The present study is mainly concerned with the nature of the demonstrative shì 是 as the historical source of the copula shì. Tracing the development of shì in bronze inscriptions from the 11th to 3rd centuries bce and comparing the patterns of change against two purportedly contemporary texts, Shī jīng 詩經 and Shàng shū 尚書, it argues that shì is not a typical demonstrative in origin because it only occurs pronominally and is strongly associated with the structural expression of focus in its early stage of development. On the basis of these observations, shì is analyzed as a special demonstrative that combines anaphoric and focalizing force to highlight and contrast a constituent that has been introduced into the discourse. It is then compared with the expressions wéi zhī 唯之and huì zhī 惠之 with the same function in the Shang oracle-bone inscriptions (13th–11th centuries bce). The investigation of whether shì is derived from wéi zhī only finds evidence indicating a morphological relationship with zhī, while the etymological link with wéi remains speculative. In conclusion, this study claims that the focalizing force of shì is important in answering the question why it is the only demonstrative evolving into a copula. The observed development of shì confirms that the origination of copulas is often related to the morphosyntactic expression of features of information structure.
* An earlier version of this paper was presented at the first lfk Society Young Scholars Symposium on August 11th–13th, 2013, at the University of Washington. I am grateful to three anonymous reviewers for their critical and insightful comments and suggestions on earlier versions of this paper. However, any errors and deficiencies remain mine alone.
Wang (1937) claims that there was no genuine copula in the Chinese language until the emergence of the copula shì/*deʔ1 是 around the 5th century—a dating he later modified to the 1st century (Wang 1958)—evolving from the demonstrative shì. Cross-linguistically, the demonstrative-to-copula pathway is commonly attested (Heine & Kuteva 2002: 108f.). Thus, it is not surprising that the dominant view concerning the historical source of the copula shì largely concurs with Wang's proposal, although some later studies, especially those based on excavated materials (Qiu 1979; Tang 1991), argue that the copular use of shì starts from the Warring States period (475–221 bce). Scholars who agree on the demonstrative origin for the copula shì mostly focus on its function as an anaphoric pronoun in critical syntactic contexts. A frequently cited study by Li and Thompson (1977), for example, postulates that the copula shì is the result of a reanalysis of the anaphoric shì used for recapitulating the topical np in the topic-comment construction, as formulated below:
This Topic-Comment structure is exemplified by (2):
In this structure, shì appears immediately after the topical np (i.e. NP 1) with which it is coreferential. After the reanalysis takes place, the referential function of shì gets lost, and shì only serves to link two nps.
Several studies, such as Feng (1993) and Shi (2001), have explored causes of the reanalysis in (1). Feng (1993) attributes the reanalysis to the loss of an obligatory pause between the topic and the comment. Shi (2001) stresses the impact of the increased use of the anaphoric shì in several new structures, and relates the reanalysis to the analogy to the svo order as well as to a change in the semantic relationship between the elements connected by shì. An important issue, however, not addressed so far, is why shì is the only demonstrative that evolved into a copula. Other demonstratives can also occur in the same “dem” slot in the topic-comment structure in (1). For example:
Opposing views (Feng 1984; Yen 1986) to the demonstrative origin invariably target on this weakness, and propose instead that the copula shì is derived from the verb/adjective shì or the particle shì with some sort of affirmative meaning. These alternative proposals, as Peyraube and Wiebusch (1994) criticize, rule out the pathway from the demonstrative shì, which occurs far more frequently, to the extent that an alternative copula interpretation becomes possible in some contexts. Therefore, the dominant opinion that favors the demonstrative origin is the most plausible interpretation except for one major obstacle: What makes shì so unique that it turns out to be the only demonstrative that develops into a copula?
What may bring new insight into the remaining issue concerning the origin of the copula shì comes from studies on the development of demonstratives in general. Diessel (1999), a comprehensive study on demonstratives, brings forward an important hypothesis regarding their grammaticalization. His major proposal is that demonstratives develop into different grammatical items depending on their syntactic positions. For copulas, he mentions two historical sources, i.e. third person pronouns and what he calls identificational demonstratives (Diessel 1999: 143ff.). The latter category refers to demonstratives occurring in copular and nonverbal clauses. According to him, the addition of this category to the conventionally recognized types of pronominal, adnominal, and adverbial demonstratives is supported by cross-linguistic evidence indicating that demonstratives in such a syntactic context can be phonologically or morphologically distinguished from other types. In terms of historical development, Diessel admits that the developmental pathway from third person pronouns to copulas follows the mechanism proposed by Li and Thompson (1977), but argues that the development from identificational demonstratives to copulas displays a different mechanism, which he illustrates with the formula in (4):
While the “dem” in (1) is coreferential with the topical np that precedes it, Diessel notes that the agreement features of “cop” in (4) appear to be determined by the predicate np that follows. For instance, in the Modern Hebrew (Ivrit) example in (5), zot, which he identifies as a copula, agrees in gender and number with the predicate noun dugma ‘example’, rather than with the subject ha-báyit:
Diessel argues that this agreement behavior of zot in Modern Hebrew is the same as that of identificational demonstratives, rather than anaphoric pronominal demonstratives. What follows is a pair of his examples showing that the anaphoric pronominal zot in (6a) agrees in gender and number with its antecedent, kasda, while the identificational ze in (6b) agrees with the noun, aba:
Although Diessel (1999) labels the new category “identificational,” he has not provided any discussion regarding any semantic or discourse importance underlying this type of demonstrative. From his examples, however, we notice that demonstratives of this type, such as ze in (6b), seem to show the function of identifying a contrastive referent in discourse. Such a function is also observable in the topic-comment construction in (1), in which the “dem” is used to identify the topical “NP 1,” i.e. the only entity, that qualifies the description of the following predicate. Therefore, although the mechanisms in (1) and (4) differ syntactically, the sentence position that the demonstrative occupies seems to be associated with the same function of identifying a contrastive referent in both cases. Such a similarity resonates with Stassen’s (1997: 76) observation that non-verbal copulas often originate from morphosyntactic devices employed to encode certain features, such as focus, related to information structure.
If the origin of a copular form is related to the structural expression of features of information structure, what prompts shì to develop toward a copula should not simply be its referential function. The sentence position of shì and its role in the information structure have been discussed extensively in studies on word order issues (e.g., Li and Thompson 1974; Ao 1983; Yin 1985; Peyraube 1997; Meisterernst 2010), but these aspects of shì have not been given much attention when examining the historical source of the copula shì. Previous studies concerning the origin of the copula shì normally take its function and usage in the classical language as the starting point of the change, and have never explored the possibility that the demonstrative shì originally possessed a certain trait that determines its later evolutional pathway. However, Qiu (1979) has reported that shì occurs exclusively in preverbal position in Bronze Inscriptions and transmitted texts from the Western Zhou (1046–771 bce), and that the postverbal occurrences of shì did not emerge until the late Spring and Autumn Period (6th–5th century bce). Since the syntactic placement of a form is strongly related to its role in information structure, it is possible that shì is a special demonstrative inherently associated with the expression of features of information structure. Thus, to answer why shì turns out to be the only demonstrative developing into a copula, we need to sort out its development from the earliest time when it emerged, rather than the period slightly before the actual grammaticalization process took place.
