从认知语言学角度看中英空间关系的表达

本文是一篇语言学毕业论文，本研究比较了64位母语为汉语(N = 32)和英语(N =32)的成人被试关于描述两种静态空间位置关系的实验结果。该实验要求他们描述一组 28 张包含两个静态物体的图片。每张图片都展现出两个物体间的一种支撑/接触或包涵关系。不同的语言有不同的方式来解码空间表达。

1. Introduction

1.1. Background
Space is a semantic domain, including location and motion. Location indicatesthe position of an entity in relation to a reference entity, and an entity in motion passesthrough many locations in space successively. Moreover, spatial expression is alinguistic form based on the abstract representation of an objective world throughhuman conceptual processing (Levinson, 2003). As a result, the notion of space playsan important role in human activities and that spatial experience constitutes theearliest, most essential and most pervasive experience of human life, as demonstratedby Johnson (1987), Landou and Zukowski (2003). Research into space can be tracedback to Kant’s philosophical notion of absolute space, possibly because of theimportance of space to human activities. A typological perspective is needed toinvestigate the domain of space. Given that human beings share a common biologicalheritage in spatial understanding, it is reasonable to speculate that spatial languagewill “closely mirror the contours of non-linguistic spatial understanding” (Bowerman1999, p. 387). However, it turns out that languages differ considerably in how theydescribe space. The idea that linguistic experience might cause reorganization ofnon-linguistic spatial knowledge has recently taken on new prominence in the face ofdiscoveries that languages vary quite widely in how they encode space (Bowerman,1996; Levinson, 2003; Lucy, 1993) For example, the Chinese locational words“在……上” encode a relationship that is encoded by several separate words inEnglish and other languages: English on refers to instances of support relationshipinvolving attachment, such as “the cup on the table” or “the painting on the wall”,whereas above refers to a general support relationship without attachment, such as “Acloud above a mountain”, over is used to describe the vertical support relationshipwith coverage, such as “A table cloth over a table”. English does not draw thisparticular distinction in its inventory of basic spatial terms; hence, English differsfrom Chinese in the way these particular spatial relationships are lexically encoded(Munnich & Landau, 2001). This example highlights the fact that the cross-linguisticdifferences occur because the concepts underlying a single term in a language can bequite different from each other. This study aims to investigate spatial reference,particularly the expression of two typical static locations encoding “ON” and “IN”from a typological point of view.
..........................

1.2. Theoretical Framework
In this study, all the collected data was analyzed on the basis of “the MotionEvent Typology” proposed by Talmy (1985, 1991, 2000), which had been revised byseveral scholars (Ameka & Essegbey, 2006; Zlatev & Yangklang, 2004; Slobin, 2004;Chen, 2005) in the same research domain. Talmy proposes that all the languages canbe pided into two types: satellite–framed and verb-framed. While others argue that itis hard to put all languages into two categories, especially for serial-verb language,bipartite-verb language and generic-verb language. Therefore, the third“equipollently-framed” language was proposed by Slobin (2004) to cover otherlanguages. The categorization of Chinese as satellite-framed language orequipollently-framed still remains controversial.
1.2.1 The Motion Event Typology
In the earliest stage, Talmy observed that languages differ in how they representan event of motion within a sentence (Talmy, 1985). Therefore, The Minimal MotionEvent was proposed to describe a full motion event which consisted of four semanticcomponents: the ‘Figure’ (the entity whose location or motion is at issue), the‘Motion’ (the presence of stasis or motion), the ‘Path’ (the stationary location or thepath of motion of the figure) and the ‘Ground’ (the entity with respect to which theFigure’s Path is characterized), but a number of additional relationships still remainedto be defined.
.......................

2. Literature Review

2.1 Previous International Studies on Spatial Language
It has been traditionally thought that the factor of language universalitydetermines the developmental speed and stages of language acquisition. However,some recent studies have suggested that language specificity becomes more and moresignificant in language acquisition.
Kant’s philosophy of spatial language describe space as a universal cognitiveprimitive. This implied that children develop basic spatial concepts first beforeacquiring spatial terms and relationships. In the Piagetian view (Piaget & Inhelder,1947), children learn to organize space according to sensorimotor, then in terms oftopological notions. They propose that the earliest spatial notions developed areconcerned with object containment and support, and then followed by the concepts ofproximity, separation, surrounding and order. The projective and Euclidean spaceinvolving a system of axes and coordinates can only be constructed at a much laterstage (Johnston, 1987; Bowerman & Choi, 2001; Hickmann, 2006). Some similarresults further demonstrate that spatial languages are acquired in a fixed sequence.According to Johnston and Slobin (1979), English children acquire spatialprepositions in a similar order: the relationship of containment in ought to be learnedbefore the acquisition of contact/support relationships and other relationships on thevertical axis, such as under and above, then relationships along the sagittal axis like infront of, behind, indicating universal mechanisms of language acquisition. Bowerman(1999) measures the time spent on the possible versus impossible event, indicatingthat infants of only months old understand that moving objects must follow acontinuous trajectory rather than moving randomly. Further studies reveal that infantsof 7-15 months can discover changes in Manner and Path and categorize events on thebasis of Manner and Path (Pulverman & Golinkoff, as cited in Ji, 2014).
...........................

