Skip to content
BY 4.0 license Open Access Published by De Gruyter Mouton December 8, 2023

Typological shift of Mandarin Chinese in terms of motion verb lexicalization pattern

  • Liu Linjun EMAIL logo and He Yingxin
From the journal Cognitive Linguistics

Abstract

Given the controversies over Mandarin Chinese in terms of Talmy’s bipartite language typology, this paper presents an exhaustive study of Chinese motion verbs collected from two authoritative dictionaries, namely, The Ancient Chinese Dictionary (2nd Edition) and The Contemporary Chinese Dictionary (7th Edition). An analysis of 662 motion verbs in ancient Chinese and 693 motion verbs in modern Chinese indicates that Mandarin Chinese has undergone a typological shift from verb-framed to satellite-framed as far as the lexicalization pattern is concerned. The typological shift seems to have been driven by two forces, the decline of monosyllabic motion verbs and the upsurge of disyllabic motion verbs, which, upon second thoughts, can be boiled down to a single but predominant process of disyllabification in Chinese, whereby two (former) roots that bear a wide range of syntactic relations are lexicalized into a disyllabic word. Thus, we see an intriguing case of how phonology and morphosyntax interact to impact the typological properties of a language.

1 Introduction

The perception of motion is an integral part of human cognition, and a motion event consists of four internal meaning components, Figure, Ground, Motion, and Path, together with an external Co-event, usually Manner or Cause (Talmy 2000b). Motion events are expressed in all languages, but languages differ in their coding of Motion. Based on the lexicalization pattern (i.e., the systematic and regular relations between inner semantic components and surface linguistic forms) of the Motion event, world languages fall into two categories: verb-framed language and satellite-framed language (Talmy 1975, 1985, 2000b). In verb-framed languages, Path, as the schematic core of a Motion event, is encoded in the main verb or verb root, and accordingly the lexicalization pattern of motion verbs is often represented as [Motion + Path]. Conversely, in satellite-framed languages, Path is expressed by satellite components such as English verb particles and Chinese verb complements, and the lexicalization pattern is formulated as [Motion + Manner/Cause].

In his Toward a Cognitive Semantics, Talmy categorizes ancient Chinese as a verb-framed language while viewing modern Chinese as a strongly satellite-framed language like English (2000b: 46), which has since given rise to a number of controversies over the typological status of Chinese. So far, four different typological categories have been assigned to Chinese, namely satellite-framed, verb-framed, parallel-framed, and equipollently-framed. Though the reason is multifold, the differences in research scope and methodology have played a central role. The research material seems to be a mixture of lexical items and syntactic patterns. While it is impossible to examine all syntactic patterns, research on motion verbs also remains partial, with the focus mostly on some high-frequency monosyllabic items and compounds of verb-complement configuration. Given this research gap and the fact that Chinese has undergone a process of disyllabification, the present study represents an exhaustive analysis of Chinese motion verbs collected in two authoritative dictionaries, The Ancient Chinese Dictionary (《古代汉语词典》) (2nd edition) (Commercial Press, 2014) and The Contemporary Chinese Dictionary (《现代汉语词典》) (7th edition) (Commercial Press, 2016), to reveal how they have evolved over time in Mandarin Chinese and consequently impacted the typological status of the language. It should also be noted that The Ancient Chinese Dictionary is based on material predominantly from the Pre-Qin and Han dynasties, when disyllabic verb compounds of modifier + head configuration began to upsurge (Dong 2011: 145), which may be a direct contributor to the typological shift of the Chinese language.

2 Controversies over Chinese typology

When ancient Chinese is concerned, most studies concur in general with Talmy (2000b); that is, ancient Chinese displays more features of a verb-framed language (see Li 1993, 1997; Peyraube 2006; Shen 2003; Xu 2006; among other). It is, nonetheless, also proposed that ancient Chinese can also be characterized as equipollently-framed, because path and manner are expressed by equivalent grammatical forms, i.e., the single path verb construction and the serial verb construction. But when frequency of occurrence is examined, the former exceeds the latter, which attests to the validity of the verb-framed categorization of ancient Chinese (Shi 2011, 2014).

The controversies, in any case, are more over modern Chinese. The first standpoint maintains that modern Chinese is more a satellite-framed language, with which most scholars agree (Chen and Zhao 2008; Deng 2020; Liu 2013; Matsumoto 2003; Peyraube 2006; Shen 2003; Talmy 2000b, 2009; etc.). The argument is mainly based on the fact that in the complex verb-directional construction, the directional verb complement, which is a closed class, serves as the satellite to encode Path, while the Co-event, usually the Manner or Cause, is conflated in the main verb. It should be noted, however, that this group of researchers seem to have divided ideas on how typical Chinese is as a satellite-framed language. For instance, Shen (2003) argues that it can be hard to distinguish the core element from the satellite element in Chinese because it lacks morphological markers, which renders modern Chinese not so typical as a satellite-framed language.

The methods used, however, vary with the research. The verb-complement construction (the directional as a sub-type of it), given its prominent status, seems to be the most examined (see Chen and Zhao 2008; Matsumoto 2003; Peyraube 2006; Shen 2003; among others), which is complemented with elicited data from picture stories (such as Deng 2020; Liu 2013) of a scale similar to Slobin (2004).

The second standpoint categorizes modern Chinese as a verb-framed language based on observations of action-result verb compounds, a sub-type of the verb-complement construction (Tai 2003), lexicalization patterns of main verbs from self-built corpora (Han 2011; Li 2017; among others), as well as comparison of modern literary works with their English translations (e.g., Wu 2015). It is very interesting to note that Tai (2003) and Shen (2003), examining the same language phenomenon, action-result verb compounds, have nonetheless ended up with conflicting conclusions, because they have diametrically opposite interpretations with regard to the head of the construction under examination.

The third view is represented by Slobin and Hoiting (1994), Zlatev and Yangklang (2004), Slobin (2004, 2006, Chen (2005), Chen and Guo (2009), and Ameka and Essegbey (2013), who argue that Chinese is an equipollently-framed language; that is, it has the features of both satellite-framed and verb-framed languages (Slobin 2004). The main reason for this typology is that Chinese is a serial verb language where Motion and Path of a Motion event are represented by separate verbs in one monoclause. Since the consecutive verbs are not morphologically marked for tense, aspect, polarity, or modality, it is difficult to determine which one bears more morphosyntactic weight. This observation actually explains the apparently contradictory conclusions between Shen (2003) and Tai (2003) discussed above.

Parallel-framed language is a new typology put forward for modern Chinese based on the observation that the core element Path in Chinese Motion events is far more complicated than in other satellite-framed languages such as English, in that it can be expressed by both directional verbs and directional verb complements in modern Chinese. In addition, the Chinese directional verb can be used independently or in combination with another directional verb or another Path verb (cf. Luo 2008; Zhong and Huang 2016).

Zhong and Huang (2016) have incorporated in their analysis the dimension of the dynamic relationship between the agentivity and the semantic core of Chinese verb-complement constructions. They argue that modern Chinese tends to adopt verb-framed language patterns in non-agentive and self-agentive Motion events or change-of-state events, whereas it prefers satellite-framed language patterns when expressing agentive events. This addition of agentivity is very interesting, in that it can explain, if only partially, the categorization of Chinese as verb-framed in Li (2017). The corpus data consist mainly of news reports on three subject matters, conflagration, coalmine mishap and volcanic eruption, none of which is highly agentive, and even the agentive events are no more than the agents’ response to natural or man-made calamities, rather than their purely volitional choice.

In short, the typology of modern Chinese remains highly contended. As suggested earlier on, this can be a liable outcome of choice of research methodology. It is not difficult to see from the highly sketchy literature review above that scholars have approached the issue primarily from two perspectives: typical morpho-syntactic configurations for Motion event representation and natural language use. At the core of the former are the Chinese verb-complement construction and the serial verb construction, which can be further boiled down to the relative weight of the component parts of the construction. If the motion verb is taken as the head and the directional or resultative complement as satellite, Chinese is a satellite-framed language; if the verb complement is taken as the head, Chinese belongs to the verb-framed language type; if the verb-complement functions as a full verb and has the same status as the motion verb, then Chinese is an equipollently-framed language; and if the head of the construction is unfixed and context-dependent, Chinese is a parallel-framed language. Ultimately, the confusion can be attributed to the fact that Chinese is morphologically impoverished language, which gives rise to a great deal of subjectivity in syntactico-semantic interpretation.

The other perspective is natural language use, which includes classic literary works, their translations, elicited data, as well as self-built corpora. The size of research material varies with the research, ranging from one hundred to ten thousand clauses. But as Li (2017) has acknowledged, language typology may be affected by different types of Motion events, which means that the sample chosen can sway the final conclusion.

Among the controversies, there is one thing that borders on a consensus; that is, Chinese has undergone a typological shift in its evolution from ancient Chinese to modern Chinese. But how come the shift remains to be further explored. Since the verb specifically refers to the verb root in Talmy’s bipartite language typology (2000b: 27), we have decided on an exhaustive approach to dictionary entries that encode motion events in Chinese in two different periods of time, because they are verb roots or the closest to verb roots given the fact that they have completed the evolution process from syntactic configurations to lexical items.

This decision is also prompted by the disyllabification process that Chinese has undergone. It is well-known that today’s morphology is yesterday’s syntax (cf. Givon 1971). Diachronically the earliest Chinese compound words have originated from syntax (Dong 2011: 25), and as a matter of fact, the process of disyllabification has changed not only the overall pattern of Chinese lexicon (disyllabic words becoming predominant in number), but also the syntax of the Chinese language. Therefore, the typological shift discussed above may well be one of its syntactic consequences.