The present study, therefore, approaches the nature of shì by examining its early development. Evidence chosen for such a diachronic investigation includes Bronze Inscriptions (bi hereinafter) from the Western Zhou (c. 1046–771 bce) to the Warring states period (475–221 bce) and two received pre-classical texts, the Shī jīng (詩經, SJ hereinafter) and the Shàng shū (尚書, SS hereinafter). Through a detailed examination of the development of shì, this study will demonstrate that shì in its early stage of development indeed appears in more restricted syntactic contexts, and is strongly associated with the expression of focus in the left periphery of the sentence. Based on this finding, the study will present a new analysis on the original function of shì, which in turn will reveal a striking link between shì and the combinations consisting of the copula wéi 唯or huì 惠and the demonstrative pronoun zhī 之in the Shang oracle-bone inscriptions (13th–11th centuries bce). Finally, we will apply the new analysis to solve the main research question of this study, namely, why shì is the only demonstrative that developed the copular use.
2 Early Development of the Demonstrative shì
2.1 Background, Data and Methodology
The study of a particular demonstrative form can hardly be justified without addressing the entire demonstrative system in which this form occurs. Unfortunately, however, no satisfactory account of the demonstrative system(s) has been proposed for the time periods covered by this study. Proposals following the traditional analysis generally assigns shì to the proximate category. However, unlike a simple division between zhè 這 ‘this’ and nà 那 ‘that’ in Modern Standard Mandarin, those proposals are invariably overcrowded by multiple forms on each side. For example, the Western Zhou demonstrative system proposed by Zhang (2006) contains four proximate forms, zī/*tsǝ 兹, cǐ/*tsheɁ 此, sī/*se 斯, and shì/*deɁ 是 (with shí/*dǝ 時 as its variant), and five distal forms, bǐ/*paiɁ 彼, yī/*Ɂi 伊, jué/*kot 厥, qí/*gǝ 其, and zhī/*tǝ 之.3 While some pairs, such as zī 兹 vs. cǐ 此, are treated as diachronic variants, the exact difference between zī 兹 or cǐ 此, on the one hand, and shì 是, on the other, remains an unsettled issue. Most studies claim that they exhibit important semantic and syntactic differences. For instance, Wei (2004: 78) packages the difference first articulated by Ma (1898 : 53) with the notion of evidentiality, and suggests that cǐ refers to something observable to the interlocutors, while shì to something unobservable. Zhang (2006: 257ff.) stresses that shì occurs more often as an anaphoric pronoun, while zī 兹 is more frequently employed exophorically as a determiner. Djamouri (2001a), however, rejects any essential semantic or syntactic differences between cǐ (and sī 斯 considered to be a variant) and shì. He claims that the true difference lies in their enunciative values. The pivotal concept underlying his proposal is the distinction between énoncé (utterance) and the énonciation (the act of uttering an utterance). The proximity expressed by shì, according to him, is anchored in the space and time when the utterance is presented; in contrast, the deictic center of cǐ (and sī) concerns the world represented by the utterance, rather than the speech event.
There are also some attempts that apply more deictic contrasts to interpret the demonstrative system of the so-called shànggǔ hànyǔ 上古漢語 (Early Old or Archaic Chinese), the earliest stage—spanning from the earliest time to the Han (206 bce–220 ce) dynasty—in an influential periodization by Wang (1958).4 A medial or a deictically neutral term is usually included in those systems, and shì is mostly assigned to that type. Feng (1983), for instance, proposes a three-term system consisting of the proximal shí/*dǝ 時, the medial shì/*deɁ 是, and the distal fú qí/*ba *gǝ 夫其 and bǐ qí/*paiɁ *gǝ 彼其 in correspondence with Classical Tibetan ɦdi, de and p’agi, respectively.5 Systems suggested by Guo (1989) and Hong (1991) evoke both the medial and the neutral terms. While the former proposal assigns shì to the medial type, the latter groups it under the neutral, which Hóng calls the “jiānzhǐ 兼指 (versatile reference)” type. None of those studies, however, offers convincing evidence in favor of a three- or even four-way system on the ground of a rigorously defined framework for analyzing demonstratives. The “jiānzhǐ” category, for example, is basically a group of demonstratives frequently employed for anaphoric use, which should not be associated directly with the actual space occupied by the referents of those forms.
In short, there are mainly two types of analyses regarding the status of shì within an overall system of demonstratives in the pre-Qin period. The first type is to treat shì as a proximate form possessing some important differences from others in the same deictic category. The second type of treatment is to establish a separate deictic term for shì. Table 1 below lists some typical proposals:
Due to the complexity of this issue, we will skip an inevitably lengthy review on those proposals, and focus on analyzing the actual use of shì. This approach is justified based on the following considerations.
First, recent studies on demonstratives (Mithun 1987; Hanks 1992; Himmelmann 1996; Tao 1999, etc.) have argued that the essential function of these expressions in naturally occurring narratives and conversations is not to direct the listener to a referent in the speech situation, but to manipulate the information flow in ongoing discourse, and to encode the speaker’s stance toward a referent. For instance, in English, Strauss (2002) reports that such features as focus and information load play an important role in the choice of demonstratives. In her analysis, the switching from that to this in the conversation in (7) for the same referent, i.e. the suitcase that Schiavo tried to sneak into a flight that she did not plan to board, is due to the speaker’s (Laura’s) intention of raising the hearer’s attention to this referent.
Cases like (7) suggest that the actual use of demonstratives in non-elicited texts is complicated by discourse and pragmatic factors. Since historical investigations rely solely on non-elicited data, it is not realistic to expect demonstratives to be used mostly for expressing deictic contrast and violate their underlying deictic value only in rare cases. On the contrary, for demonstratives with no clear historical evidence indicating their deictic features, it might be more useful to study their actual usage and compare it with universals or tendencies informed by living languages. For example, if we find that a form in Classical Chinese is often associated with referents of high focus and high information load, we could at least conclude that this form behaves similarly to the proximate demonstrative in English.
It is also important to consider that since demonstratives occur frequently and extensively as cohesive devices, and are common historical sources of various grammatical items (Diessel 1999; Heine and Kuteva 2002), a demonstrative may become specialized in serving functions other than expressing deictic contrast at some point in its course of development. Synchronically, it is thus possible to encounter a system consisting of multiple demonstratives at different developmental stages. In this case, it might be even wrong to assume that all demonstratives in this system should contrast deictically.