2.2 Previous Domestic Studies into Spatial Language

Shen (1985) compares English-Chinese spatial expressions with Figure andGround, even prepositions. The result shows that English Figures appear ahead ofGround, while Chinese Ground is often used before Figure. Furthermore, Chineseprepositions are always presented as auxiliary words before verbs, whereas Englishverbs are regarded as adverbs, followed by prepositions. Xing (1996) has studied theinterior structure of “zai…li” and “zai…zhong”, showing the differences andsimilarities between these two locative words.
.......................
3. Methodology...........................21
3.1 Research Questions...................21
3.2 Subjects......................22
4. Results and Discussion.....................27
4.2 The Semantic Pattern of “+V–S”............29
4.3 The Semantic Pattern of “–V+S”................31
5. Conclusions.......................43

4. Results and Discussion

4.1 The Semantic Pattern of “+V+S”
Table 1 shows the general condition of “+V+S” used by English and Chinesespeakers in the static description task of “IN” and “ON”. A first ANOVA performedon the semantic pattern of “+V+S” shows a significant effect of language groupF(1,62)=88.51; P<0.001. Chinese speakers tend to use more specific verbs plusspecific satellites than English subjects in all static location scenes. The overallpattern can also be observed within each item type (Typical-“ON”, EXP-“ON”,Typical-“IN” and EXP-“IN”). The location type main effect is also significantF(1,62)=30.593; P<0.001, as well as the paired comparison (P<0.001) underBonferroni correction. Speakers seem to obtain generally higher score in ON relatedscenes rather than IN. The two-way interaction of typicality and language F(1, 62)=19.858; P<0.001 is significant as shown in Fig. 1a. Paired comparison finds that Chinesespeakers achieve higher scores in typical scenes, scoring much lower score thanEnglish speakers when the scene is atypical. This indicates that Chinese speakersshow a strong preference to specific verbs plus specific satellites to encode typicalscenes with different spatial relationships, including both IN and ON. Furthermore,Fig.1b shows the significance of the two-way interaction between location type andtypicality F(1, 62)=24.484; P＜0.001. Paired comparison finds that ON items elicitmore specific verbs plus specific satellites than items encoding IN when the scene isatypical, whereas these two types do not differ from each other in the typical scene.And three-way interaction of location type, typicality and language F(1, 62)=2.581; P＞0.05 is not significant.
.........................

5. Conclusions
The study shows some significant differences in the way that English and Chinesespeakers describe static locations encoding ON and IN. From the above analysis, weinfer that language group (English VS Chinese), location type (“ON” and “IN”scenes), and Typicality (Typical VS Exploratory ) as well play an important role in thesemantic patterns chosen to encode the static location scenes. The effect of languagegroup is most significant, especially in the semantic pattern of “–V+S”. Englishspeakers tend to use more general verbs and more specific satellites ( namely“–V+S”) , while Chinese speakers rely much more heavily on specific verbs ( namely“+V+S” and “+V–S” ) in all static location types. However, English specific verbs arenonetheless more frequent with non-prototypical situations, and generally focus onPOS, VJ and DIS. As for Chinese speakers, they clearly differentiate prototypicalcontainment situations (preposition “里”and ‘in’ with specific verbs) fromnon-prototypical ones in the static location description task. Chinese speakers relyheavily on specific verbs (generally with specific satellites) to encode all static scenes,except for the EXP-IN, and their verbs frequently encode VM and SMM. In summary,the language differences between English and Chinese are observed in the currentstudy, showing that language types, typicality and even location types stronglyinfluence how speakers describe static locations. These differences help to prove thatlanguage-specific factors strongly influence how native speakers talk about staticlocation scenes. This is consistent with current theories put forward by Talmy.
reference(omitted)