Given the complexity of motion event coding in Mandarin Chinese, focus on established wordhood can bring us to the core of the issue – Talmy’s typology is based primarily on the lexicalization patterns of Motion events in any case. Dictionary entry definitions can also help resolve some of the confusions in weight assignment to the component parts of the Motion event. It is anticipated that the endeavor can present a panoramic view of Chinese motion verbs at two sections of their evolution in terms of Talmy’s lexicalization pattern theory and in turn unveil the underlying motivations for the typological shift.

3 Methodology

3.1 Theoretical preliminaries

According to Talmy (2000b), a Motion event refers to the situation which includes motion or the continuation of a stationary location. Basically, an event of Motion contains one animate or inanimate entity moving or located with respect to another entity.

A Motion event can be analyzed into four internal components, namely, Figure, Ground, Motion, and Path. The Figure, which is often expressed by a subject noun, refers to a physical object that is moving or located. The Ground, which is relatively stationary, functions as a reference object with respect to the Figure’s path or site. The third internal component Motion is an abstract concept. Two states of Motion are structurally distinguished in human language: one is the occurrence of motion (represented by MOVE), and the other is the locatedness with a zero-path (represented by BELOC). To be more specific, the component of Motion refers to the presence (MOVE) or absence (BELOC) of translational motion in which the location of the Figure changes in a certain time period. In other words, not all kinds of motion that the Figure can exhibit are involved in the framework of Talmy’s motion event theory. The fourth internal component, Path, is the path followed or the site occupied by the Figure with reference to the Ground.

Besides the above four internal components, an event of Motion is habitually associated with an external Co-event, most often bearing the relation of Manner or Cause to the main Motion event. In the Manner relation, the Co-event normally conceptualized as an additional activity that the Figure exhibits often co-occurs with the main motion, supporting the Motion event by elaborating the main motion. In the Cause relation, the Co-event can precede or co-occur with the framing event as an essential prerequisite or motivation for the main motion. In brief, Manner refers to how the motion is carried out, and Cause denotes the external cause of the motion.

In Talmy’s lexicalization pattern theory, lexicalization is involved where a specific semantic component is regularly associated with a particular morpheme. The process also addresses the case in which a group of semantic elements that bear particular relations to each other is connected with a morpheme, forming the whole meaning of the morpheme (2000b: 24). In a word, lexicalization is the systematic and regular relations between meaning and form. However, meaning components and surface forms are largely not in a one-to-one relationship. A single meaning component can be represented by a combination of surface linguistic forms and a single surface form can also express a combination of meaning components (Talmy 2000b: 21). The process of lexicalization is actually the mapping process between the meaning domain and the language surface domain. For the convenience of comparing languages with very different word structures, verb here specifically refers to the verb root[1] (Talmy 2000b: 27).

In satellite-framed languages, Path is characteristically mapped onto the satellite while the Co-event, usually as Manner or Cause, is conflated into the main verb. In verb-framed languages, the semantic component Path is characteristically mapped onto the verb root while the Co-event, usually as Manner or Cause, is expressed in a separate lexical item from the main verb like gerunds. English is a most representative of the former type and Spanish is typical of the latter. The lexicalization pattern of motion verbs in a satellite-framed language can be formulated as [Motion + Manner/Cause] and that of motion verbs in a verb-framed language as [Motion + Path], which can be schematically represented in Figures 1 and 2 respectively, the joined lines indicating the conflation of the semantic components.

Figure 1: 
Satellite-framed lexicalization pattern.
Figure 1:

Satellite-framed lexicalization pattern.

Figure 2: 
Verb-framed lexicalization pattern.
Figure 2:

Verb-framed lexicalization pattern.

3.2 Practical considerations

3.2.1 Research questions

The general goal of this study is to further investigate the language typology of ancient Chinese and modern Chinese and then account for the very likely typological shift. Specifically, this study tries to answer the following three research questions:

  1. Which typological category do ancient Chinese and modern Chinese belong to respectively, satellite-framed or verb-framed?

  2. How does Chinese typological category shift from ancient Chinese to modern Chinese?

  3. How has the process of disyllabification that Chinese has undergone in its evolution from ancient Chinese to modern Chinese affected the typology of Chinese?

3.2.2 Research material

As already mentioned, the research material of the present study is the motion verbs from two authoritative dictionaries, namely, The Ancient Chinese Dictionary (2nd Edition) and The Contemporary Chinese Dictionary (7th Edition), both of which are the latest editions published by the Commercial Press, an enduring publishing house in China.

The Ancient Chinese Dictionary, as the title suggests, is the source of motion verbs in ancient Chinese. The entries in the dictionary, 14,200 monosyllabic and 28,000 disyllabic, are primarily selected from representative ancient Chinese books and documents, such as The Analects, Mencius, A History of the Later Han Dynasty, etc., predominantly from the Pre-Qin and Han dynasties, which coincide with the beginning of the upsurge of disyllabic verb compounds of modifier + head configuration (Dong 2011: 145). If the compound happens to be a motion verb, it satisfies the [Motion + Manner] lexicalization pattern and adds to this specific lexical inventory, and ultimately skews the language towards a satellite-framed language. The Contemporary Chinese Dictionary serves as the source of motion verbs in modern Chinese, 56,000 dictionary entries making up the mainstream vocabulary of Mandarin Chinese.

It should be acknowledged, nonetheless, that dictionary resources may represent varieties other than spontaneous language use, which can be particularly true of The Ancient Chinese Dictionary, though some of its resources, such as The Analects, are claimed to be close to the spoken variety. Another limitation of our decision on the present research material is that the frequency dimension is largely neglected though we are of the view that it takes a threshold frequency for a linguistic configuration to be lexicalized and listed as a dictionary entry. As a comprise, we try to incorporate a frequency examination of a small but important subcategory of motion verbs, the so-called directional verbs (see Section 5.2.3).

3.2.3 Research procedure

Motion verbs are collected in four steps. Step one is distinguishing words from bound morphemes. This study aims to investigate words, so bound morphemes need to be excluded. The Contemporary Chinese Dictionary has made the distinction by annotating word-class information for words. The Ancient Chinese Dictionary lacks such information, but each entry is followed by at least one example sentence, which means that the entry can stand independently as a word. So we are in a position to treat entries in The Ancient Chinese Dictionary as words.

The second step is identifying verbs. In Talmy (2000b: 27), where the focus is mainly on verb roots across various languages, the Chinese verb roots are supposed to be able to stand alone as entire words, which indicates that we can directly treat verb entries in the Chinese dictionaries as verb roots and collect them as research material. The Contemporary Chinese Dictionary has labeled its entries with word-class information; when it comes to The Ancient Chinese Dictionary, we make reference to entry definitions and the syntactic slots in which they occur in the example sentences to identify verbs.

The third step is identifying motion verbs. In the framework of Talmy’s motion event theory, the semantic component of Motion does not include all kinds of motions that a Figure could exhibit. Instead, it refers specifically to the occurrence (MOVE) or nonoccurrence (BELOC) of translational motion in which the location of the Figure changes with respect to the Ground in a certain time. Particularly, it excludes self-contained motions such as dilation, rotation and oscillation, in which the Figure does not change its location with respect to the Ground (Talmy 2000b: 25–26). Therefore, Chinese verbs like 转 (zhuàn, ‘to rotate’) and 膨胀 (péngzhàng, ‘to dilate’) are excluded from our study. Also excluded are those motion verbs that involve the movement of a particular body part while the whole body does not change its location with respect to the Ground, such as 顉 (qīn, ‘to nod’), 耸肩 (sǒngjiān, ‘to shrug’), etc., because the Figure is usually conceptualized as an indivisible zero-dimensional unit, even though its internal components form multidimensional relationships (Talmy 2000a: 183).

Then the lexicalization pattern of each identified motion verb is analyzed and subcategorized. When a verb is polysemous and shows more than one lexicalization pattern, we label each with a subscript number. For instance, in modern Chinese, 跑 (pǎo) is of the [Motion + Manner] lexicalization pattern when expressing the meaning ‘to run’ and ‘to walk’, but of [Motion + Path] when used to mean ‘to leave’. So we end up with 跑1 and 跑2 to acknowledge the two lexicalization patterns. The same applies to words that show more than two lexicalization patterns. To capture this feature in our calculation, we assign value “1” to each motion verb, and divide the value by the number of lexicalization patterns that a motion verb exhibits. In the case of 跑, the two patterns 跑1 and 跑2 will each get a value of “0.5”.

To guarantee maximal reliability, we have two coders to label the words. The initial inter-coder agreement is 91.3 %. All discrepancies are then resolved by turning to a third coder before statistics are calculated and sorted out for research purposes.

4 Results and discussion

4.1 Two general reversals

The statistical results obtained from applying the research procedure as delineated above show two general reversals. The first reversal concerns the relative dominance of monosyllabic vis-a-vis disyllabic motion verbs in ancient and modern Chinese. As shown in Table 1, a total of 662 motion verbs are identified in The Ancient Chinese Dictionary, 439 of them being monosyllabic and 223 disyllabic. When it comes to The Contemporary Chinese Dictionary, 693[2] satisfy the criteria of motion verbs, 150 being monosyllabic, 521 disyllabic, and 22 polysyllabic. That is, disyllabic motions verbs have greatly outnumbered monosyllabic motion verbs in modern Chinese.