For reasons stated above, we will not pursue the position of shì in a demonstrative system, but focus only on examining its actual use. As for the selection of research data, because the bi come with more reliable dating than transmitted texts, the present study will first establish a diachronic account of the development of shì in 2.2 and 2.3 based on bi data, and then evaluate tendencies deduced from the bi against evidence from the SJ and the SS in 2.4. The bi data used in this study are primarily collected from the Jīnwén yǐndé 金文引得 (Huadong shifan daxue zhongguo wenzi yanjiu yu yingyong zhongxin, comp., 2001; JWYD hereinafter), which is a concordance of 5,758 inscriptions from the Western Zhou in Volume 1 and 7,692 inscriptions from the Spring and Autumn and the Warring States periods in Volume 2. To ensure the inclusion of all important tokens, this study also consulted the concordance of the Yīn Zhōu jīnwén jíchéng 殷周金文集成 ((IA, CASS), ed., 1984–1994; Jicheng hereinafter), i.e. the Yīn Zhōu jīnwén jíchéng yǐndé 殷周金文集成引得 (Zhang 2001; JCYD hereinafter). Nevertheless, these concordances are only used for the purpose of collecting tokens. Transcription and interpretation of inscriptions are based on various references, including the Shāng Zhōu qīngtóngqì míngwén xuǎn 商周青銅器銘文選 (Ma et al., eds., 1986–1990; MWX hereafter), the Jīnwén gǔlín 金文詁林 (Zhou et al. 1974–5; JWGL hereinafter) and other published studies on bronze inscriptions.
2.2 Western Zhou Period (11th–8th Centuries bce)
In the Western Zhou data, we find only seven tokens of shì occurring exclusively as a pronominal.8 In these examples, shì is mostly coreferential with an np referring to a concrete object in the previous discourse except for one token (i.e. example 10, below) of shì referring to a proposition expressed previously.
The earliest token, as in (8), is found in the inscription on a bronze vessel from around the 10th century bce.9 This token occurs in a sentence expressing a prayerful wish, a common element in the dedication part that concludes a bronze inscription (Shaughnessy 1991: 83). Shì appears as a left dislocated object, and is co-referential with a prior np, “the ding-tripod of Lǚ.”
Since the inscription is physically on the vessel, it is possible to take shì in (8) as self-pointing, that is, the exophoric use of a proximate demonstrative. However, there are instances, such as (9), in favor of the anaphoric interpretation:
The referent of shì, the royal gift of a four-horse carriage, is unlikely to have been present when the narrator recounted the event. Therefore, the anaphoric interpretation seems more effective to account for all instances of shì referring to a concrete object.
The syntactic position of shì in (8) and (9) is typical to all identified tokens in the Western Zhou and subsequent time periods. In the Western Zhou, six out of the seven tokens of shì occur as the object in the preverbal position. The left dislocated shì reflects an ex-situ strategy for encoding focus or, more precisely, the identificational focus distinguished by Kiss (1995; 1998; 2001). The identificational focus expresses exhaustive identification, and involves syntactic reordering in languages that display a particular structural position for foci of this type. The other type, the information focus, conveys only new, non-presupposed information, and does not involve any syntactic reordering. In our translation, cleft sentences are adopted for reflecting the analysis that shì expresses the identificational focus.
As mentioned, there is one token of shì referring to a previously expressed proposition. This token is also the only instance of shì occurring after the verb that governs it. In (10), shì follows the word wéi 隹, but the string “wéi shì” is placed before the vo-predicate, “sàng wǒ guó 喪我國 (lose our territories).”
Shì in (10) refers to a hypothetical situation, i.e. blindly following the king’s decisions, that has been described in the previous discourse and would incur the consequence expressed by the following predicate, “to lose our territories”. In the pre-classical language, wéi functions as a copula and a particle marking focalization (Pulleyblank 1995: 131).15 The causal meaning present in (10) should be related to the explanatory force of the copula wéi. In his study on copulas in the oracle-bone inscriptions (13th–11th centuries bce), Takashima (1990) proposes that wéi, differing from the other Shang copula huì 惠, has the illocutionary meaning of providing a comment or explanation, and the clause in which wéi occurs often provides a reason or cause related to the preceding sentence or clause. What follows is an example for the explanatory wéi:
Djamouri’s (2001b) analysis would take two clauses in (11a) and (11b) as one, but still regard wéi as a copula linking two propositions in a “quasi-equative” clause.
The copula wéi can also occur in focalization constructions, and has evolved into a focalization marker by the late Shang (Takashima 1990; Djamouri 2001b). For instance:
Because both the copular and focalization marking functions of wéi survived into the Western Zhou (Qiu 2008: 185ff.), it is difficult to determine whether wéi in (10) is only responsible for the causal meaning as a copula, or if it also assumes the role as a focalization marker that highlights the proposition expressed by shì. Nevertheless, neither interpretation undermines the claim that shì is associated with the identificational focus placed in the left periphery of the sentence, since it is certain that the speaker, i.e. the king, in (10) intends to identify the conduct of blindly following his order as the exact cause for the loss of their state.
2.3 The Spring and Autumn and the Warring States Periods (8th–3rd Centuries bce)
The number of tokens of the demonstrative shì increases significantly in this period. Many of them show usages identical or similar to those described in 2.2. We will first review those usages in 2.3.1, and then move to discuss two important changes in 2.3.2.
2.3.1 Continuation of Earlier Usages
Among the total of 45 tokens, 22 display the typical pattern illustrated by example (8): Shì occurs as a pronominal demonstrative employed to track a prior np referring to a bronze vessel, and appears as an ex-situ focus landing in preverbal position. Here is an example:
The last two sentences in (13) express the same prayerful wish that their descendants would forever use those bronze bells. Both zhī 之 and shì are employed to track the same referent, the harmonic bells, in the previous discourse. While the first sentence shows the regular order of svo, shì in the second sentence occurs at the preverbal position following the subject.
In this period, there are more instances of shì referring to a proposition, and all of them happen to occur as causal expressions as well. However, there are no more instances of shì preceded by wéi. We identify three tokens of shì that can be associated with the expression of reason or cause in inscriptions dating from the late Spring and Autumn period. In all three cases, shì stands alone as a causal phrase before the main predicate. What follows is an inscription that contains two of the tokens:
Shì in (14) refers to the noble origin recounted in the prior discourse. Since it is normal for causal phrases to appear adverbially, it is difficult to determine whether shì is meant to be encoded as the identificational focus, which would be rendered as “it was due to this” in the translation, or if it is merely an adverb meaning “thus, therefore”. The ambiguity, however, reflects a transient stage as shì occurs increasingly in this syntactic context and eventually evolves into a sentence connective.
The third token of shì occurring alone as a causal phrase involves some controversy in interpretation due to the reduplication marks next to the graphs representing the proper noun, “哀成叔”:
Due to the reduplication marks, there seems to be no better solution than taking “嘉是隹哀成叔” as forming a sentence. Preliminary studies by Zhao (1981) and Zhang (1981) both apply this punctuation, and read the sentence as ‘Jia is exactly the Āi Chéng Shū.’ This reading is adopted in MWX, but rejected by many other studies because it is abnormal for Jia, who is the narrator of this inscription, to address himself with a posthumous name. Cai (1985) later argues that although reduplication marks normally indicate reduplication, they may sometimes be used to mark proper nouns. If the proper noun “哀成叔” is not meant to be reduplicated, the whole string of “嘉是隹哀成叔之鼎永用禋祀” is a well-formed sentence, in which yòng is the main verb and wéi marks the focused object “哀成叔之鼎.”17 The function of shì in this analysis, serving as the adverbial causal phrase, is the same as that of shì in (14).
