This study aims to explore linguistic differences between fictive motion expressions and physical motion event expressions in Chinese. Although both types of expressions are associated with dynamic linguistic forms, they describe different types of semantic content. Using authentic data, this study examines motion verbs, motion verb constructions, the complexity of ground elements, and alternative manner expressions for Chinese fictive motion events, the results of which are compared with those of previous studies on physical motion events. It is found that Chinese fictive motion expressions are very different from Chinese physical motion event expressions in terms of the above four aspects. The results lend support to the hypothesis that fictive motion occurs more in verb-framed languages.
This study aims to explore the linguistic differences of fictive motion expressions and physical motion event expressions.
Within Talmy’s framework of motion event theory, a motion event is mainly composed of four semantic elements, i.e., the Figure, the Ground, the activating process, and the association function (Talmy, 2000b: 226–227). A motion event either designates a physical motion event with displacement or a stationary locative event (Talmy, 2000b: 25). Both physical motion events and locative events have a Figure and sometimes a Ground. Their differences lie in, firstly, the activating process, which is “motion” for physical motion events and “stationariness” for locative events; and secondly, the association function, which is “the path of the motion on the part of the Figure with respect to the Ground” for physical motion events and “the site occupied by the Figure in terms of the Ground” for locative events (ibid).
Fictive motion is represented in sentences that describe static spatial entities with dynamic linguistic forms (motion verbs and directional prepositions). This definition shows that fictive motion expressions have the same semantic elements as locative events but use linguistic forms typically employed to represent physical motion events. In this study, fictive motion events involve the fictive movement on the part of the Figure as represented in fictive motion sentences.
Fictive motion events and physical motion events share similar types of linguistic forms, i.e., motion verbs and directional prepositions, but they describe different semantic content. Physical motion event expressions faithfully depict a situation involving physical motion, whereas fictive motion event expressions unfaithfully refer to a static entity with dynamic linguistic forms. Experimental studies have shown that the comprehension of fictive motion sentences involves the mental simulation of motion (Matlock, 2004, 2006; Matlock, Holmes, Srinivasan, & Ramscar, 2011; Saygin, Mccullough, Alac, & Emmorey, 2010; Singh & Mishra, 2010). Despite this connection of fictive motion with mental simulation of motion, the static semantic content may still render the linguistic features of fictive motion sentences different from those of physical motion event sentences. This study tries to explore the linguistic differences between these two types of sentences.
In the next section, the most relevant literature will be reviewed, including the study on the lexicalization patterns of physical motion event expressions, the study on the linguistic features of fictive motion events, and the study on locative events. Section 3 explains some issues about the methodology. Results are illustrated in Section 4. Section 5 focuses on the discussion of the results, after which conclusions are made in Section 6.
2 Literature Review
2.1 The Studies on the Lexicalization Patterns of Physical Motion Events
Talmy (2000b: 213–288) classified world languages into satellite-framed languages (S-languages) and verb-framed languages (V-languages) based on whether the path of the movement is expressed in the satellite or the verb in motion event expressions. Apart from the location of path information, the two types of languages were also found to be different in other aspects. Relevant here are the expressiveness of the manner information and the complexity of the ground entities. It was found that, compared with V-languages, S-languages are more flexible in expressing manner information because the verb is available to encode the manner information in addition to other linguistic units (Slobin, 2004: 250–253). S-languages are richer in the types of manner verbs used in motion event expressions than V-languages (Slobin, 1997: 458). In addition, manner verbs occur more frequently in S-languages than in V-languages (ibid). With regard to ground entities, S-languages tend to have more ground elements per verb than V-languages, while V-languages are more likely to employ motion verbs without mentioning any ground element (Slobin, 1997: 442).
Since the seminal work of Talmy (2000b), the lexicalization patterns of physical motion events in various languages have been explored extensively. Studies on Chinese physical motion events focus on whether Chinese is an S-language or a V-language (Chen, 2007; Chen & Guo, 2009; Shen, 2003; Slobin, 2004; Tai, 2003; Talmy, 2000b) and the transition of Chinese from being a V-language to being an S-language (Shi & Wu, 2014). The status of modern Chinese in this dichotomy is a contentious issue, but a general agreement has been reached that Chinese has been under the development from being a V-language to being an S-language.
The data used in analyzing physical motion event expressions in modern Chinese fall into two major types, i.e., introspective data (Shen, 2003; Tai, 2003; Talmy, 2000b) and usage-based data (Chen, 2007; Chen & Guo, 2009), within which usage-based data are either elicited spoken motion event expressions through experiments (Chen, 2007) or motion event sentences collected from written materials (Chen & Guo, 2009). Each type of data has its advantages and disadvantages. Introspective data are easy to manipulate and played an important role in the early development of the theory, but they cannot show how language is used in real situations. Elicited spoken data can test colloquial usage. They represent lab speech and are confined in terms of the materials used as well as participants’ online time pressure. Usage-based written data can provide a complicated yet authentic picture of how motion events are expressed. For the current study, I will focus on studies with written data for the reason that this is a comparative study between physical motion events and fictive motion events, and that fictive motion expressions are more comprehensively observed in written texts.