Table 1:

Motion verbs in terms of syllable number.

Motion verbs Monosyllabic Disyllabic Total
Ancient Chinese 439 223 662
Modern Chinese 150 521 671

The second reversal is seen in the dominant lexicalization pattern of the two inventories of Chinese motion verbs. Table 2 presents the lexicalization patterns in ancient Chinese and Table 3 lists those in modern Chinese. Note that each dictionary entry is assigned a value of “1”. When the entry word expresses more than one pattern of translational motion, it is subject to analysis into different lexicalization patterns, which entails that the value “1” will be divided by the number of lexicalization patterns, thus the decimals in the tables.

Table 2:

Frequency and proportion of lexicalization patterns in ancient Chinese.

Lexicalization pattern Frequency Proportion
[Motion] 6.5 0.98 %
[Motion + Path] 274.82 41.51 %
[Motion + Manner] 211.32 31.92 %
[Motion + Path + Manner] 148.32 22.41 %
[Motion + Path + Cause] 3 0.45 %
[Motion + Manner + Cause] 16 2.42 %
[Motion + Path + Manner + Cause] 2 0.30 %
Total 661.96 99.99 %
  1. Note: The data in the table are rounded to two decimal places, thus the errors.

Table 3:

Frequency and proportion of lexicalization patterns in modern Chinese.

Lexicalization pattern Frequency Proportion
[Motion] 4 0.60 %
[Motion + Path] 167.33 24.94 %
[Motion + Manner] 301.83 44.98 %
[Motion + Path + Manner] 150.83 22.48 %
[Motion + Path + Cause] 7 1.04 %
[Motion + Manner + Cause] 33 4.92 %
[Motion + Path + Manner + Cause] 7 1.04 %
Total 670.99 100 %
  1. Note: The data in the table are rounded to two decimal places, thus the errors.

As can be seen from Tables 2 and 3, there are altogether seven lexicalization patterns exhibited in both ancient and modern Chinese, which indicates that Chinese has a long history of using a richer variety of lexicalization patterns than established in Talmy (2000b), namely, [Motion + Path], [Motion + Manner] and [Motion + Cause]. What is more impressive, however, is that there is a reversal in the relative frequency/proportion of the two most dominant patterns, [Motion + Path] and [Motion + Manner]. In Table 2, [Motion + Path], with a frequency of 274.82, accounts for 41.51 % of all motion verbs, contributing the most to the typological properties of ancient Chinese as a verb-framed language, with [Motion + Manner] coming the second (31.92 %). When Table 3 is concerned, [Motion + Manner] rises to 44.98 % in proportion, the largest for any single pattern, surpassing [Motion + Path] by 20.04 %. If all patterns that include the semantic component [Manner] are calculated, the proportion can rise up to an indeed overwhelming 73.26 % (cf. 49.50 % for Path). The typological consequence of this reversal is that modern Chinese has been greatly pulled towards the end of satellite-frame languages. Given the specific figures of Chinese lexicalization patterns in the two different eras, if there is a near consensus on ancient Chinese as a verb-framed language, it might be safe to conclude that modern Chinese is more a satellite-framed language.

4.2 Tendencies in relation to syllable number

Given the first reversal discussed above, it seems that the process of disyllabification has a great impact on the typological shift from ancient Chinese to modern Chinese. Following the overall frequency and proportion analysis, this section analyzes the two inventories of motion verbs in terms of syllable number.[3] The statistics are represented in Figures 3 and 4.

Figure 3: 
Distribution of monosyllabic and disyllabic motion verbs across lexicalization patterns in ancient Chinese.
Figure 3:

Distribution of monosyllabic and disyllabic motion verbs across lexicalization patterns in ancient Chinese.

Figure 4: 
Distribution of monosyllabic and disyllabic modern Chinese motion verbs across lexicalization patterns.
Figure 4:

Distribution of monosyllabic and disyllabic modern Chinese motion verbs across lexicalization patterns.

From Figure 3, it can be seen that in ancient Chinese, the frequency order of lexicalization patterns of monosyllabic motion verbs is exactly the same as the overall order of ancient Chinese, [Motion + Path] being the largest. But when it comes to disyllabic motion verbs, the order differs from the overall one in that the pattern [Motion + Manner] accounts for the largest (82.5), pushing [Motion + Path] to the second place (78). So it seems that monosyllabic motion verbs contribute more than disyllabic ones to the verb-framed language properties of ancient Chinese.

As far as Figure 4 is concerned, the frequency order of the lexicalization patterns of monosyllabic motion verbs in modern Chinese is roughly the same as the overall order and the top three patterns are [Motion + Manner], [Motion + Path] and [Motion + Path + Manner]. But when it comes to disyllabic motion verbs, the frequency order of lexicalization patterns is somewhat different, the three most frequently occurring patterns being [Motion + Manner], [Motion + Path + Manner] and [Motion + Path]. In both groups of motion verbs, the representative pattern [Motion + Manner] accounts for the largest in proportion, but in sheer number, the disyllabic (261.5) far exceed the monosyllabic (56.33). Therefore, it is the disyllabic motion verbs that have contributed more to the satellited-framed language properties of modern Chinese.

As mentioned earlier on, the semantic components of Path and Manner are directly related to Talmy’s bipartite language typology, so a separate statistical analysis is made on the distribution of Path and Manner in monosyllabic vis-a-vis disyllabic motion verbs in order to examine their influence on the typological category of the Chinese language. The results are listed in Tables 4 and 5, where the bracketed are the numbers of motion verbs in each category.

Table 4:

Distribution of Path and Manner in motion verbs of ancient Chinese.

Monosyllabic motion verbs (439) Disyllabic motion verbs (223) Difference
Path 65.86 % 62.33 % 3.53 %
Manner 53.33 % 64.35 % −11.02 %
Difference 12.53 % −1.02 % /
Table 5:

Distribution of Path and Manner in motion verbs of modern Chinese.

Monosyllabic motion verbs (150) Disyllabic motion verbs (521) Difference
Path 49.79 % 48.53 % 1.26 %
Manner 64.77 % 76.70 % −11.93 %
Difference −14.98 % −28.17 /

Table 4 shows that the semantic components of Path and Manner exist extensively in both monosyllabic and disyllabic motion verbs of ancient Chinese. The proportion of Path in monosyllabic motion verbs is similar to that of disyllabic motion verbs, and Path is slightly more distributed in monosyllabic motion verbs (by 3.53 %). In contrast, Manner is seen more in disyllabic motion verbs than in monosyllabic motion verbs by 11.02 percentage points. In addition, in disyllabic motion verbs of ancient Chinese, the proportion of Path and Manner is very close (with a slight difference of −1.02 %), whereas in monosyllabic motion verbs, the proportion of Path exceeds that of Manner by 12.53 percentage points.

Table 5 demonstrates the extensive existence of the semantic components Manner and Path in both groups of motion verbs in modern Chinese. But we can see an obvious upsurge in the proportion of Manner and a decline in that of Path. Specifically, the proportion of Path in both groups is under 50 %, which is far lower than that in ancient Chinese (65.86 % and 62.33 % respectively). Moreover, Path is a little more distributed in monosyllabic motion verbs than in disyllabic ones. When it comes to Manner, it is much more distributed in disyllabic motion verbs than in monosyllabic motion verbs. The proportion of Manner in disyllabic motion verbs is 11.93 percentage points higher than that in monosyllabic ones. Additionally, in both monosyllabic and disyllabic motion verbs, the proportion of Manner is much higher than that of Path, 14.98 % for monosyllabic motion verbs and 28.17 % for disyllabic motion verbs.

4.3 Section summary

To sum up, Chinese has indeed undergone a typological shift in its evolution from ancient Chinese to modern Chinese in terms of lexicalization pattern. Ancient Chinese is more towards the end of verb-framed languages, while modern Chinese shows more characteristics of a satellite-framed language. When the lexicalization pattern is examined in relation to syllable number, Path is more distributed in monosyllabic motion verbs, while Manner is more incorporated in disyllabic ones, be it in ancient or modern Chinese, which seems to suggest that monosyllabic motion verbs contribute more to the verb-framed properties of the Chinese language and disyllabic ones enhance its satellite-framed properties.

This is, in fact, not very difficult to explain. As is known, words in ancient Chinese are predominantly monosyllabic. When a monosyllabic word is passed down to modern Chinese, it can be taken as a legacy of ancient Chinese, thus more conservative in its lexicalization pattern. We can thus infer that the typological shift in Chinese can be attributed, if only in part, to the process of disyllabification, a process at the interface between morphology and syntax predominant from the Pre-Qin and Han dynasties, which coincides with the beginning of the upsurge of disyllabic verb compounds of modifier + head configuration (Dong 2011: 145).

The process may also be liable for the richer variety of lexicalization patterns (seven of them) than what has been proposed by Talmy (2000b) (three mainly), though some of them account for a very small proportion and have little influence on the typological properties of the Chinese language. They may very well be the outcome of syntactic configurations condensed into words. The following section investigates more into this while providing a more comprehensive explanation of the observed typological shift.

5 Reasons for the typological shift of Chinese

The reason analysis is to be done in terms of the number of syllables of the motion verbs collected from the two source dictionaries, with the focus on resolving the last research question, namely, “How has the process of disyllabification that Chinese has undergone in its evolution from ancient Chinese to modern Chinese affected the typology of Chinese?”.