In the bi from the Warring States period, all instances of shì referring to a reason or cause are found in the expression shì yǐ 是以 ‘due to this (?), thus’. There are in total five tokens of “shì yǐ ” from two famous vessels from the state of Zhōngshān 中山. What follows is one of them:
The word yǐ 以meaning ‘because of, due to’ is derived from the verb yǐ which originally means ‘to take, to use.’ The verb yǐ meaning ‘to use’ is still attested in the Late Spring and Autumn period:
In (17), shì is used to track a concrete np, i.e. ‘ritual utensils as dowries’, and occurs as a left dislocated object before the verb yǐ ‘to use’. The same word order is seen in the causal phrase “shì yǐ ” in (16). The only difference is that shì refers to a proposition, rather than to a concrete object, in (16).18 This ordering indicates that the causal phrase “shì yǐ ” is derived from a focalization structure that encodes the proposition referred to by shì as the prominent segment. However, as we have admitted in the case of shì in (14), it is difficult to determine whether the string “shì yǐ ”, which occurs before the vo-predicate “cì zhì jué mìng 賜之氒命” in (16), has turned into a device employed purely for linking two propositions in the causal relationship. We may contend, however, that the placement of shì before the main predicate still coincides with the structural position where the identificational focus is expected.
2.3.2 Two New Developments
The first significant development is the occurrence of shì between the left dislocated object and the verb in the [(S) O shì V] and the [(S) wéi O shì V] constructions. The former pattern, four tokens in total, is attested only in the Spring and Autumn period. What follow are two of them:
The construction with wéi 隹（唯）as a focus marker is found in three tokens from the Warring States period.19 Here is an example:
The referent of shì in the earlier [(S) O shì V] pattern can be ambiguous if the subject is elided. For example, the underlined sentence in (19)—disregarding the context—could be interpreted as “it was this (principle, person, etc.) that Heaven’s mandate will follow.” Such ambiguity does not exist in (20) due to the occurrence of wéi as the focus marker, which is rendered by the word only in the translation. In addition, the construction with wéi is also more flexible in marking different constituents. What follows is a case in which the subject is marked as the focus:
Since this construction is not restricted to highlighting the object, it is more precise to formalize it as the [wéi X shì Y] construction.
When shì appears in a separate clause/sentence, its function in terms of information flow is to reactivate and highlight an entity or proposition in the background. In the two newly emerging constructions, shì is still co-referential with the highlighted np, such as ‘four positions’ in (18), ‘Heaven’s mandate’ in (19), and ‘Protector’ in (20). The only difference is that the need of reactivation has become virtually none with the focused np appearing right before shì in the same clause. In this regard, the rise of the [(S) O shì V] construction in the Spring and Autumn period reflects a change whereby a global focalization construction that operates beyond clause boundaries turns into a mono-clausal one. With the need of reactivation diminishing in the mono-clausal construction, the anaphoric force of shì becomes redundant, which might lead to its transition to a marker of focalization.
The second important development, for which only one token is found, is the occurrence of shì in the adnominal position:
Although they are written with the same graph, adnominal and pronominal shì share few common traits. While pronominal shì is associated with a prior np and is exclusively anaphoric, the adnominal shì in (22) modifies the np following it, and is most likely exophoric, since no co-referent seems present in the surrounding discourse.21 Moreover, the np marked by shì in (22) also displays the regular svo order. On the grounds of these differences and the fact that this bronze vessel is from the state of Yān, which belongs to the northern cultural sphere in Lǐ’s (1985) classification, we might entertain the possibility that adnominal shì in (22) exhibits a regional feature of North China during the Spring and Autumn period.22
2.4 Comparison of Major Trends Revealed by the bi, the SJ, and the SS
Table 2 provides the number of tokens for each use in the bi from the Western Zhou and subsequent time periods. From the distribution of those tokens, we can see that shì occurs originally and predominantly as a demonstrative pronoun employed to reactivate and highlight an entity or proposition that has been established in the previous discourse, and possesses a dual force of anaphora and focalization in all time periods. The [(S) O shì V] construction is attested only in the Spring and Autumn period, while the [wéi X shì Y] construction appears only in the Warring States period. Although it might be debatable whether shì is a focus marker in these two constructions (Meisterernst 2010), the function of shì is labeled as focalization marking on the basis of its obligatory occurrence in these two focalization constructions. The earliest occurrence of adnominal shì is in the Spring and Autumn period, which is a later development of shì as well.
Although transmitted texts of the pre-Han period tend to be less translucent in terms of dating, they can be incorporated to test tendencies suggested by the bi. Considering the time depth of the language reflected in the texts, we will use the SJ and the SS ( jīnwén 今文 ‘New Text’ version) for comparison, and test the following hypotheses:
(23)Hypotheses to be tested in the SS and the SJ
Tokens of anaphoric and focalizing shì occur extensively in every section/chapter. For this use of shì, we consider only tokens of shì occurring as the object. The occurrence of shì in sentence/clause-connecting expressions, such as “shì yǐ 是以” (one token in the SS and seven tokens in the SJ) and “shì yòng 是用” (seven tokens in the SJ), will not be included.
Tokens of focalization marking shì occur in sections/chapters dating later than the Western Zhou.
Tokens of adnominal shì occur in sections/chapters dating later than the Western Zhou.
Tables 3 and 4 provide the number of tokens for the three uses of shì in the SJ and the SS respectively. The sections/chapters in which those tokens are found are divided into three groups based on their composition time in these two tables. First, in Table 3, informed by the discussion by Zhang (2006: 2ff., 20f.), the first group of texts reflects the language of the early Western Zhou period (11th–10th centuries bce), the second group mainly reflects the late Western Zhou (9th–8th centuries bce), and the last group, the Spring and Autumn period (8th–5th centuries bce).
The demonstrative shì occurs 149 times in the SJ. Hypothesis (i) is supported by a total of 59 tokens of anaphoric and focalizing shì in all three groups of texts. The lowest frequency of the [wéi X shì Y] structure in the SJ also confirms the hypothesis that this pattern arose latest in the Warring States period. However, the SJ seems to suggest that the [(S) O shì V] structure and the adnominal shì were in active use during the late Western Zhou.
There is a possible token of adnominal shì in a poem (Mao 244) from the first group:
It is unclear how Karlgren reads shì in (24), but it is certain that he does not take it as the determiner. Syntactically speaking, this analysis is certainly possible, and the line in this reading strongly resembles a quote from the announcement issued by King Wu following his conquest of the Shang:
However, the adnominal reading would make shì 是 a proximate demonstrative in (24). Since plenty of evidence suggests that zī 茲 is undoubtedly the proximate demonstrative in the Western Zhou period, we remain skeptical toward this token of shì.