2.2 The Studies on the Linguistic Features of Fictive Motion Events
The linguistic phenomenon of fictive motion has been noticed widely (Bennett, 1975: 35–40; Dowty, 1979: 67; Jackendoff, 1983: 168–174; Levin, 1993: 256; Thompson, 2013: 91–121), but Talmy (1996, 2000a) is the first one who has extensively examined it. Fictive motion expressions describe static scenes with dynamic linguistic forms. This apparent incongruence between the semantic content and the linguistic conceptualization has been examined in terms of cognitive mechanisms, such as mental scanning (Langacker, 1986, 1999, 2005), Conceptual Integration Theory (Fauconnier & Turner, 2002: 376–380), Conceptual Metaphor Theory (Lakoff & Turner, 1989: 144–146), and metonymy (Caballero, 2006: 161; 239; Fauconnier & Turner, 2002: 378).
More relevant here are the studies on the linguistic features of fictive motion expressions. Matsumoto (1996) made the proposal that one type of fictive motion expressions, namely, coextension paths, is constrained when expressing manner information and path information. A coextension path sentence is a “depiction of the form, orientation, or location of a spatially extended object in terms of a path over the object’s extent” (Talmy, 2000a: 138). The example below is a typical coextension path sentence in English. In this sentence, the configuration and location of the fence is defined in terms of a fictive movement along the fence with respect to its two ends.
(1) The fence goes from the plateau to the valley. (ibid)
What Matsumoto (1996) proposed is that, for coextension path expressions, some path information must be included and manner information irrelevant to the path component cannot be expressed. This is generally borne out in different languages (Cappelli, 2012; Rojo & Valenzuela, 2003).
Another aspect is comparing the linguistic features of physical motion event expressions with fictive motion expressions. One type of fictive motion, i.e., visual paths, was compared with physical motion events in terms of the complexity of the path (Slobin, 2009). S-languages tend to have more complex path information than V-languages (Slobin, 2004: 239–240). Four languages of different types (two S-languages and two V-languages) were examined. It was found that the linguistic features in terms of complex path information of physical motion event expressions are carried over to visual path sentences (Slobin, 2009: 211–212). Similar findings were observed in a study comparing English (an S-language) and Spanish (a V-language) when they express visual paths and physical motion events (Cifuentes-Férez, 2014). When expressing visual paths, English uses more and richer manner verbs than Spanish (Cifuentes-Férez, 2014: 234), and English is better at encoding complex combinations of Path plus Ground (Cifuentes-Férez, 2014: 226).
2.3 The Studies on Locative Events
Studies on locative events are relatively fewer, though locative events are a composing member of motion events within Talmy’s motion event theory. Fictive motion expressions are one of several linguistic strategies in representing a locative scene. It was found that when expressing locative scenes, Serbian (an S-language) uses more manner verbs than French (a V-language) (Stosic & Sarda, 2009: 50–51). This indicates that, similar to the case of motion event expressions, manner is more salient in S-languages than in V-languages for sentences depicting locative situations. On the other hand, French adopts more fictive motion expressions than Serbian (Stosic & Sarda, 2009: 57). It was thus hypothesized that fictive motion expressions are limited in manner-salient languages, i.e., S-languages (Stosic & Sarda, 2009). This study is interesting here because this hypothesis is relevant to our findings.
This study is a comparative one. It mainly compared the linguistic features of fictive motion and physical motion events in Chinese. For physical motion events, the results from a previous study (Chen & Guo, 2009) were adopted. Chen & Guo (2009) studied the linguistic features of physical motion events in Chinese with data collected from novels. In addition to this study, data from two other studies (Özçalışkan & Slobin, 2003; Slobin, 1996) were also cited in order to compare Chinese fictive motion expressions with typical S-languages and V-languages. Fictive motion expressions were collected manually from written texts (See Section 3.3).
3.2 Aspects to Compare
There are four aspects to compare. The first one is motion verbs in terms of both type and token. The type of motion verbs is decided by the semantic elements expressed. Possible semantic elements include path information, manner information, and motion information without obvious path or manner elements. The second point for comparison is associated with the first one, and it is the motion verb constructions employed. Serial verb constructions are frequently observed in Chinese, so they are highly expected to be found. Motion verb constructions refer to the types of combination of semantic elements expressed in the verbal predicates (Chen & Guo, 2009: 1755–1757). The third aspect to compare is the complexity of the ground elements, which is represented by the number of ground entities encoded. The last thing to look at is the employment of additional manner information apart from the manner information expressed by the main verb. These four aspects are chosen mainly because they have been studied for physical motion event expressions (Chen & Guo, 2009; Özçalışkan & Slobin, 2003; Slobin, 1996).