5.1 Monosyllabic motion verbs

As has been pointed out in Section 4.2, the proportion of the semantic component Path in monosyllabic motion verbs has dropped by 16.07 % (from 65.86 % to 49.79 %) in the evolution process from ancient Chinese to modern Chinese, which can be a vital reason for the Chinese typological shift. We have therefore investigated all monosyllabic motion verbs in ancient Chinese which contain Path. There are 303 of them, out of a total of 439, which gives this specific lexicalization pattern the single most prominent status in ancient Chinese. Word-by-word examination reveals that 222 have undergone changes that contribute to the decline in this lexicalization pattern and ultimately the typological shift of the Chinese language.

The changes fall into five categories as presented in Table 6. In terms of the number of words involved, the most seen category is for the monosyllabic motion verbs of ancient Chinese to lose their motion verb status in modern Chinese. There are 143 of them, making up 64.41 % of the investigated. These verbs are found to be of various conflation patterns. For instance, 废 (fèi, ‘to collapse/fall’), with the semantic component Path in ancient Chinese, is no longer a motion verb in modern Chinese. Likewise, 北 (běi), meaning ‘to escape/retreat’ in ancient Chinese, has become primarily a noun, referring to the cardinal direction ‘north’; 蠢 (chǔn, ‘to wriggle (as of worms)’, which is of [Motion + Manner] conflation, represents a case of motion verb turning predominantly adjectival in modern Chinese, meaning ‘stupid’. There is also a group of motion verbs with the semantic component Cause that have lost the verb status, such as 逖 (), which used to mean ‘to cause to move faraway’, has evolved into an adjective reserved for formal use to mean ‘distant/remote/faraway’. The loss of the causative use of such words is actually in line with the historical evolution of the Chinese resultative construction: it is prevalent for ancient Chinese to make causative use of verbs and adjectives, which declines with the emergence of the resultative construction (Shen 2003).

Table 6:

Changes in the Path component in ancient Chinese monosyllabic motion.

Verbs
Changes Number Percentage
Loss of motion verb status 143 64.41 %
Loss of free morpheme status 40 18.02 %
Conflation of variant forms 19 8.56 %
Loss of form altogether 10 4.50 %
Drop of [Motion + Path] lexicalization pattern 10 4.50 %
Total 222 99.99 %
  1. Note: The data in the table are rounded to two decimal places, thus the errors.

The second (40 or 18.02 %) come the monosyllabic motion verbs that have evolved into bound morphemes or prepositions in modern Chinese, thus losing their word/verb status. For example, 薄 (), which means ‘to approach’ in ancient Chinese, is found in modern Chinese in a very limited number of di- or multi-syllabic words such as 薄暮 (bómù, ‘toward evening’) and 日薄西山 (rìbóxīshān, ‘The sun is sinking in the west.’) so far as the motion sense is concerned. Then are 19 motion verbs that have been conflated into nine by dropping the variant forms, which reduces the number of monosyllabic motion verbs by 10. Another 10 motion verbs are simply dropped by modern Chinese – they are no longer used at all. The remaining 10, though not big in number, can be intriguing, because they have in any case changed their lexicalization pattern altogether (such as 迈 as illustrated in Table 7) or lost the [Motion + Path] pattern when they embody more than one lexicalization pattern (such as 乘 in Table 7).

Table 7:

Comparison of meanings and lexicalization patterns of monosyllabic motion verbs in ancient and modern Chinese.

Motion verb Ancient Chinese Modern Chinese
Meaning Lexicalization pattern Meaning Lexicalization pattern
mài to move forward [Motion + Path] to stride [Motion + Manner]
chéng
  1. to rise up

  2. to dive or ride

  3. to chase



  1. [Motion + Path]

  2. 2-3[Motion + Manner]





to take a journey by means of vehicles or animals [Motion + Manner]

5.2 Disyllabic motion verbs

Disyllabification is a dominant rhythmic tendency in modern Chinese, making disyllabic words the bulk of Chinese vocabulary. Of disyllabic compounds, verbs, with a percentage of 36.4 %, come second after nouns (51 %). When their internal structures are examined, three stand out, with verb-object compounds taking up 39.7 %, coordinating compounds 27 %, and modifier-head compounds 23.3 % (Yuan and Huang 1998).

As has been shown in Section 4, the process of disyllabification has a vital effect on the Chinese typological shift. The sheer number of disyllabic motion verbs has increased from 223 in ancient Chinese to 521 in modern Chinese, a difference as big as 298, more than doubling the number in ancient Chinese. The change in semantic components is as impressive: the proportion of Path in disyllabic motion verbs has declined by 13.8 % while the proportion of Manner has increased by 12.35 % during the evolution process. This section, therefore, presents an exhaustive analysis of the internal structure of Chinese disyllabic motion verbs.

It should be noted, however, that not all disyllabic words in Chinese have an internal structure. Rather, some are simply of one single morpheme. With reference to Dictionary of Chinese Lianmian Words [4] (Xie 2011), 19 ancient Chinese motion verbs are identified as of this case and nine are thus established in modern Chinese, so altogether 28 are left out of consideration.

The remaining bulk is then compounds of the following configurations in order of instantiation frequency: modifier-head compounds, coordinating compounds, verb-object compounds, verb-complement compounds, and subject-predicate compounds (for specific figures see Table 8).

Table 8:

Frequency counts of different structure types of Chinese disyllabic motion verbs.

Type Ancient Chinese Modern Chinese
Frequency Proportion Frequency Proportion
Coordinating compounds 101.5 49.75 % 166 32.42 %
Modifier-head compounds 53.5 26.23 % 208 40.63 %
Verb-object compounds 41 20.10 % 101 19.73 %
Verb-complement compounds 5 2.45 % 37 7.23 %
Subject-predicate compounds 3 1.47 % 0 /
Total 204 100 % 512a 100.01 %
  1. Note: The data in the table are rounded to two decimal places, thus the errors. aNine lianmian words have been deducted from the total of 521 disyllabic motion verbs in modern Chinese.

When Table 8 is examined, we can see a considerable decline in the proportion of coordinating compounds, from 49.75 % in ancient Chinese to 32.42 % in modern Chinese. Complementing this decline is an upsurge in the proportion of modifier-head compounds in modern Chinese. Given the features of Chinese grammar, the Manner component is the default candidate for verb modifiers. This evolutionary change then stands out as an important reason for the typological shift of the Chinese language.

In the meantime, it might be interesting to note that the verb-complement structure type makes up only a small fraction of the motion verb inventory, be it in ancient Chinese (2.45 %) or in modern Chinese (7.23 %). But when previous studies are referred to, it constitutes a major concern (see literature review in Section 2). Although earlier endeavors are not confined to lexicalized words, the current research suggests that so far as lexicalization is concerned, the issue of Chinese typology has to be addressed on a much broader scale than this particular pattern. This finding may also partly explain why there are so many controversies over the typological category of modern Chinese. In what follows are detailed discussions of each lexical pattern.

5.2.1 Coordinating compounds

Coordination of monosyllabic words can meet the requirements of disyllabification without drastically altering the basic semantics of the original monosyllables, which explains why it is favored in the process of disyllabification. With regard to the semantic relation between the component roots, coordinating compound words fall into compounds of synonymous roots, antonymous roots, and similar roots, where “similar” means that the component roots are of the same category and somewhat related in semantic meaning (Dong and Yin 2021). The results of our analysis are presented in Table 9.

Table 9:

Number and proportion of three types of Chinese coordinating motion verbs.

Coordinating compound motion verbs Ancient Chinese Modern Chinese
Number Proportion Number Proportion
Of synonymous roots 54 53.20 % 101 60.84 %
Of similar roots 32.5 32.02 % 56 33.73 %
Of antonymous roots 15 14.78 % 9 5.42 %
Total 101.5 100 % 166 99.99 %
  1. Note: The data in the table are rounded to two decimal places, thus the errors.

According to Table 9, in both ancient and modern Chinese, coordinating motion verbs of synonymous roots make up the largest category, accounting for more than half of their respective total, which gives this combination the dominant status. Coordinating motion verbs of similar roots come second and those of antonymous roots are the least seen. In what follows, the three types are discussed one by one.

Coordinating motion verbs of synonymous roots can be illustrated in (1).

(1)
a. 陨坠 (yǔnzhuì, ‘to fall’) (ancient Chinese)
b. 飞翔 (fēixiáng, ‘to fly’) (modern Chinese)

The ancient Chinese motion verb 陨坠 consists of two verbal roots 陨 and 坠, both meaning “to fall”. They also contain the same semantic components, namely, Motion and Path. Similarly, in modern Chinese, the disyllabic motion verb 飞翔, whose lexicalization pattern is [Motion + Manner], is composed of two synonymous verbal roots 飞 and 翔, both of the same semantic components Motion and Manner. It can be seen that the two synonymous roots of the same semantic components are simply juxtaposed to form disyllabic motion verbs.