In Table 4, chapters in the first group are those which are commonly regarded as being created during the Western Zhou, the second group are generally suspected of being composed during the Spring and Autumn or the Warring States period, and the last group contains chapters composed no earlier than the Warring States period and possibly as late as the Han period.
The frequency of shì in the SS—17 times in total—is far lower. The anaphoric and focalizing shì is found in chapters dating to the Western Zhou, but its frequency is not high either, if we adopt the conventional dating of “Hóng fàn.”23 The non-occurrence of adnominal shì seems to support hypothesis (iii), if we assume that later composers had access to some earlier texts or intentionally kept a more conservative style of language.24 Two tokens of [wéi X shì Y] from the “Lì zhèng” chapter, however, do not support hypothesis (ii).
To sum up, if we consider all the evidences examined so far, it is possible to push the date of the [(S) O shì V] structure and the adnominal shì slightly earlier to the end of the Western Zhou based on the SJ, but it is still more plausible to maintain that [wéi X shì Y] arose later. The low frequency of [wéi X shì Y] in the SJ and its major distribution pattern in the SS chapters confirm that the active use of this structure emerged in the Warring States period.
3 The Origin of the Demonstrative shì 是
As mentioned in 2.1, shì is often assigned to the same deictic category as zī 茲 or cǐ 此, and many studies suggest that the major difference is that the latter two forms are mainly used for expressing deictic contrast, while shì is weak in deictic meaning. The present study agrees that shì is exclusively in the anaphoric use till the appearance of the adnominal shì around the Spring and Autumn period. However, we must not overlook important syntactic differences between shì, on the one hand, and zī or cǐ, on the other. Compared to 42 (out of the total of 45) tokens of the adnominal zī and 4 (out of the total of 6) tokens of the adnominal cǐ, there is only one token of adnominal shì in our bi data. More important, our diachronic investigation reveals that shì in origin is indeed not a regular demonstrative like zī or cǐ. It contains both the anaphoric force to track an np or a proposition expressed in the previous discourse and the focalizing force to encode the identificational focus. Therefore, we propose that shì in origin is an anaphoric and focalizing pronominal demonstrative strongly related to the structural expression of the identificational focus.
Before the emergence of shì, it was common for zhī/*tə 之 to appear as the focused object marked by the focalizing copula wéi/*wi 隹（唯）or huì/*wîs （惠）, the choice of which is mainly conditioned by the controllability of the action and the presence/absence of modality (Takashima 1990), in Shang Chinese. Zhī in (26a) and (26b), for example, is co-referential with an np, ‘Zhǐ Jiā’, in the previous clause, and precedes the verb cóng ‘to follow’:
The strings of wéi zhī and huì zhī are only attested in the Shang period, while the new form shì with the same function emerged in the subsequent Western Zhou period. If not coincidental, it seems reasonable to suggest that shì is in fact derived from the Shang expression wéi zhī or huì zhī. This proposal has a great advantage in accounting for the focalizing force as well as the special structural position that shì occupies in the sentence. Since shì appears neutral in modality, we may even narrow the source down down to wéi zhī.
Nevertheless, although the function of shì perfectly matches that of wéi zhī (and huì zhī), phonologically speaking it remains difficult to justify the following processes of sound change:
While the changes involving the weakening of *ɢwi are entirely speculative, the vowel change from *ə to *e, albeit an irregular one in oc, is possible. The same alternation is found between the proximate demonstrative zī/*tsə 茲 in the pre-classical language and cǐ/*tsheʔ 此 in the classical language. Pulleyblank (2000) has discussed this vowel alternation in the pair zī and cǐ and the pair shì 是 and shí/*dǝ 時, which he considers to be a form morphologically related to zhī 之, only “stronger” and “more independent”. He claims that cǐ is derived from zī in the same way that shì is derived from shí, but provides no further explanation.26 Although we are unclear on the exact nature of the *ə vs. *e alternation, the fact that the same alternation is found in two sets of demonstratives indicates that such a connection should not be accidental.27
Zhī/*tə 之, according to Djamouri (1999), had lost its deictic meaning and became distance neutral in the Western Zhou.28 It is thus predictable that the newly emerged form, shì, does not inherit the original deictic force, but only the anaphoric function of zhī. Following the disappearance of the combinations of wéi zhī and huì zhī, shì emerges as an anaphoric demonstrative specialized in activating a prior np or proposition at the preverbal position. In this respect, shì can be interpreted as an allomorph of zhī at the preverbal position that encodes the identificational focus in the sentence.
It has been suggested that the commonly attested focalization construction [wéi X shì Y] in the classical language is the convergence of two focus-marking strategies: an old one that employs wéi to mark the following constituent as the focus and a new one that uses shì to mark the preceding constituent as the focus (Wang 1958; Tang 1991). Meisterernst (2010) opposes the traditional analysis of shì as a focus marker. Her proposal regards [NP s (wéi) NP o shì V] as a cleft structure that does not involve the left dislocation of NP o, and takes the function of shì as marking embedded nominalization. The hypothesis proposed by the present study, however, implies that from the earliest time, shì, if not etymologically related to wéi, is at least associated with the same type of focalization operated by the focalizing copula wéi, and that the [S O shì V] structure emerging later is simply an innovation that turns a global focalization operation across clausal boundaries into a mono-clausal construction. The addition of wéi, which had ceased to function as a copula and had become a focus marker, in the Warring States period serves to clarify the potential ambiguity and make the structure more flexible in highlighting diverse types of constituents.
It needs to be noted that multiple constructions can result in the sov order in the classical language, and among them, zhī seemingly displays a similar function of marking focalization in the [NP s NP o zhī VP] construction, which overtook the [NP s NP o shì VP] by the late Warring States period (Yin 1985; Peyraube 1997). However, the present study maintains that zhī should not be analyzed as a resumptive pronoun coreferential with the preceding NP o. The [np zhī vp] pattern in general is an innovation based on earlier patterns of [vp zhī np] and [NP 1 zhī NP 2]. The occurrence of zhī in the [np zhī vp] pattern is a critical diagnostic context that confirms the transition of shì from a demonstrative to a marker of embedded nominalization as generally claimed or an attributive marker as argued by Yue (1998). Therefore, in the case of the [NP s NP o zhī VP] construction, we find Meisterernst’s (2010) analysis more acceptable in treating zhī as marking an embedded nominal phrase.29
Finally, our bi data suggest that the occurrence of shì at the adnominal position may have started as a regional innovation. Although adnominal zhī is rare in the classical language, its occurrence in the pre-classical language is not. In the late Shang period, adnominal zhī is frequently found before time expressions and occasionally before other nouns (Djamouri 1994; Yue 1998; Deng 2011). In the Western Zhou period, the occurrence of [zhī np] becomes extremely rare (one token in our bi data). The commonly encountered patterns became [NP 1 zhī NP 2] and [vp zhī np]. Nevertheless, Yue (1998) suggests that zhī is still the determiner in these two common patterns. Thus, it is not impossible to propose that adnominal shì is also etymologically related to zhī.30
The present investigation finds that shì originally is only used pronominally and is strongly associated with the structural expression of the identificational focus. We identify shì as an anaphoric and focalizing demonstrative pronoun that functions in the same way as “wéi zhī ” and “huì zhī ” in the Shang time. Although we have not found solid evidence in support of the hypothesis that shì is directly derived from wéi zhī, the morphological link between shì and zhī is well supported by their phonological forms and their syntactic and pragmatic functions.