3.3 Fictive Motion Data
Fictive motion data were collected manually from published written texts (See the Appendix). This set of texts includes travelling essays and notes. The contents of this set of published books are not purely introductions of different places. Rather, the descriptions of locations and scenic spots are threaded by stories involving human beings. Novels, though used for analyzing physical motion expressions, were not chosen for the collection of fictive motion data because fictive motion expressions rarely occur in novels.
This study focuses on one type of fictive motion expressions, namely, coextension paths. Coextension paths are the most frequently observed and most often discussed type of fictive motion. The definition and illustration of coextension paths can be found in Section 2.2.
The written texts were examined sentence by sentence and all coextension path expressions were manually selected. There are 355 episodes1 containing coextension paths, and 462 clauses2 are identified within the 355 episodes. In Chen & Guo (2009), there are 180 episodes containing 520 clauses.
The data were coded with uam CorpusTool (O’Donnell, 2008), which allows manual annotation of collections of texts. The annotation schemes are based on the four aspects to compare, namely, the motion verbs, motion verb constructions, the number of ground entities in each clause, and additional manner information apart from the main verb. After the annotation, the findings were compared with those in the three studies (Chen & Guo, 2009; Özçalışkan & Slobin, 2003; Slobin, 1996).
4.1 Motion Verb Use: Types
As mentioned in Section 3.2, the type of a motion verb is based on the semantic information encoded. Four types of semantic elements were identified in the motion verbs employed in Chinese fictive motion expressions, including manner, deictic, path, and general extension. The manner information covers dimensions such as the motor pattern, rate, rhythm, posture, affect, and evaluative factors (Slobin, 2004: 255). Deictic information can either be ‘toward the speaker’ and ‘in a direction other than toward the speaker’ (Talmy, 2000b: 56). The path information pertains to the trajectory over which the Figure moves with respect to another entity (Chen & Guo, 2009: 1755). General extension is motion that is not obviously associated with manner or path; rather, it is about the extending motion of some entity. General extension is a semantic element uniquely encoded in verbs of coextension paths, as illustrated in the following example. The verb 延伸 (yánshēn; extend) in (2) is a general extension verb. In this sentence, the extension of the road is described as the extending of the road.
That is a highway that extends out from the airport.
Table 1 below presents all types of verbs and their frequencies in Chinese fictive motion expressions, including manner verbs, path verbs, deictic verbs, general extension verbs, manner + path verbs, general extension + manner verbs, general extension + path verbs, general extension + deictic verbs, and path + deictic verbs.
Manner verbs, path verbs, deictic verbs, and general extension verbs are verbs that encode the corresponding semantic elements. Notice that apart from verbs with a single semantic element, this study also includes verbs with more than one semantic element, including manner + path verbs,4 general extension + manner verbs, general extension + path verbs, general extension + deictic verbs, and path + deictic verbs. As can be seen from Table 1, these polymorphemic verbs are usually compound words. They are identified as one word because in Modern Chinese Dictionary (2012), each of them is listed as one word entry. Such words were deleted from the study in (Chen & Guo, 2009) because they couldn’t be classified easily and they occurred rarely. In my opinion, one way to deal with such compound words is to treat them as polymorphemic words. Furthermore, they occur so frequently in fictive motion expressions that they cannot be simply ignored. The sentence in (3) contains a manner + path motion verb.
The highway meanders through the dense forest.
Neutral verbs were identified in (Chen & Guo, 2009) and were treated as manner verbs in their analysis. In this study, they were coded as manner verbs from the start.
As stated before, S-languages and V-languages are different in the type and token of manner verbs. Manner verbs tend to be richer in terms of type and more frequent in terms of frequency in S-languages than in V-languages (Özçalışkan & Slobin, 2003). For example, as shown in Table 2 below, English, as an S-language, has more types of manner verbs than Turkish, which is a V-language (ibid). The number (41) of manner verb types in Chinese physical motion event expressions is higher than that in Turkish (26) and lower than that in English (64), being almost in the middle of these two.
The number (29) of manner verb types in Chinese fictive motion expressions is also between that in English and that in Turkish, but it is much closer to Turkish, which is the V-language’s end. In terms of the number of verb types, Chinese fictive motion expressions have significantly fewer types of manner verbs than Chinese physical motion event expressions (z = 6.47, p < .01, 99% ci: 30.5% – 64.74%).6 The lower number of manner verb types in Chinese fictive motion expressions is not due to the relatively lower number of fictive motion expressions, as we can see the number (63) of the types of path verbs in fictive motion expressions is significantly higher than that in physical motion event expressions in Chinese (13) (z = 3.12, p < .01, 99% ci: 5.91% – 40.75%). This indicates that the types of manner verbs that can be used in Chinese fictive motion expressions are limited if we compare them with Chinese physical motion expressions and typical S-languages.