When it comes to the lexicalization pattern of such motion verbs, the most dominant pattern in ancient Chinese, as demonstrated in Table 10, is [Motion + Path] with the number of 21, accounting for 38.89 %, while in modern Chinese, the most dominant is [Motion + Manner], 46 in number or 45.10 % in proportion. This is in line with the overall typological shift from ancient Chinese to modern Chinese. There is, however, something that seems to stray from the general shifting pattern; that is, verbs of [Motion + Path] have also increased in number as well as in proportion, from 21 (38.89 %) in ancient Chinese to 44 (43.14 %) in modern Chinese. A closer look, in any case, can switch off this alarm. The biggest increase is seen in the lexicalization pattern [Motion + Manner], by 9.91 %, which can dwarf that of [Motion + Path] (4.25 %). On top of that, there is the general trend of disyllabification, which can stand as good reason for the increase, especially given the slight advantage of [Motion + Path] over other lexicalization patterns in ancient Chinese and the coordinating characteristic of this type of motion compounds.

Table 10:

Chinese coordinating motion verbs of synonymous roots.

Lexicalization pattern Ancient Chinese Modern Chinese
[Motion] / 2 (1.96 %)
[Motion + Path] 21 (38.89 %) 44 (43.14 %)
[Motion + Manner] 19 (35.19 %) 46 (45.10 %)
[Motion + Path + Manner] 14 (25.93 %) 10 (9.80 %)
Total 54 (100 %) 102 (100 %)

The results of coordinating motion verbs of antonymous roots are listed in Table 11.

Table 11:

Chinese coordinating motion verbs of antonymous roots.

Lexicalization pattern Ancient Chinese Modern Chinese
[Motion + Path] 13 (86.67 %) 9 (100 %)
[Motion + Manner] 2 (13.33 %) /
Total 15 (100 %) 9 (100 %)

In comparison with coordinating motion verbs of synonymous roots, the number of disyllabic motion verbs with antonymous roots is much smaller, which can be attributed to the relatively small number of monosyllabic free verbal morphemes that are antonymous and the relative difficulty with which antonyms come together to express one meaning. The dominant lexicalization pattern is [Motion + Path] in both ancient and modern Chinese. See (2) for illustration.

(2)
a. 沉浮 (chénfú, ‘to sink and float’) (ancient Chinese)
b. 出入 (chūrù, ‘to go out and come in’) (modern Chinese)

The disyllabic motion verb 沉浮 is composed of two antonymous verbal roots 沉 (‘to sink’) and 浮 (‘to float’), both encoding the semantic components of Motion and Manner. Similarly, the two verbal roots of the modern Chinese motion verb 出入 also share the same semantic components, namely, Motion and Path.

The third category is coordinating motion verbs of similar roots, which can be illustrated by (3).

(3)
跋涉 (báshè, ‘to walk on mountains and across rivers’)

This example is found in ancient Chinese as well as in modern Chinese. Both 跋 and 涉 express the meaning of walking, but emphasize different venues for walking. So they are related in meaning, but it is a relation beyond synonymy and antonymy.

Statistics in Table 12 show that the dominant lexicalization pattern of this compound category in ancient Chinese is [Motion + Path] whereas that in modern Chinese is [Motion + Manner], which also helps the Chinese language to gather momentum to shift from a verb-framed language to a satellite-framed one in terms of lexicalization pattern.

Table 12:

Chinese coordinating motion verbs with similar roots.

Lexicalization pattern Ancient Chinese Modern Chinese
[Motion] 1.5 (4.62 %) 1 (1.79 %)
[Motion + Path] 18 (55.38 %) 16 (28.57 %)
[Motion + Manner] 5 (15.38 %) 28 (50.00 %)
[Motion + Path + Manner] 8 (24.62 %) 10 (17.86 %)
[Motion + Manner + Cause] / 1 (1.79 %)
Total 32.5 (100 %) 56 (100.01 %)
  1. Note: The data in the table are rounded to two decimal places, thus the errors.

To recap, in ancient Chinese, synonymous verbal roots that contain Motion and Path seem easier to be juxtaposed to form disyllabic motion verbs. But in modern Chinese, synonymous verbal roots conflating Motion and Manner enjoy greater ease to be combined to form disyllabic motion verbs. As for the second largest category, coordinating motion verbs with similar roots, the verbal root containing Motion and Path is more likely to combine with another verbal root in ancient Chinese, but in modern Chinese, the verbal root including Motion and Manner is more liable to form a coordinating motion verb with a similar root. Such differences between ancient and modern Chinese have driven the language to change its typological category. When it comes to coordinating motion verbs with antonymous roots, verbal roots encoding Motion and Path are more capable of forming disyllabic motion verbs in both ancient and modern Chinese. But this type is of a fairly small number, thus of very limited impact on the Chinese typology. It should be noted that as the norm coordination does not change the lexicalization pattern of the disyllabic word from its component monosyllables; this process impacts language typology by instantiating more frequently in semantic patterns with Manner than those with Path.

5.2.2 Modifier-head compounds

As the name to this group of compounds indicates, the former root is the modifier of the latter, the structural head. When the head is a verb, the modifier is by default a manner adverbial. This is attested by the two motion verb inventories; that is, the semantic component of Manner is encoded in every modifier-head motion verb, though Path and Cause may also be incorporated. When Table 8 is referred to, there has been a significant increase in motion verbs of modifier-head configuration from ancient Chinese to modern Chinese (by 14.40 percentage points), which makes this group of motion verbs an important driving force in the typological shift of the Chinese language. As a matter of fact, modifier-head compounds stand as the single most important contributor to Chinese typological shift in terms of sheer number. The specific results of lexicalization pattern analysis are presented in Table 13.

Table 13:

Lexicalization pattern analysis of Chinese modifier-head motion verbs.

Lexicalization pattern Ancient Chinese Modern Chinese
[Motion + Manner] 30.5 (57.01 %) 119 (55.61 %)
[Motion + Manner + Path] 23 (42.99 %) 83 (38.79 %)
[Motion + Manner + Cause] / 8 (3.74 %)
[Motion + Manner + Path + Cause] / 4 (1.87 %)
Total 53.5 (100.00 %) 214 (100.01 %)
  1. Note: The data in the table are rounded to two decimal places, thus the errors.

When the modifiers are examined more closely, they seem to fall into different categories. First, they can be nominals functioning as the manner adverbial of the head verbs. In (4), for example, both 蛇 (‘snake’) and 酒 (‘wine’) are nouns, indicating the manner of the action coded in the compound head.

(4)
a. 蛇行 (shéxíng, ‘to move like a snake’) (ancient Chinese)
b. 酒驾 (jiǔjià, ‘to drive after drinking’) (modern Chinese)

The modifier can also be verbs, as illustrated in (5), where 抄 highlights taking a tortuous route as an important attack strategy and 赶 means to do something in a hurried manner.

(5)
a. 抄袭 (chāoxí, ‘to launch a surprise attack by making a detour’) (ancient Chinese)
b. 赶赴 (gǎnfù, ‘to hurriedly go to some place’) (modern Chinese)

The modifier, of course, is also coded in adverbs, the default word class for adverbial modifiers. In (6), 时 in 时迈 is an adverb, meaning ‘punctually’, and 独 in 独行 is of the same word class, meaning ‘alone’.

(6)
a. 时迈 (shímài, ‘to patrol punctually’) (ancient Chinese)
b. 独行 (dúxíng, ‘to walk alone’) (modern Chinese)

When we compare the motion verbs in ancient and modern Chinese, 21 disyllabic motion verbs are found in both lists, and 20 of them have maintained the same word structure over time, leaving 奔走 (bēnzǒu) an exceptional case. In ancient Chinese, it means ‘to run’ and belongs to the coordinating type of synonymous roots, in that both 奔 and 走 mean ‘to run’ and conflate the semantic components of Motion and Manner. However, in modern Chinese, 奔走 has come to mean ‘to run quickly’, and thus belonging to the type of modifier-head compound, with 奔 modifying the translational motion of 走 by giving prominence to the speed. As a matter of fact, 走 originally meant “to run”, and it was later on that it lost the semantic component of “fast” to mean “to walk”. In modern Chinese, it also functions as an ideograph, where the speed meaning is still existent to mark its semantic origin.

Perhaps not too much can be read into this individual case, but when the predominant role of modifier-head compound motion verbs is taken into consideration, it may represent a still ongoing trend. In any case, the slot immediately preceding the head verb is by default for the adverbial, of which manner is a ready candidate. Though in Talmy’s framework Manner is not taken as an internal component of a Motion event, it is nonetheless the case for an action to occur in a certain manner, which makes manner also an intrinsic dimension. This observation is also attested by a more general principle, the principle of iconicity: the manner adverbial is the closest to the head verb of all adjunct elements in Mandarin Chinese. Given this relational advantage, the lexicalization of modifier + head also precedes other syntactic configurations in the process of disyllabification (cf. Dong 2011, p. 156). Furthermore, there is the key mechanism of reanalysis in grammaticalization, which may very well enable a previously coordinating element to be reanalyzed as a modifier.

In short, the semantic component of Manner is included in each and every modifier-head motion verb. Unlike in ancient Chinese, such compound words in modern Chinese have not only witnessed a drastic surge in number and proportion, but also greater diversity in lexicalization pattern, which can be a crucial reason for Chinese to transit from a verb-framed language to a satellite-framed one.

5.2.3 Verb-complement compounds

In verb-complement compounds, the latter root complements the former, usually denoting the result (e.g. 向迩, xiàngěr, ‘to move close’) or direction (e.g. 赶上, gǎnshàng, ‘to catch up’) of an action. As noted in literature review, their interpretation is among the most contested, and one major difficulty concerns the establishment of the core of the verb-complement compound. Some linguists suggest that the root is the core while others think that the complement root is more important according to the principle of end-weight (Li and Cao 2015), a dispute that has given rise to different assignments of the same compound to different lexicalization patterns.

