion is a universal phenomenon that humans around the world speaking all kinds of languages experience every day. However, the ways in which this experience is encoded in different languages can vary in typologically different patterns (Talmy 1975, 1985; Aske 1989; Slobin 1996). A motion event is “a situation containing movement or the maintenance of a stationary location” (Talmy, 1957). In a motion event, six motion semantic components are involved: “Motion”, “Path”,

“Figure”, “Ground”, “Manner”, and “Cause”. Languages encode motion semantic components into different surface elements which include “verbs”, “adpositions", and “subordinate clauses”. Some languages tend to encode Path into verbs, such as Spanish (e.g. El chico entró el cuarto “The boy entered the room”). Other languages tend to express Path with adpositions, such as English (e.g. The boy walked into the room). Talmy (1985) classified these languages into two different categories : Verb-framed languages (V-Verb-framed languages), which tend to encode Path with the verbs.

Such languages include Spanish, French, Arabic, and Japanese; and Satellite-framed languages (S-framed languages), which tend to encode Path with a satellite component adhering to the main verb. Such languages include English, German, and Mandarin. Because Verb-framed languages and Satellite-framed languages are typologically different, speakers of these languages focus on different semantic components to encode motion.

‧

國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

In addition to the typological differences found in encoding motion events, many studies have also been conducted on children’s motion expressions across different languages (Choi & Bowerman 1991; Özyurek & Özçaliskan 2000;

Parafragou et al. 2002; Hickmann & Hendriks, 2010). Each language trains its young learners to attend to a particular dimension in the course of language acquisition (Slobin, 1996). Thus, when learning their native language, children must determine which aspects of their mental representations they should focus on, and which aspects can be realized through the grammar of their native language. Two questions continue to intrigue researchers: “When do children learn to think and speak in a language-specific way when they talk about motion?” and “how do children acquire adult-like and language-specific rhetorical style in motion encoding?” Many researchers have reported that the motion expressions of children learning different languages begin to show some language-specific fashion early on (Choi &

Bowerman 1991; Özyurek & Özçaliskan 2000; Parafragou et al. 2002). Children start to encode the semantic elements of motion lexically and syntactically in accordance to their native languages allow them to. As children grow older, their motion expressions continue to be affected by the language-specific rhetorical style and become more adult-like.

The typology of Mandarin in motion events has been investigated by many researchers. Talmy (1991) suggested that Mandarin is an S-framed language. Tai (2003) suggested that Mandarin more closely resembles a V-framed language. Other researchers have argued that Mandarin falls within neither of these frameworks and should be classified as an equipollently-framed language, a third type of motion

‧

國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

encoding category (Slobin, 2004; Chen, 2005; Chen & Guo, 2009). The way Mandarin-speaking children express motion has attracted the interest of many scholars (Chen, 2005; Guo & Chen, 2009; Lin, 2006; Huang, 2012). Research on Mandarin speaking children’s motion expressions has focused almost exclusively on children’s motion expressions when narrating stories (Chen 2005, Guo and Chen 2009, Lin 2006). Chen (2005) and Guo and Chen (2009) studied Beijing Mandarin speaking children’s narration of the Frog Story, and found that children’s linguistic constructions of motion tended to become more equipollently-framed as the language learner grows older and contends that both Manner and Path are salient in children’s encoding. In another study, Lin (2006) used animated motion pictures to elicit Taiwanese Mandarin children’s motion narration, and found that Manner played an important role in motion event expressions. Among all these studies, only Huang (2012) conducted research on the way Mandarin speaking children encode motion in daily conversation rather than in narration. In her study, she investigated children’s language in daily conversation, and found that Path was the most productive verb.

The results of these studies above on Mandarin children’s motion acquisition are somewhat inconsistent because their findings were different in the aspect of the use of motion verbs and the constructions in children’s motion encoding. First, Manner verbs were more frequently used by children in the studies with narrative data, while Path verbs were more frequently used in the study with conversational data. Second, the preferable motion construction was three-verb construction (Manner + Path + Deictic) in the studies with narrative data, but one-verb construction in the study with conversational data. Previous studies seem to present

‧

國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

conflicting results on how Mandarin speaking children learn to express motion events in different discourse genres.

In fact, some previous studies have implied that children’s language expressions could be different in narration and conversation (MacLachlan &

Chapman, 1988; Dollaghan et al., 1990; Wagner et al., 2000), such as fluency, syntactic complexity and mean length of utterance (MLU). This begs the question of whether children’s expressions of motion event also show different characteristics in certain aspects. While most researchers have collected data of children’s motion expressions from elicited narratives (Slobin 1996; Özyurek & Özçaliskan 2000;

Chen 2005; Parafragou et al. 2006), only a small number of them have collected data from natural conversations (Choi & Bowerman, 1991; Zheng & Golding-Meadow, 2002; Hohenstein, 2013), and even fewer have studied the ways in which genre differences may affect children’s expressions of motion events. In one study, Selimis and Katis (2010) examined the role of context in acquisition of motion encoding, and found that both English-speaking and Greek-speaking children used more types of manner verbs in conversation than in narration when encoding motion events.

Selimis and Katis have proposed that Manner might be more prominent in conversation. However, more studies are needed to support this claim.

In conclusion, previous research have studied the emergence and the development of children’s language-specific patterns of encoding motion. However, few studies have examined how genre difference affects the development of children’s motion encoding, as implied by the inconsistency in the studies of Mandarin-speaking children. Previous studies on how Mandarin-speaking children

‧

國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

encode motion have not included both narration and daily conversation in their data sets, and the coding and the organization of the data is not consistent. This has yielded conflicting conclusions, and whether children express motion differently in different genres still remains unclear. The aim of this study is to achieve a more comprehensive understanding of how Mandarin-speaking children encode motion by re-examining and comparing children's motion expressions in conversation and narration. This study is also valuable in examining how strong language-specific patterns in children’s motion expressions reveal in different discourse genres.