4.2 Motion Verb Use: Tokens
Following the line in the last section, manner verbs occur more frequently in S-languages than in V-languages. As illustrated in Table 3 below, English (an S-language) has a much higher percentage of manner verbs (53%) than path verbs (27%), whereas Turkish (a V-language) uses more path verbs (59%) than manner verbs (34%). The percentage (45.3%) of manner verbs in Chinese physical motion event expressions fall between these two languages, and the difference between the percentage of manner verbs (45.3%) and that of path verbs (53.1%) is not that big.
Table 3 shows that the case of fictive motion expressions in Chinese goes to the extreme of V-languages. Chinese fictive motion expressions use much more path verbs (55.4%) than manner verbs (8.5%). The percentage of manner verbs (8.5%) is even lower than that in Turkish (34%), and it is significantly lower than that in Chinese physical motion event expressions (45.3%) (z = 15.51, p < .01, 99% ci: 31.28% – 42.32%). The column labeled as “other” covers cases of manner + path verbs, general extension + manner verbs, general extension + path verbs, general extension + deictic verbs, and path + deictic verbs, among which manner + path verbs were observed 13 times and general extension + manner verbs occurred once. If we add manner + path verbs and general extension + manner verbs into manner verbs, the percentage is lower still.
4.3 Motion Verb Constructions
Chinese allows serial verb constructions that can combine up to three verbal components at the same time. It can be observed in Table 4 (adapted from Chen & Guo, 2009: 1760) that motion verb constructions in Chinese physical motion event expressions include manner + path + deictic, manner + path, manner + deictic, manner only, path + deictic, path + path, path only, and deictic only. These eight constructions fall into four types,8 i.e., manner + path, manner only, path only, and deictic only. Among these four types, manner + path accounts for a large percentage (62.31%). However, Chinese is able to use manner only and path only motion verb constructions, suggesting that both manner verbs and path verbs can be used as independent main verbs in Chinese physical motion event expressions.
There are more types of motion verb constructions in Chinese fictive motion expressions than in Chinese physical motion expressions since one more semantic element, i.e., general extension, is involved. As shown in Table 5, five additional types are identified, including general extension + path + deictic, general extension + path, general extension + manner, general extension + deictic, and general extension only.
The distribution of motion verb constructions in fictive motion expressions is very different from that in physical motion expressions. The construction manner + path only takes up 12.77%, which is significantly lower than the percentage in physical motion expressions (62.31%) (z = 15.89, p < .01, 99% ci: 42.29% – 0.56.79%). Constructions containing only manner verbs (3.03%) have a much lower proportion than those containing only path verbs (48.27%). The fewer employment of manner verbs compared with path verbs is also instantiated through the comparison of the construction general extension + manner (0.22%) and the construction general extension + path (17.10%). Constructions containing general extension verbs are an important composing element of constructions employed to represent fictive motion.
4.4 Description of Ground Elements
As reviewed in Section 2.1, S-languages tend to encode more ground elements than V-languages. In Table 6, English (an S-language; 96%) employs more clauses with ground elements than Spanish (a V-language; 81%) does when expressing physical motion events. The case (83%) of Chinese physical motion event expressions is more similar to that of Spanish. When it comes to Chinese fictive motion expressions, the clauses (61%) encoding ground elements are significantly lower than Chinese physical motion event expressions (83%) (z = 7.74, p < .01, 99% ci: 15.37% – 28.63%), and it is even fewer than Spanish.
It can be seen from Table 7 that a considerable number (39%) of Chinese fictive motion expressions do not have any ground element. Similar to Chinese physical motion expressions, Chinese fictive motion expressions usually only encode one ground entity when it comes to plus-ground clauses. Only 2% of the clauses have two ground entities, and none encode three ground entities.
4.5 Alternative Expressions of Manner
Both S-languages and V-languages use alternative manner expressions in addition to manner verbs. In Table 8, it can be seen that the number of the total motion descriptions containing alternative manner expressions are similar for Chinese, English, and Spanish physical motion event expressions (102, 107, and 113 respectively). Also, all of the three languages use more adverbials than descriptions in expressing physical motion events (98/4, 73/34, and 93/20 respectively). This tendency is observed for Chinese fictive motion expressions (154/24). One difference between Chinese fictive motion expressions and the other three types of motion expressions is that the number of motion descriptions containing alternative manner expressions in the former one (178) is much higher than the latter three (102, 107, and 113 respectively), and the former one is significantly higher than that of Chinese physical motion event expressions (z = 6.55, p < .01, 99% ci: 12.19% – 25.63%).