According to the semantic characteristics of the complementary morpheme, verb-complement motion verbs of ancient and modern Chinese are listed in Table 14. The figures show that Path has become the most seen semantic component in this type of verb compounds in modern Chinese, in particular those with directional complements (14 out of a total of 15). This is at variance with the general trend of the typological shift of Chinese, but given the general idea that today’s morphology is yesterday’s syntax, there is not much to be alarmed.

Table 14:

Chinese verb-complement motion verbs.

Complement type Lexicalization pattern Ancient Chinese Modern Chinese
Result [Motion + Path] 1 (20.00 %) 8 (21.62 %)
[Motion + Manner] 3 (60.00 %) 7 (18.92 %)
[Motion + Path + Manner] 1 (20.00 %) 7 (18.92 %)
Direction [Motion + Path] / 14 (37.84 %)
[Motion + Path + Manner] / 1 (2.70 %)
Total 5 (100.00 %) 37 (100.00 %)

First, the sheer number of verb-complement disyllabic motion verbs is far from big, five in ancient Chinese and 37 in modern Chinese. Second, the imbalance is mainly seen in the compounds with directional complements (14 of [Motion + Path] and one of [Motion + Path + Manner]). When the 15 compounds are closely examined, the directional roots make up a closed class of four members only, 来 (lái, ‘to come’), 去 (, ‘to go’), 回 (huí, ‘to return’), and 上 (shàng, ‘(to go) up). As a matter of fact, directional verbs are no more than a small closed word class as a whole. More importantly, these motion verbs can be taken as lexicalized syntactic configurations; they are lacking in ancient Chinese simply because they were yet to be lexicalized. Moreover, lexicalization is a matter of degree and the establishment of wordhood is never an easy job. When such verbs are interpreted as syntactic rather than lexical units, directional roots serve the function of satellites in Talmy’s terminology. Even when the wordhood of the 14 compounds of [Motion + Path] are taken for granted, there seems to be an imbalance between their use as verbs and that as directional complements. The Frequency Table of Word Subcategories in the Modern Chinese Corpus (现代汉语语料库词语分词类频率表)[5] is a source that differentiates the two uses, and the relevant results are presented in Table 15. Overall, the ranking of their complement use is much higher than that of their verb use, implying that they are more frequently used as satellites. Although there are five of them (in bold font) that are more frequently used as verbs, a close examination reveals that none of them ranks high enough to reverse the trend. Then it seems that these motion verbs, though of [Motion + Path] semantic conflation, can also drive the Chinese language towards the end of satellite-framed languages in actual use.

Table 15:

Functionally differentiated ranking and frequency counting of Chinese directional verbs.

Word form Ranking as verb Frequency counts Ranking as complement Frequency counts
出来 (chūlái, ‘to come out’) 4,873 241 149 6,335
出去 (chūqù, ‘to go out’) 1,405 952 2,498 519
过来 (guòlái, ‘to come over’) 5,276 218 858 1,554
过去 (guòqù, ‘to go over’) 5,505 208 755 1,772
回来 (huílái, ‘to come back’) 3,531 351 587 2,211
回去 (huíqù, ‘to go back’) 6,112 183 1,650 807
进来 (jìnlái, ‘to come in’) 2,364 552 3,794 323
进去 (jìnqù, ‘to go in’) 2,173 477 4,409 271
上来 (shànglái, ‘to come up’) 10,836 88 2,008 662
上去 (shàngqù, ‘to go up’) 2,321 560 4,050 297
下来 (xiàlái, ‘to come down’) 925 1,446 809 1,661
下去 (xiàqù, ‘to go down’) 1,224 1,099 1,331 1,001
归来 (guīlái, ‘to come back’) 8,128 128 / /
退回 (tuìhuí, ‘to go back’) 12,718 71 / /
Total 6,574 17,413

5.2.4 Verb-object compounds

In verb-object compound motion verbs, the former root represents the action and the latter signifies the entity that the action commands or involves. Since verbal morphemes and nominal morphemes are the most among free monosyllabic morphemes in Chinese (Yin 1984), abundant material can be fed for the lexicalization of disyllabic verb-object compounds. As shown in Table 16, there is a significant upsurge in the number of verb-object motion verbs from 41 in ancient Chinese to 113 in modern Chinese. In terms of lexicalization pattern, the greatest contribution is again from [Motion + Manner] (11 vs. 56.5). When the two more complicated patterns are also taken into account, the tendency can be even more conspicuous, with an addition of 26 (23 + 3) in modern Chinese vis-a-vis 7 in ancient Chinese.

Table 16:

Lexicalization patterns of Chinese verb-object compounds.

Lexicalization pattern Ancient Chinese Modern Chinese
Number Proportion Number Proportion
[Motion + Path] 23 56.09 % 30.5 26.76 %
[Motion + Manner] 11 26.82 % 57.5 50.44 %
[Motion + Path + Manner] 7 17.09 % 23 20.17 %
[Motion + Manner + Cause] 0 / 3 2.63 %
Total 41 100 % 114 100 %

The reason lies again at the morphology-syntax interface. When the compounds of this category are examined, many of them, if not all, are simply fusions of the predicate verb with its object, some still allowing syntactic operations, such as addition of aspect markers in between. The V-O relation is far from homogeneous. A total of 13 role types of objects has actually been established in modern Chinese[6] and all can occur immediately post the predicate verb. See (7) for illustration of some role types with the verb 遛 (‘to walk/stroll’).

(7)
a. 遛马 (liùmǎ, ‘to walk a horse’)
b. 遛鸟 (liùniǎo, ‘to take a walk with pet birds in a park or a quiet place’)
c. 遛弯儿 (liùwǎnr, ‘to take a walk or go for a stroll’)
d. 遛早儿 (liùzǎor, ‘to take a morning stroll’)

To sum up, in ancient Chinese, path verbs whose lexicalization pattern is [Motion + Path] are more capable of combining with a nominal root to form disyllabic motion verbs. However, in modern Chinese, manner verbs with the lexicalization pattern of [Motion + Manner] are more capable of forming new verb-object motion verbs. This difference between ancient and modern Chinese adds to the momentum of the typological shift of the language.

In reversal to verb-object compound motion verbs are those of the subject-predicate configuration. Only three of them are found in ancient Chinese, as represented in Table 17.

Table 17:

Subject-predicate motion verbs in ancient Chinese.

Lexicalization pattern Ancient Chinese Meaning
[Motion + Path] 日下 rìxià ‘The sun goes down.’
[Motion + Manner] 骁腾 xiāoténg ‘The horse gallops away.’
[Motion + Path + Manner] 鹰扬 yīngyáng ‘The eagle flies high.’

In the three cases, 日 (‘sun’), 骁 (‘horse’) and 鹰 (‘eagle’) are the entities that respectively conduct the motion of 下 (‘to go down’), 腾 (‘to gallop’) and 扬 (‘to fly high’). As shown in the above table, the meaning of the subject-predicate motion verb can be represented by a full sentence, which to a certain extent explains why there are merely three such motion verbs in ancient Chinese only. Unlike object which falls within VP in terms of syntactic analysis, subject lies outside VP, which renders it difficult to fuse with the verb and undergo the process of lexicalization. As a matter of fact, modern Chinese seems to have dropped this lexicalization pattern altogether so far as motion verbs are concerned. Here again we see the interfacing between morphology and syntax.

5.3 Some remaining issues

5.3.1 Cooccurrence of path and manner in one single motion verb

Apart from Talmy’s bipartite typology, there are other language types that have been assigned to the Chinese language, and the lexicalization pattern analysis of Chinese motion verbs has indeed manifested some features of equipollently-framed languages. Specifically, Tables 2 and 3 show that the pattern [Motion + Path + Manner], which characterizes the equipollently-framed language, accounts for 22.41 % in ancient Chinese and 22.48 % in modern Chinese. High as the figures are, they still lag significantly behind those for [Motion + Path] and [Motion + Manner] in both tables, implying that [Motion + Path + Manner], though frequently instantiated over time, is never dominant in the evolution of the language. In any case, it is necessary to account for its persistent existence.

When the monosyllables are concerned, there are 101 (or 87.32 when weighted) motion verbs in ancient Chinese and 25 (22.33 after weighting) in modern Chinese that are of the lexicalization concerned. The sharp decline is driven by multiple reasons, such as loss of motion verb status, conflation of variant forms (see Section 5.1 for details), but we want to focus on two important cognitive bases here for the occurrence of both Path and Manner in one single syllable. The first one is that some motions inherently consist of both semantic components, such as 滴 () and 沥 (), both meaning ‘([of liquid] to drip in small drops’), in that liquid, when dripping, will go downward by default. And it is no wonder that the English explanation ‘drip’ also embodies both semantic features. Some other English examples include ‘roam’ when meaning ‘to walk or travel around an area without any definite aim or direction’, ‘bail’ when meaning ‘to leave a place, especially quickly’, etc. The second concerns motion verbs converted from nouns, which is a prevalent phenomenon in Chinese. For instance, 波 (), originally a noun meaning ‘wave’, comes to express ‘to walk along the river’ as a motion verb. Again, Chinese is not alone in this regard, because the English word ‘beach’ can be a ready counterpart when it is used to mean ‘to come or bring something out of the water and onto the beach’. In both cases, the Path component is metonymically specified by their respective nominal use. Therefore, it may well be a language universal for some motion verbs to consist of both Path and Manner in their lexicalization pattern, monosyllables included.