Our analysis of shì has an important implication for the mechanism proposed by Li and Thompson (1977) for the development of the demonstrative shì. Let us revisit the function of shì in formula (1):
In our analysis, shì does not simply reactivate a prior referent, but also highlights and contrasts it in the current discourse. Therefore, the string [shì
2], which might appear to be in the normal subject-predicate structure, is in fact a focalization construction. Based on this analysis, example (2) should be rendered as follows:
This analysis predicts that shì conveys more salience in information structure than normal anaphoric forms, and accounts for the emphatic sense commonly associated with shì.
Moreover, the present study also argues that instead of relating the current clause to a previous one, shì evolved to highlight a coreferential segment within the same clause, first in the [(S) O shì V] construction in the Spring and Autumn period and then in the [wéi X shì Y] construction in the Warring States period. As the anaphoric force of shì became virtually redundant, the effective function of shì changed to signal focalization only. It is important to note that the rise of the copular use of shì through the mechanism in (1), similarly, involves the loss of anaphoric force, and features the same direction of change from a bi-clausal (i.e. [NP 1][DEM NP 2]) to a mono-clausal construction (i.e. [NP 1 COP NP 2]). If Qiu (1979) and Tang (1991) are correct in dating the earliest copular use of shī to the Warring States period, the rise of the focalization marking shì happens as a prelude to its emerging transition to a copula, indicating that the grammaticalization processes that turn shì into a focalization marking device and a copula are related events.
As our investigation sorts out the early development of shì, what becomes clear is that shì is not a typical demonstrative in origin, and that its evolutionary pathway from a demonstrative to a copula is not only based on its anaphoric force, but on its association with the structural expression of focus as well. The combination of the anaphoric and focalizing functions furnishes shì with the capacity to reactivate a prior referent at the structural position that encodes the identificational focus in the current discourse. Therefore, it is important to acknowledge that the function of identifying a contrastive referent might be the feature essentially shared by the demonstrative and the copula shì. This hypothesis is in line with Stassen’s (1997: 76) cross-linguistic observation on the origin of non-verbal copulas, and calls for further research on the semantic and discourse importance of the category of “identificational” demonstratives proposed by Diessel (1999).
Cai Yunzhang 蔡運章. 1985. Aichengshu dingming kaoshi 哀成叔鼎銘考釋 [Deciphering the inscription of the Aichengshu ding-tripod]. Zhongyuan wenwu 中原文物 [Cultural Relics of the Central Plain] 4: 56–62.
Chen Mengjia 陳夢家. 1954–6 . Xi Zhou tongqi duandai 西周銅器斷代 [Periodization of Western Zhou bronzes]. Kaogu xuebao 考古學報[ Journal of Archaeology]. Beijing: Zhonghua Book Company.
Djamouri Redouane . 1999. The evolution of Zhi in Archaic Chinese. In Peyraube Alain and Chaofen Sun (eds.) Studies on Chinese Historical Syntax and Morphology: Linguistic Essays in Honor of Mei Tsu-lin33–47. Paris: École des Hautes Études en Sciences Sociales, Centre de Recherches Linguistique sur l’Asie Orientale.
Djamouri Redouane . 2001a. Système des pronoms démonstratifs en chinois bas archaïque. In Djamouri Redouane (ed.) Collected Essays in Ancient Chinese Grammar161–176. Paris: Ecole des Hautes Études en Sciences Sociales, Centre de Recherches Linguistique sur l’Asie Orientale.
Feng Chuntian 馮春田. 1984. Cong Wang Chong Lun heng kan youguan xici ‘shi’ de wenti 從王充《論衡》看有關繫詞“是”的問題 [Issues concerning the copula ‘shi’ based on the texts of the Lun heng by Wang Chong]. In Cheng Xiangqing (ed.) Liang Han hanyu yanjiu 兩漢漢語研究 [Research on the Chinese Language during the Western and the Eastern Han Dynasty] 341–364. Jinan: Shandong Education Press.
Feng Shengli 馮勝利. 2013. Shanggu yinyun yanjiu de xin shijiao 上古音韻研究的新視角 [New perspectives in the study of Old Chinese phonology]. In Shi Feng and Gang Peng (eds.) Da jiang dong qu - Wang Shiyuan jiaoshou bashi sui heshou wenji 大江東去——王士元教授八十歲賀壽文集 [The Great River Flows East: Festschrift in Honor of Professor William S.-Y. Wang] 71–84. Hong Kong: City University of Hong Kong Press.
Feng Zheng 馮蒸. 1983. Guanyu Hanzang yuxi kongjian zhishici de jige wenti 關於漢藏語系空間指示詞的幾個問題 [Several issues concerning the spatial demonstratives in Sino-Tibetan languages]. In sya ron sō 均社論叢 [Journal of In sya] 13: 1–12.
Guo Moruo 郭沫若. 1935 . Liang Zhou jinwenci daxi tulu kaoshi 兩周金文辭大系圖錄考釋 [Pictures and Decipherment of Bronze Inscriptions of the Western and the Eastern Zhou Dynasties]. Beijing: Science Press.
Guo Xiliang 郭錫良. 1989. Shi lun shanggu hanyu zhishidaici de tixi 試論上古漢語指示代詞的體系 [A tentative discussion on the system of demonstratives in Archaic Chinese]. Yuyan wenxi xueshu lunwenji 語言文字學術論文集 [Proceedings of Research on Language and Writing]. Beijing: Knowledge Press. Also in Guo Xiliang (1997) Hanyu shi lun ji 漢語史論集 [An Anthology on History of the Chinese Language] 75–96. Beijing: Commercial Press.
Hong Bo 洪波. 1991. Shanggu hanyu zhidaici shumian tixi de zai yanjiu 上古漢語指代詞書面體系的再研究 [Re-examining the demonstrative system of written Archaic Chinese]. Yuyan yanjiu lun cong 語言研究論叢 vol. 6. Also in Hong Bo (1999) Han Tai yu zhuizhi 漢台語錐指 [Research on the Chinese and Tai Languages] 1–31. Tianjin: Nankai University Press.
Huang Shengzhang 黃盛璋. 1983. Xianqin gu hanyu zhishici yanjiu 先秦古漢語指示詞研究 [Study on demonstratives in Pre-Qin Chinese]. Yuyan yanjiu 語言研究 [Language Research] 2: 136–157.