S-languages and V-languages are different in terms of whether the alternative manner expressions modify manner verbs or non-manner verbs. As shown in the last column of Table 8, English (an S-language; 73%) tends to use alternative manner expressions to modify manner verbs. On the contrary, Turkish (a V-language; 39%) tends to use alternative manner expressions to modify verbs other than manner verbs. Chinese physical motion expressions are different from both English and Turkish in that they use all the alternative manner expressions to modify manner verbs. The percentage of alternative manner expressions modifying manner verbs in Chinese fictive motion expressions (19%) is even less than that in Turkish physical motion expressions.
5.1 Motion Verbs and Motion Verb Constructions
Chinese physical motion expressions employ more types of manner verbs than path verbs. In contrast, Chinese fictive motion expressions use much more types of path verbs than manner verbs. Compared with Chinese physical motion expressions, Chinese fictive motion expressions have more path verb types but less manner verb types. The number of manner verb types in Chinese fictive motion expressions is close to that in Turkish, which is a typical V-language.
In terms of frequency, the percentage of path verbs in Chinese fictive motion expressions is similar to that in Chinese physical motion expressions and Turkish (a V-language), but the percentage of manner verbs in the former is much lower than the latter two as well as English (an S-language). Chinese fictive motion expressions use path verbs most, followed by general extension verbs.
The predominant motion verb construction in Chinese physical motion expressions is manner + path, i.e., the verb-complement construction. Manner only and path only constructions are identified but their percentages are relatively lower. The distribution of motion verb constructions is different in Chinese fictive motion expressions. The construction type “manner + path” is significantly fewer. The major construction type is the path only construction, followed by the general extension only construction.
The features of motion verbs and motion verb constructions in fictive motion expressions can be explained by the function of fictive motion sentences. As reviewed in Section 2.1, Matsumoto (1996) proposed two conditions on coextension paths, i.e., the manner condition and the path condition. For the manner condition, no manner information can be encoded if it is irrelevant to path; for the path condition, some path information must be expressed in coextension paths. These two conditions can generally, though not strictly, be applied to Chinese fictive motion expressions. The main reason is that it is the path-like configuration of the Figure that motivates the use of coextension paths. Thus, path is the core element in such expressions. Some of the manner verbs in fictive motion expressions are indeed helpful in depicting the path of the fictive movement, such as the manner verb 刺 (cì; stab) in (4), but some of them seem to be less related to path, such as 冲 (chōng; rush) and 扎 (zhā; jump) in (5). The verb 刺 (cì; stab) in (4) not only depicts the manner of the fictive movement, but also indicates the configuration of the Figure (snow peak) and the path of the fictive movement. It is not surprising to find such manner verbs since sometimes people deliberately use fictive motion expressions as a rhetorical device to make the writing more vivid and expressive.
The snow peak, which is white and glistening, stabs into the vast blue clear sky.
jiǎo-shān-shàng zhí chōng-xià-lái, yì-tóu
Cape-Mount-on all-the-way rush-down-come head
zhā-jìn-le bó-hǎi àn-biān.
jump-into-pfv Bohai-Sea bank
The Great Wall rushes down all the way from the Cape Mount of Yan Mountain and jumps into the bank of the Bohai Sea.
5.2 Ground Elements
Compared with Chinese physical motion event expressions, Chinese fictive motion expressions employ significantly fewer ground elements. The ground elements encoded in Chinese fictive motion expressions are even 20% fewer than those in Spanish, which is a V-language. A considerable amount of Chinese fictive motion expressions do not have any ground element. According to findings from previous studies that V-languages tend to encode fewer ground entities, fictive motion expressions behave more like a V-language in terms of ground encoding.
The function of coextension path sentences is to delineate the configuration of a linear, extending spatial entity. When no ground element is used as the reference point to define the scope of the extending Figure, the boundary of the Figure usually can be inferred from the context. In (6), the temples of different sizes are conceptualized as one long, extending entity that extends to some unspecified place. We can infer from the adverbial phrase 一路 (yílù; all-the-way) that the temples are located along the road.
zhè-eryěshì dà-bó-gōng, nà-er yě shì
hereagainis Tua-Pek-Kong-Temple there again is
dà-bó-gōng, dà-dà-xiǎo-xiǎo-de tǔ-dì-miào yí-lù
Tua-Pek-Kong-Temple big-and-small temple all-the-way
Here is a Tua-Pek-Kong-Temple, and there is a Tua-Pek-Kong-Temple. Temples of various sizes were built along (something).