When it comes to disyllabic motion verbs, Section 5.2 has shown that the two syllables are in different relations, and those of the [Motion + Path + Manner] pattern are no exception. Table 18 illustrates some representative cases.

Table 18:

Morphemic analysis of disyllabic motion verbs of [Motion + Path + Manner].

Motion verb Meaning Manner of compounding Semantics of morpheme 1 Semantics of morpheme 2
滴沥 dīlì ‘(of water) to drip sparsely downward’ coordination of verbal roots Manner + Path Manner + Path
鼓行 gǔxíng ‘to march forward while beating the drum’ modifier + head Manner Path
背驰 bèichí ‘to run in the opposite direction’ modifier + head Path Manner
乘危 chéngwēi ‘to climb high’ verb + complement Manner Path
揭河 jiēhé ‘to pull up one’s clothing and cross the river’ verb + object Manner Path

It can be seen from Table 18 that the motion verbs bearing features of equipollently-framed languages are the outcome of various manners of compounding, when the component morphemes contribute either Path/Manner or both to the disyllabic verb. Since compounding is in the main a process of two monosyllables coming together, disyllabification has thus played a role in generating verbs with the semantic components of both Path and Manner, which also explains to a certain extent the roughly unchanged percentage of this category of motion verbs in modern Chinese irrespective of the sharp drop (from 87.32 to 20.83) of monosyllabic motion verbs of this lexicalization pattern in the evolution process of the language.

All things considered, [Motion + Path + Manner] can be best taken as necessary supplement to the more prevalent lexicalization patterns in Chinese. There is no clearcut division between different language types; what a specific typological category captures is no more than the predominant properties of a language.

5.3.2 A glimpse of motion verb lexicalization in three intervening periods

The research material of the present study represents two ends of a fairly long history of lexical evolution of the Chinese language. To give a touch of dynamicity to this research, we refer to three periodized dictionaries to examine if the trend we have been arguing for is attested in the three time periods, namely the Tang and the Five Dynasties (618–960, henceforth Tang & Five for short), the Song Dynasty (960–1279), and the Yuan Dynasty (1271–1368). The number of the identified motion verbs is presented in Table 19 in terms of their syllable composition. It can be seen that the proportion of disyllables is the biggest for all three periods, though there is a drop in the Yuan Dynasty. And accompanying the drop of disyllables in the Yuan period is an increase in the monosyllables, 47 or 43.93 % of a total of 107. The change may well be the outcome of the dictionary sources of this period shifting more towards the spoken variety (say the Yuan dramas) than those of the two early periods, which indicates that Talmy’s typology is variety sensitive.

Table 19:

Number of motion verbs in three time periods.a

Dynasties Monosyllabic Disyllabic Multisyllabic Total
Tang & Five 47 (33.81 %) 89 (64.03 %) 3 (2.16 %) 139 (100.00 %)
Song 6 (8.11 %) 60 (81.08 %) 8 (10.81 %) 74 (100.00 %)
Yuan 47 (43.93 %) 48 (44.86 %) 12 (11.21 %) 107 (100.00 %)
  1. aOverall, the figures presented in the table are much smaller than those of the two dictionaries we choose as our main sources of data. This has something to do with the rationale of dictionary compilation. The three periodized dictionaries are compiled to capture the predominant lexical features of their respective periods. And they are of a very moderate size, around one-fourth the volume of the other two.

The lexicalization patterns are presented in Table 20, which shows that the general trend is for the Path component to decline and the Manner component to increase over time. Specifically, [Motion + Path] ranks the highest in Tang & Five, but it is taken over by [Motion + Manner] in Song. Though [Motion + Manner] falls to the second place in Yuan, the most prevalent pattern [Motion + Path + Manner], nonetheless, also contains the semantic component Manner. The two bottom lines highlight this overall trend, with “Path total” exceeding “Manner total” in Tang & Five, but exceeded by [Manner total] in the two later periods. The upsurge of Path in the Yuan period can be explained in terms of the fairly large proportion of monosyllabic motion verbs, which are more conservative in the disyllabification process and most probably bear more properties of ancient Chinese. Anyhow, 81.16 (75.85 %), the figure of “Manner total” of Yuan, is prominent enough.

Table 20:

Lexicalization pattern analysis of motion verbs in three time periods.

Lexicalization pattern Tang & Five Song Yuan
[Motion + Path] 52 (37.41 %) 17 (22.97 %) 24.83 (23.21 %)
[Motion + Manner] 46.5 (33.45 %) 29 (39.19 %) 34.83 (32.55 %)
[Motion + Path + Manner] 34 (24.46 %) 22.5 (30.41 %) 41 (38.32 %)
[Motion + Path + Cause] 1 (0.72 %) / 1 (0.93 %)
[Motion + Manner + Cause] 4.5 (3.24 %) 5.5 (7.43 %) 5 (4.67 %)
[Motion + Path + Manner + Cause] 1 (0.72 %) / 0.33 (0.31 %)
Total 139 (100.00 %) 74 (100.00 %) 106.99 (99.99 %)
Path total 88 (63.31 %) 39.5 (53.38 %) 67.16 (62.77 %)
Manner total 86 (61.87 %) 57 (77.03 %) 81.16 (75.85 %)
  1. Note: The data in the table are rounded to two decimal places, thus the errors.

The Path and Manner distribution of the motion verbs located in the three periodized dictionaries is listed in Table 21 in terms of syllable number. In each cell, the figure to the right of the slash is the total of monosyllables/disyllables and that to the left is the number of motion words containing the specific semantic component. It can be seen that in all cases of the disyllables, the distribution of Manner exceeds that of Path; on top of this, the difference increases over time, from 4.50 % in Tang & Five, to 21.67 % in Song and 25.00 % in Yuan. The reverse is true of the monosyllables, except in the case of Song, which may well be a chance error, given the fairly small denominator 6. The results again attest to the conservativeness of monosyllables and the pulling effect of the disyllabification process on the Chinese language towards the satellite-framed end of Talmy’s bipartite language typology.

Table 21:

Path and Manner distribution in monosyllabic vis-à-vis disyllabic motion verbs in three time periods.

Dynasty Monosyllabic motion verbs Disyllabic motion verbs
Path Manner Path Manner
Tang & Five 31.5/47

(67.02 %)
25/47

(53.19 %)
54/89

(60.67 %)
58/89

(65.17 %)
Song 2.5/6

(41.67 %)
4/6

(66.67 %)
32/60

(53.33 %)
45/60

(75.00 %)
Yuan 30.16/47

(64.17 %)
28/47

(59.57 %)
28/48

(58.33 %)
40/48

(83.33 %)

6 Conclusions

An exhaustive analysis of motion verbs as characterized by Talmy (2000b) in The Ancient Chinese Dictionary (2nd Edition) and The Contemporary Chinese Dictionary (7th Edition) shows that Chinese has undergone a typological shift in its evolution from ancient Chinese to modern Chinese. Specifically, ancient Chinese is more towards the end of verb-framed languages, while modern Chinese shows more characteristics of a satellite-framed language in terms of lexicalization pattern.

The typological shift seems to have been driven by two forces, the decline of monosyllabic motion verbs and the upsurge of disyllabic motion verbs, which, upon second thoughts, can be boiled down to a single but predominant process of disyllabification in Chinese lexicalization. In any case, the process should be held, at least partially, responsible for the decline of monosyllabic motion verbs, if only in the sense that it has laid off many monosyllabic motion verbs by coding the meanings in their newly emerged disyllabic lexical counterparts. There must have been intense competitions between the different forms, but when disyllabification is an overwhelming trend, the outcome is not difficult to predict.

Moreover, there are altogether seven lexicalization patterns exhibited in both ancient Chinese and modern Chinese, which indicates that Chinese employs a richer variety of lexicalization patterns than proposed in Talmy (2000b), though some of them account for a small proportion and have a minimal bearing on the typological shift of the language. This fact may be accounted for in terms of a very distinctive property of Chinese motion verbs – the disyllabic ones consist of two (former) roots that bear a wide range of syntactic relations. Thus, we see an intriguing case of how phonology and morphosyntax interact to impact the typological properties of a language.

Data availability statement

The datasets generated during and/or analyzed during the current study are available in the OSF.io repository, https://osf.io/3n9mv/.


Corresponding author: Liu Linjun, Beijing Language and Culture University, Beijing, China, E-mail:

Acknowledgments

Heartfelt gratitude to Professor GAO Yuan for first directing our attention to the possible consequences of the disyllabification process in Mandarin Chinese though not necessarily within the framework of language typology. Thanks are also due to the anonymous reviewers for comments on a previous version of this article.