Li Xueqin 李學勤. 2008. Zheng ren jinwen liangzhong dui du 鄭人金文兩種對讀 [Comparison of two bronze inscriptions of the Zheng people]. Zhonghua guoxue yanjiu 中華國學研究 [Research on the Traditional Study] 1: 85–88.
Ma Jianzhong 馬建忠. 1898 . Ma shi wentong 馬氏文通 [Ma’s Grammar of Classical Chinese]. Beijing: Commercial Press.
Pan Wuyun 潘悟雲. 2001. Shanggu zhidaici de qiangdiao shi he ruohua shi 上古指代词的强调式和弱化式 [The stressed and weakened forms of demonstratives in Archaic Chinese]. In Fan Kaitai and Huyang Qi (eds.) Yuyan wenti zai renshi 語言問題再認識 [Rethink Some Linguistic Issues]. Shanghai: Shanghai Education Press. Also in Pan Wuyun (2002) Zhuming zhongnian yuyan xuejia zixuanji Pan Wuyun juan 著名中年語言學家自選集－潘悟雲卷 [Anthology of Famous Middle-aged Linguists - Volume of Pan Wuyun] 293-311. Hefei: Anwei Education Press.
Peyraube A and Wiebusch Thekla . 1994. Problems relating to the history of different copulas in Ancient Chinese. In Chen, Matthew Y. and Ovid J. L. Tzeng (eds.), In Honor of William S.-Y. Wang: Interdisciplinary Studies on Language and Language Change383–404. Taipei: Pyramid Press.
Pulleyblank Edwin G. 2000. The morphology of demonstrative pronouns in Classical Chinese. In Wang, Samuel H., Feng-fu Tsao and Chin-fa Lien (eds.), Selected papers from the Fifth International Conference on Chinese Linguistics1–25. Taipei: Crane Publishing Co., Ltd.
Qian Zongwu 錢宗武. 2004. Jinwen Shangshu yufa yanjiu 今文尚書語言研究 [A Study on the Grammar in the New Text Version of the Shangshu]. Beijing: Commercial Press.
Qiu Xiejun 裘燮君. 2008. Shang Zhou xuci yanjiu 商周虛詞研究 [Study on “Empty Words” in the Shang and the Zhou dynasties]. Beijing: Zhonghua Book Company.
Qiu Xigui 裘錫圭. 1979. Tantan guwenzi ziliao dui guhanyu yanjiu de zhongyaoxing 談談股文字資料對古漢語研究的重要性 [On the importance of paleographic materials in the research on the Archaic Chinese language]. Zhongguo yuwen 中國語文 [Studies of the Chinese Language] 6: 437–442 458.
Qiu Xigui 裘錫圭. 2002. Bingong xu mingwen kaoshi 燹公盨公銘文考釋 [Deciphering the inscription of Bingong xu-container]. Zhongguo lishi wenwu 中國歷史文物 [ Journal of National Museum of China] 6: 23–24.
Shi Yuzhi 石毓智. 2001. Hanyu yufahua de licheng 漢語語法化的歷程 [The Grammaticalization Process of the Chinese Language]. Beijing: Peking University Press.
Tai James H.-Y. and Chan Marjorie K.M. 1999. Some reflections on the periodization of the Chinese language studies in Chinese historical syntax and morphology. In Peyraube, Alain and Chaofen Sun (eds.), Linguistic Essays in Honor of Mei Tsu-lin (Collection des Cahiers de Linguistique d’Asie Orientale)223–239. Paris: École des Hautes Études en Sciences Sociales, Centre de Recherches Linguistique sur l’Asie Orientale.
Tang Lan 唐蘭. 1986. Xizhou qingtongqi mingwen fendai shizheng 西周青銅器銘文分代史徵 [Periodization of Western Zhou Bronze Inscriptions]. Beijing: Zhonghua Book Company.
Tang Yuming 唐鈺明. 1991. Shanggu panduanju de bianhua kaocha 上古判斷句的變換考察 [Examination on movements in declarative sentences in Archaic Chinese]. Zhongguo yuwen 中國語文 [Studies of the Chinese Language] 5: 388–389.
Wang Li 王力. 1937. Zhongguo wenfa de xici [中國文法中的繫詞 The copula in Chinese grammar]. Tsing Hua Journal of Chinese Studies 清華學報 8.1: 1–67.
Wang Li 王力. 1958. Hanyu shi gao 漢語史稿 [Manuscript on the History of the Chinese Language]. Beijing: Science Press.
Wei Peiquan 魏培泉. 2004. Han Wei liuchao chengdaici yanjiu 漢魏六朝稱代詞研究 [A Study of Pronouns in the Han Wei and Six Dynasties]. Taipei: Academia Sinica.
Yang Bojun 楊伯峻 and He Leshi 何樂士. 1992 . Guhanyu yufa jiqi fazhan 古漢語語法及其發展 [Archaic Chinese Grammar and its Development]. Beijing: Yuwen Press.
Yin Guoguang 殷國光. 1985. Xianqin dai yufa biaozhi de binyu qianzhi jushi chutan 先秦漢語帶語法標志的賓語前置句式初探 [Preliminaries with marked preposed objects in Pre-Qin Chinese]. Yuyan yanjiu 語言研究 [Language Research] 2: 162–171.
Yue Anne O. 1998. Zhi 之 in pre-Qin Chinese. T’oung Pao 84: 239–292.
Zhang Yachu 張亞初. 2001. Yin Zhou jinwen jicheng yinde 殷周金文集成引得 [A Concordance of the Collection of Shang and Zhou Bronze Inscriptions]. Beijing: Zhonghua Book Company. [ JCYD]
Zhang Yujin 張玉金. 2004. Xizhou hanyu yufa yanjiu 西周漢語語法研究 [A Study on Western Zhou Chinese Grammar]. Beijing: Commercial Press.
Zhang Yujin 張玉金. 2006. Xizhou hanyu daici yanjiu 西周漢語代詞研究 [A Study on Pronouns in Western Zhou Chinese]. Beijing: Zhonghua Book Company.
Zhang Zhenglang 張政烺. 1981. Aichengshu ding shiwen 哀成叔鼎釋文 [Interpretation of the inscription of Aichengshu ding-tripod]. Guwenzi yanjiu 古文字研究 [Study of Paleography] 5: 27–83.
Zhao Zhenhua 趙振華. 1981. Aichengshu ding de mingwen he niandai 哀成叔鼎的銘文和年代 [The inscription and the dating of the Aichengshu ding-tripod]. Wenwu 文物 [Cultural Relics] 7: 68–69.
Zhou Fagao 周法高 et al. 1974–5. Jinwen gulin 金文詁林 [A Concordance and Dictionary of Bronze Script]. Hong Kong: The Chinese University Press. [ JWGL]
Unless otherwise noted, the reconstructions follow the “Minimal oc system” (ocm) in Schuessler (2007).