5.3 Alternative Manner Expressions
Chinese fictive motion expressions employ more alternative manner expressions than Chinese physical motion expressions, as well as English (an S-language) and Turkish (a V-language). This seems to violate Matsumoto’s manner condition mentioned above. However, a closer examination of the alternative manner expressions in Chinese fictive motion expressions reveals that, in most cases, alternative manner expressions contribute to the depiction of the path of the fictive movement and the configuration of the Figure, as shown below in sentence (7). Of course, similar to manner verbs, some alternative manner expressions are used to enhance the poetic flavor of the writing, as in (8).
Huaihai Road extends eastward from Xujiahui, and it sweeps straight over Shanghai.
When the night falls, the city becomes quiet, and small alleys extend silently.
Alternative manner expressions are used more to modify manner verbs in S-languages than in V-languages. This is within our expectation since there are not so many manner verbs in V-languages. Chinese is an exception in this respect in that all the alternative manner expressions exclusively modify manner verbs in physical motion event expressions. Chinese fictive motion expressions are more similar to V-languages in that the alternative manner expressions tend to modify non-manner verbs. This is mainly due to the low frequency of manner verbs occurring in fictive motion expressions. If we take into account the fact that many manner verbs are associated with the path of the fictive movement, then the percentage of alternative manner expressions modifying pure manner verbs is even lower.
5.4 General Discussion
Based on the differences between S-languages and V-languages in terms of motion verbs, motion verb constructions, the encoding of ground entities, and alternative path expressions, Chinese fictive motion expressions are more like a V-language. More path verb types are employed while limited types of manner verbs are used. In terms of the tokens, path verbs are much more frequent than manner verbs. When it comes to motion verb constructions, the dominant construction is path only construction; and the construction manner + path, which is a salient feature of physical motion event expressions, is used significantly less frequently. The percentage of clauses containing some ground elements is even lower than typical V-languages, and for clauses with some ground elements, there is usually only one ground entity. Alternative manner expressions modify manner verbs less frequently than when they modify non-manner verbs. Table 9 below is a summary of the comparison.
The similarity between Chinese fictive motion expressions and V-languages supports the hypothesis that fictive motion expressions occur more in V-languages. It was hypothesized that in manner-salient languages, namely, S-languages, the use of fictive motion sentences is limited, and that fictive motion expressions are more salient in V-languages than in S-languages (Stosic & Sarda, 2009). This hypothesis is supported by other studies on fictive motion expressions in modern Chinese from the perspective of verbal predicate patterns (Ma, 2015, 2016). Scholars generally agree that ancient Chinese is a V-language and that Chinese has been developing from a V-language to an S-language. It was found that many verbal predicate patterns in modern Chinese fictive motion expressions are inherited from ancient Chinese, such as monosyllabic verbs, disyllabic words, adverb + monosyllabic verb constructions, and Chinese idioms (Ma, 2015, 2016: 34). This is consistent with the findings of this study. Modern Chinese is relatively rich in verbal predicate patterns expressing manner information, and patterns irrelevant to manner information are limited. Manner information can be expressed in fictive motion expressions, but it is encoded much less compared to path information. A large proportion of fictive motion expressions need verbal constructions that do not express manner. As a V-language, ancient Chinese can provide such verbal predicate patterns that do not convey manner information. Thus, ancient verbal predicate patterns are frequently resorted to when we need to use fictive motion expressions. The employment of ancient verbal predicate patterns allows the absence of manner information. Another relevant point is that, while speakers of S-languages are better at describing the process of motion, speakers of V-languages tend to give information about the background of motion events by describing aspects of the static scene (Slobin, 1997). Fictive motion is one of the linguistic devices to depict static scenes.
Many previous studies focus on the similarity and association between fictive motion and physical motion events, such as the involvement of mental simulation of physical motion (Matlock, 2004, 2006) or mental scanning (Langacker, 1986, 1999, 2005) and the carrying over of linguistic features from motion event expressions to fictive motion expressions (Cifuentes-Férez, 2014; Slobin, 2004, 2009). The differences between fictive motion expressions and physical motion event expressions do not gain much attention. This study examined the differences between fictive motion expressions and physical motion event expressions through comparing their linguistic features. The comparison between the linguistic features of Chinese fictive motion sentences and those of Chinese physical motion event expressions, with English (a typical S-language) and Spanish and Turkish (typical V-languages) as reference languages as well, leads to the conclusion that Chinese fictive motion events and physical motion events are represented differently, and the linguistic features of Chinese fictive motion expressions are more like those of a V-language. Despite the facts that motion is simulated in the comprehension of fictive motion (Matlock, 2004, 2006) and that the linguistic features of expressions representing physical motion events are transferred to those representing fictive motion (Slobin, 2009; Cifuentes-Férez, 2014), fictive motion is still represented differently from physical motion events in language. The words and patterns used in fictive motion sentences are confined by the static nature of the scenes described. This conclusion is probably also associated with the importance of path information which motivates the use of fictive motion sentences as proposed by Matsumoto (1996). The necessity of making the path information explicit and manner information implicit results in the choice of path verbs and patterns of V-languages. Further studies are needed to examine whether fictive motion occurs more in V-languages, whether fictive motion suppresses linguistic features of S-languages, and also the reasons for these interesting phenomena.