References

Ameka, K. Felix & James Essegbey. 2013. Serialising languages: Satellite-framed, verb-framed or neither. Ghana Journal of Linguistics 02(01). 19–38.Search in Google Scholar

Chen, Jia & Youbin Zhao. 2008. Yě lùn xiàndàihànyǔ de yùndòngshìjiàn cíhuìhuà yǔyì biānmǎ móshì [On Chinese as a satellite-framed language]. Journal of the University of Electronic Science and Technology (Social Science Edition) 10(06). 51–55+58.Search in Google Scholar

Chen, Liang. 2005. The acquisition and use of motion event expressions in Chinese. Lafayette, LA: University of Louisiana at Lafayette Unpublished doctoral dissertation.Search in Google Scholar

Chen, Liang & Jiansheng Guo. 2009. Motion events in Chinese novels: Evidence for an equipollently-framed language. Journal of Pragmatics 41(09). 1749–1766. https://doi.org/10.1016/j.pragma.2008.10.015.Search in Google Scholar

Deng, Yu. 2020. Hànyǔ zhuàngtài biànhuà shìjiàn de cíhuìhuà lèixíngxué shízhèng yánjiū [An empirical study on the lexicalization typology of state change event in Mandarin Chinese]. Foreign Language Research 43(05). 45–51.Search in Google Scholar

Dong, Xiufang. 2011. Cíhuìhuà: Hànyǔ shuāngyīncí de yǎnshēng hé fāzhǎn [Lexicalization: The origin and evolution of Chinese disyllabic words]. Beijing: The Commercial Press.Search in Google Scholar

Dong, Xiufang & Huixia Yin. 2021. Cóng lèixíngxué shìjiǎo kàn hànyǔ zhōng bìnglièshì fùhécí de tèdiǎn [On the characteristics of Chinese coordinating compounds from the typological perspective]. Journal of Hebei Normal University (Philosophy and Social Science Edition) 44(05). 116–124.Search in Google Scholar

Givon, Talmy. 1971. Historical syntax and synchronic morphology: An archeologist’s fieldtrip. Chicago Linguistic Society 7. 394–415.Search in Google Scholar

Han, Chunlan. 2011. Hànyǔ yùndòngshìjiàn cíhuìhuà móshì lèixíng guīshǔ tànjiū [A study of typology attribution of lexicalization pattern about motion event in Chinese]. Journal of Ocean University of China (Social Science Edition) 24(05). 98–101.Search in Google Scholar

Huang, Borong & Xudong Liao. 2017. Xiàndài hànyǔ [Modern Chinese], Updated 6th edn., vol. 1. Beijing: Higher Education Press.Search in Google Scholar

Li, Fengxiang. 1993. A diachronic study of v-v compounds in Chinese. Buffalo, NY: State University of New York dissertation.Search in Google Scholar

Li, Fengxiang. 1997. Cross-linguistic lexicalization patterns: Diachronic evidence from verb-complement compounds in Chinese. Sprachtypologie und Universalienforschung 50(03). 229–252. https://doi.org/10.1524/stuf.1997.50.3.229.Search in Google Scholar

Li, Fuyin. 2017. Diǎnxíng wèiyí yùndòng shìjiàn biǎozhēng zhōng de lùjìng yàosù [Path component in typical translocational events]. Foreign Language Education 38(04). 1–6.Search in Google Scholar

Li, Liyun & Zhao Cao. 2015. Hànyǔ dòngbǔshì fùhécí shēncéng jiégòu tànxī [Deep structure of Chinese verb-complement compounds]. Journal of Hebei Normal University 38(03). 94–99.Search in Google Scholar

Liu, Yan. 2013. Xiàndài hànyǔ yùndòng shìjiàn biǎodá móshì yánjiū [A study on the expression of motion event in Mandarin Chinese]. Tianjin: Nankai University dissertation.Search in Google Scholar

Luo, Xinghuan. 2008. Yīng-hàn yùndòng shìjiàn cíhuìhuà móshì de lèixíngxué yánjiū [A typological study of lexicalization patterns of English and Chinese motion events]. Foreign Language Education 29(03). 29–33.Search in Google Scholar

Matsumoto, Yo. 2003. Typologies of lexicalization patterns and event integration: Clarifications and reformulations. In Shuji Chiba (ed.), Empirical and theoretical investigations into language: A festschrift for Masaru Kajita, 403–418. Tokyo: Kaitakusha.Search in Google Scholar

Peyraube, Alain. 2006. Motion events in Chinese: A diachronic study of directional complements. In Maya Hickmann & Stephane Robert (eds.), Space in languages: Linguistic systems and cognitive categories, 121–135. Amsterdam: John Benjamins.10.1075/tsl.66.08peySearch in Google Scholar

Shen, Jiaxuan. 2003. Xiàndài hànyǔ “dòngbǔjiégòu” de lèixíngxué kǎochá [The resultative construction in Chinese: A typological perspective]. Chinese Teaching in the World 17(03). 17–23+2.Search in Google Scholar

Shi, Wenlei. 2011. Hànyǔ yùndòng shìjiàn cíhuà lèixíng de lìshí zhuǎnyí [The pattern shift of the motion event integration in Chinese: A typological study based on morpho-syntactic features]. Studies of the Chinese Language 60(06). 483–498+575.Search in Google Scholar

Shi, Wenlei. 2014. Hànyǔ yùndòng shìjiàn cíhuà lèixíng de lìshí zhuǎnyí [Diachrony of motion expressions in Chinese]. Beijing: The Commercial Press.Search in Google Scholar

Slobin, Dan I. 2004. The many ways to search for a frog: Linguistic typology and the expression of motion events. In Sven Strömqvist & Ludo Verhoeven (eds.), Relating events in narrative: Typological and contextual perspectives, 219–257. Mahwah, NJ: Lwarence Erlbaum Associates.Search in Google Scholar

Slobin, Dan I. 2006. What makes manner of motion salient? Explorations in linguistic typology, discourse, and cognition. In Maya Hickmann & Stephane Robert (eds.), Space in languages and cognitive categories, 59–81. Philadelphia: John Benjamins.10.1075/tsl.66.05sloSearch in Google Scholar

Slobin, Dan I. & Nini Hoiting. 1994. Reference to movement in spoken and signed languages: Typological considerations. In Proceedings of the 20th Annual Meeting of the BLS, 487–505.10.3765/bls.v20i1.1466Search in Google Scholar

Tai, James. 2003. Cognitive relativism: Resultative construction in Chinese. Language and Linguistics 04(02). 301–316.Search in Google Scholar

Talmy, Leonard. 1975. Semantics and syntax of motion. In John Kimball (ed.), Syntax and semantics: Vol. IV, 181–238. NY: Academic Press.10.1163/9789004368828_008Search in Google Scholar

Talmy, Leonard. 1985. Lexicalization patterns: Semantic structure in lexical forms. In Timothy Shopen (ed.), Language typology and syntactic description Vol. III: Grammatical categories and the lexicon, 57–149. Cambridge: Cambridge University Press.Search in Google Scholar

Talmy, Leonard. 2000a. Toward a cognitive semantics: Vol. I: Concept structuring systems. Cambridge: MIT Press.10.7551/mitpress/6847.001.0001Search in Google Scholar

Talmy, Leonard. 2000b. Toward a cognitive semantics: Vol. II: Typology and process in concept structuring. Cambridge: MIT Press.10.7551/mitpress/6848.001.0001Search in Google Scholar

Talmy, Leonard. 2009. Main verb properties and equipollent framing. In Jiansheng Guo, Elena Lieven, Nancy Budwig, Susan Ervin Tripp, Keiko Nakamura & Seyda Özçalışkan (eds.), Crosslinguistic approaches to the psychology of language: Research in the tradition of Dan Isaac Slobin, 389–402. Mahwah, NJ: Lawrence Erlbaum Associates.Search in Google Scholar

Wu, Jianwei. 2015. Yīnghàn yùndòng shìjiàn de kuàngjià jiégòu bǐjiào yánjiū [Reviewing the typological categorization of motion events in Chinese and English]. Journal of East China University of Science and Technology (Social Science Edition) 30(05). 103–109+116.Search in Google Scholar

Xie, Jifeng. 2011. Hànyǔ liánmiáncí cídiǎn [Dictionary of Chinese lianmian words]. Beijing: Foreign Language Teaching and Research Press.Search in Google Scholar

Xu, Dan. 2006. Typological change in Chinese syntax. NY: Oxford University Press.10.1093/acprof:oso/9780199297566.001.0001Search in Google Scholar

Yin, Binyong. 1984. Hànyǔ yǔsù de dìngliàng yánjiū [A quantitative study of Chinese morphemes]. Studies of the Chinese Language 33(03). 9–17.Search in Google Scholar

Yuan, Chunfa & Changning Huang. 1998. Jīyú yǔsù shùjùkù de hànyǔ yǔsù jí gòucí yánjiū [A data bank based research on Chinese morphemes and word formation]. Chinese Teaching in the World 44(02). 7–12.Search in Google Scholar

Zhang, Bin. 2010. Xiàndài hànyǔ miáoxiě yǔfǎ [A descriptive grammar of Mandarin Chinese]. Beijing: The Commercial Press.Search in Google Scholar

Zhong, Shuneng & Ruifang Huang. 2016. Hànyǔ dòngbǔjiégòu lèixíngxué de rènzhī yánjiū [Research on verb-complement structure typology from the perspective of cognitive linguistics]. Journal of Foreign Languages 39(03). 20–30.Search in Google Scholar

Zlatev, Jordan & Peng Yangklang. 2004. A third way to travel: The place of Thai in motion-event typology. In Sven Strömqvist & Ludo Verhoeven (eds.), Relating events in narrative: Typological and contextual perspectives, 159–190. Hillsdale, NJ: Lawrence Erlbaum.Search in Google Scholar

Received: 2023-01-07
Accepted: 2023-11-04
Published Online: 2023-12-08
Published in Print: 2024-02-26

© 2023 the author(s), published by De Gruyter, Berlin/Boston

This work is licensed under the Creative Commons Attribution 4.0 International License.

Downloaded on 25.5.2024 from https://www.degruyter.com/document/doi/10.1515/cog-2022-0106/html
Scroll to top button