Appendix Data Source
Notes about Living in a Mountain, Qiuyu Yu, 1995, Taipei: Er-ya Press.
A Cultural Journey, Qiuyu Yu, 2001, Shanghai: Oriental Publishing Centre.
No Boundary for Travellers, Qiuyu Yu, 2001, Beijing: Huayi Publishing House.
Sigh in a Thousand Years, Qiuyu Yu, 2002, Beijing: Writers Publishing House.
Run away for Fifteen Years, Qiuyu Yu, 2004, Haikou: Nanhai Publishing Company.
Yu Qiuyu’s Prose: The Illustrated Collector’s Edition, Qiuyu Yu, 2005, Beijing: Peoples Literature Publishing House
《东西南北人：中国人的性格与文化》，博文(Ed.), 2001, 北京：当代世界出版社。
People from East, West, South and North: Chinese Personality and Culture, Wen Bo (Ed.), 2001, Beijing: The Contemporary World Press.
Excellent Chinese Travel Essays in the 20th Century, Fei Lin (Ed.), 2004, Beijing: People’s Literature Publishing House.
Modern and Contemporary Travel Essays by Famous Scholars with Photos:
The Temple of Earth and Me, Tiesheng Shi, 2002, Jinan: Shandong Pictorial Publisher House.
Rain in Kunming, Cengqi Wang, 2004, Jinan: Shandong Pictorial Publisher House.
Story of the Hulan River, Hong Xiao, 2003, Jinan: Shandong Pictorial Publisher House.
Qinhuai River with the Sound of Oars and the Shadow of Lights, Ziqing Zhu, 2002, Jinan: Shandong Pictorial Publisher House.
Notes When Travelling in Europe, Ziqing Zhu, 2002, Jinan: Shandong Pictorial Publisher House.
Cambridge in my Eyes, Zhimo Xu, 2002, Jinan: Shandong Pictorial Publisher House.
Spring of Diaotai, Dafu Yu, 2002, Jinan: Shandong Pictorial Publisher House.
Snowscape in the West Lake, Jingwen Zhong, 2004, Jinan: Shandong Pictorial Publisher House.
Memories of the Lake, Juren Cao, 2002, Jinan: Shandong Pictorial Publisher House.
Notes When Travelling in Beihai, Xiang Zhu, 2003, Jinan: Shandong Pictorial Publisher House.
At the Bottom of Flaming Mountains, Xianlin Ji, 2004, Jinan: Shandong Pictorial Publisher House.
Cappelli G. 2012. Travelling in space: Spatial representation in English and Italian tourism discourse. Textus: English Studies in Italy , 25(1): 19–36.
Chen L. , & Guo J. 2009. Motion events in Chinese novels: Evidence for an equipollentaly-framed language. Journal of Pragmatics , (41): 1749–1766.
Cifuentes-Férez P. 2014. A closer look at Paths of vision, Manner of vision and their translation from English into Spanish. Languages in Contrast , 14(2): 214–250.
Dictionary Editorial Office, Institute of Linguistics, Chinese Academy of Social Sciences. (Eds.). 2012. 现代汉语词典 [Modern Chinese Dictionary] (6th ed.). Beijing: The Commercial Press.
Langacker R. W. 1986. Abstract motion. In Nikiforidou V. , VanClay M. , Niepokuj M. , & Feder D. (Eds.), Proceedings of the Twelfth Annual Meeting of the Berkeley Linguistics Society, 455–471. Berkley: University of California, Berkeley Linguistics Society.
Langacker R. W. 2005. Dynamicity, fictivity, and scanning: The imaginative basis of logic and linguistic meaning. In Pecher D. , & Zwaan R. A. (Eds.), Grounding Cognition: The Role of Perception and Action in Memory, Language, and Thinking, 164–197. New York: Cambridge University Press.
Ma S. 2015. Verbal predicates in Chinese fictive motion expressions. Springer Lecture Notes in Chinese Lexical Semantics, Lecture Notes in Artificial Intelligence , 9332 : 589 – 598.
Matlock T. 2006. Depicting fictive motion in drawings. In Luchenbroers J. (Ed.), Cognitive Linguistics: Investigations Across Languages, Fields, and Philosophical Boundaries, 67–85. Amsterdam: John H. Benjamins.
Matlock T. , Holmes K. J. , Srinivasan M. , & Ramscar M. (2011). Even abstract motion influences the understanding of time. Metaphor and Symbol , 26: 260–271.
O’Donnell M. 2008. The uam CorpusTool: Software for corpus annotation and exploration. In Callejas Bretones , Carmen M. et al. (Eds), Applied Linguistics Now: Understanding Language and Mind/La lingüística Aplicada Noy: Comprendiendo el Lenguaje y la Mente, 1433–1447. Almería: Universidad de Almería.
Özçalişkan Ş. , & Slobin D. I. 2003. Codability effects on the expression of manner of motion in Turkish and English. In Özsoy A. S. , Akar D. , Nakipoğlu-Demiralp M. , Taylan E. E. , & Aksu-Koç A. (Eds.), Studies in Turkish Linguistics, 259–270. Istanbul: Boğaziçi University Press.
Saygin A. P. , Mccullough S. , Alac M. , & Emmorey K. 2010. Modulation of bold response in motion-sensitive lateral temporal cortex by real and fictive motion sentences. Journal of Cognitive Neuroscience , 22(11): 2480–2490.
Slobin D. I. 1996. Two ways to travel: Verbs of motion in English and Spanish. In Shibatani M. , & Thompson S. A. (Eds.), Grammatical Constructions: Their Form and Meaning, 195–220. Oxford: Clarendon Press.
Slobin D. I. 1997. Mind, code and text. In Bybee J. , Haiman J. , & Thompson S. A. (Eds.), Essays on Language Function and Language Type: Dedicted to T. Givón, 437–467. Amsterdam/Philadelphia: John Benjamins.
Slobin D. I. 2004. The many ways to search for a frog: Linguistic typology and the expression of motion events. In Strömqvist S. , & Verhoeven L. (Eds.), Relating Events in Narrative: Typological and Contextual Perspectives, Vol. 2, 219–257. Mahwah, nj: Lawrence Erlbaum Associates.
Slobin D. I. 2009. Relations between paths of motion and paths of vision: A crosslinguistic and developmental exploration. In Gathercole V. C. M. (Ed.), Routes to Language: Studies in Honor of Melissa Bowerman, 197–221. New York: Psychology Press.
Stosic D. , & Sarda L. 2009. The many ways to be located: The expression of fictive motion in French and Serbian. In Vukanović M. B. , & Grmuša L. G. (Eds.), Space and Time in Language and Literature, 39–60. Newcastle: Cambridge Scholars Publishing.
Talmy L. 1996. Fictive motion in language and “ception”. In P. Bloom, M. A. Peterson, L. Nadel, & M. F. Garrett (Eds.), Language and Space (pp. 211–276). Cambridge, Mass.: The mit Press.
An episode is “the movement of a major protagonist, beginning from a stationary position and continuing to move until arriving at another stationary position where a plot-advancing event occurs” ( Özçalışkan & Slobin, 2003: 260). The protagonists in fictive motion expressions are the fictive Figure.
A clause is “a linguistic unit that contains a unified predicate” (Chen & Guo, 2009: 1755). A predicate can either be a motion verb or a motion verb construction (ibid).
All the Chinese examples are from the data collected for this study unless otherwise specified. The illustration of a Chinese example is composed of four parts. The first line is the original Chinese fictive motion sentence; the second line is the corresponding transcription in pinyin; the third line is the literal translation of each Chinese word in English that may sometimes sacrifice the idiomaticality of English in order to show the original structure of Chinese; and the last line is the English translation of the Chinese sentence. The abbreviations in the second line are based on (Li & Thompson, 1981).
This type of verbs is not an exception to the manner/result complementarity because the path element here specifies the direction or the contour of the movement rather than the result of the movement.
The co-verb 在 (zài; locative-marker) was treated as a path verb when it follows a manner verb in (Chen & Guo, 2009) because it expresses a sense of path in such cases. To make the analyses comparable, this co-verb is also counted as a path verb here.
In this study, the Z-test (see Agresti & Franklin, 2013, p. 470 for the formula of Z-test) was used to decide whether the difference between two proportions is significant or not.
The data in the tables and texts about English and Turkish are from (Özçalışkan & Slobin, 2003), and the data about English and Spanish come from (Slobin, 1996).
It is pointed out in (Chen & Guo, 2009) that verbs encoding manner + deictic information is treated as manner only verbs because the two deictic verbs encode a very general sense of path of motion which primarily focuses on the directional relationship with the speaker and they are categorized as “neutral verbs” by some researchers (Özçalışkan & Slobin, 2003). The same is true for verbs encoding manner + path + deictic information and verbs expressing path + deictic information. To make the analyses comparable, deictic information is not counted when there is other semantic information (such as manner information or path information) preceding it for motion verb constructions in fictive motion expressions.