We will return to this sentence, and provide a different translation that reflects our analysis of shì as an anaphoric and focalizing demonstrative. The deictic nature of shì, as discussed in the following section, remains an open issue in the field. The question mark here also indicates such uncertainty.
To note, zhī 之 is assigned to the proximate category in an earlier study by him (see Zhang 2004: 98ff.)
This widely used periodization has many problems, such as the lack of distinction between literary and vernacular language, and failure to capture significant transitional periods when applied to the history of Chinese grammar. For detailed discussion, see Tai and Chan (1999).
Coblin (1986: 149) lists the same correspondences between Chinese shí and shì and Tibetan ɦdi and de. Pulleyblank (2000), however, suggests that it is zhī 之that corresponds to Tibetan de ‘that’, and shì 是, which he reconstructs as *ătàjʔ, corresponds to Tibetan ɦdi ‘this.’ He also points out that the deictic contrast between Tibetan ɦdi and de should be regarded as “independent specialization,” and cannot be applied to Chinese.
Only forms identified as pronominal and adnominal demonstratives are included in the comparison. Pronunciations are only provided for the first occurrence of each form in the table.
Pulleyblank (1995: 79) admits the occasional use of zhī 之, which he mainly discusses in his section of 3rd person pronouns, as an adnominal demonstrative, and translates it as ‘this.’
The JCYD includes one more token of shì from Hū Ding 曶鼎 (Jicheng 5.2838). The alleged presence of this word is dismissed for reasons stated in fn.19.
Chen (1954–6 : 131) dates this vessel to the reign of King Kāng (1005/3–978 bce), while Tang (1986: 344) assigns it to the reign of King Mù (956–918 bce). The dating to the early Western Zhou in the Jicheng agrees with Chen’s opinion, while the MWX, assigning this piece to the mid-Western Zhou, agrees with Tang’s opinion.
The “#” symbol indicates the beginning and the end of an inscription.
For the reading of kǎo/*khûʔ 考 as xiào/*khrûs 孝, see Tang (1986: 344).
As mentioned, the deictic nature of shì is an open issue. The question mark intends to reflect the controversy.
The interpretation of the phrase shèng mǎ 乘馬 as a carriage drawn by four horses follows Chen (1992: 601).
There are different interpretations of these two sentences. We adopted what seems the most straightforward reading suggested by Guo (1935 : 137).
Not all researchers accept the copular function of wéi. The copula wéi is interpreted by some as an assertive modal particle or as an adverb modifying a nominal predicate.
The interpretation of xiào 孝 as meaning jìng 敬 ‘to respect’ follows Tang (1986). The reading that takes yǔ/*ŋaʔ 語 ‘to speak, tell’ as a phonetic loan for yù/*ŋwâʔ 娛 ‘to entertain’ follows MWX.
Cai (1985) suggests the punctuation that takes “嘉是隹哀成叔之鼎” as a sentence. His translation reads, “Jia therefore had the ding-tripod cast for Āi Chéng Shū”. This reading, as he admits, requires an insertion of an elided verb, “to cast,” after shì wéi 是隹, which he takes as forming a phrase equivalent to shì yǐ 是以 ‘thus, therefore’ in classical Chinese. We follow Cài’s analysis of the reduplication marks, but do not agree with his punctuation because it seems unlikely that the main verb would be omitted in this way. Li (2008) proposes yet another punctuation, in which Jiā belongs to the previous sentence, and “是隹哀成叔” stands alone as a single sentence. Due to limited length, we will not discuss why this punctuation makes no better sense than what is proposed in (15) in terms of content, but will only point out the disadvantage that the referent of shì in his interpretation is a person, a very unlikely usage of shì as an independent pronoun.
Whether shì refers to an np or a proposition expressed in the previous discourse is the key to distinguishing whether yǐ is used to introduce an instrument or a reason. Based on this criterion, we only recognized five tokens of shì yǐ expressing the causal relationship.
The occurrence of shì as one of the two missing graphs in “必唯朕[禾][是]賞” in Hū Dǐng 曶鼎 (Jicheng 5.2838), a Western Zhou piece, is suggested in several major references (JCYD, JWYD, and MWX). It is, however, unlikely that the [(S) wéi O shì V] construction would be encountered in a Western Zhou inscription. It is also uncertain whether the space for the alleged missing graphs on the rubbing was meant to accommodate two graphs. Chen (1954–56), for example, supplements only one graph, hé 禾 ‘millet plant’ after zhèn 朕.
Guo (1935 : 228) interprets xián/*gin 賢 as jìn/*dzinʔ 賮 meaning ‘to present (tribute).’ Phonologically this is possible since some mc dz- initials may come from oc *sg- (Sagart 1999: 63–65). But, since Fújí is the recipient of the tribute, we expect the occurrence of yú 于 in this sentence to mark the source, “Xiǎnyú.” The MWX suggests that xián 賢 has the meaning of shèng/*lhəŋh 勝 ‘to conquer, overcome’ based on the occurrence of the phrase suì xián 歲賢 in the Huáinánzĭ 淮南子, which was composed during the Western Han (206 bce–220 ce). Nevertheless, the face value of xián/*gin 賢 ‘wise’ can be extended to mean ‘to outwit’ (Behr, p.c.), which works very well in this context. This interpretation remains the most plausible one among all readings proposed so far.
The context does not provide clear indication of whether shì marks a particular deictic contrast. But, since shì can be interpreted as self-pointing in this case, we might speculate that it marks proximity.
Behr (2017) notes that the word ‘qie-vase’ here might be a substrate word, related to Proto-Japanese *ke (<*ka-Ci) > Old Japanese ke ‘vessel, container’ or Middle Korean kali ‘fish pot, container’ (< Old Korean *ka[l,t]).
Although the previous consensus is that the “Hóng fàn” chapter contains popular cosmological notions, such as wǔxíng 五行, of the Warring States period, some recent studies, such as Qiu (2002), based themselves on a newly found bronze vessel, have argue that this chapter might have been created during the early Western Zhou.
The demonstrative written as shí/*dǝ 時 does occur adnominally in the SS. But, it is dubious to assume a diachronic connection between shí 時 and shì 是. See fn. 26 for explanation.
Schuessler’s (2007) reconstruction does not adopt the theory that mc velar fricative and glottal initials reflect a set of uvular and labiouvular initials in oc. Accepting Baxter and Sagart’s (2014: 43ff.) new reconstruction, I have modified the initial of wéi from Schuessler’s *w- to *ɢw -.
The demonstrative written as shí 時 occurs only in the more archaic parts of the SJ and the SS, and is commonly regarded as representing an earlier form of shì 是. However, our bi data clearly supports the emergence of shì in the earlier half of the Western Zhou. The syntactic functions of these two in the SJ and the SS, according to Deng (2011: 267ff.), also show a complementary distribution. On these grounds, it is better to regard them as contemporary forms, both derived from zhī 之.
Pan (2001) proposes a distinction between stressed (or emphatic) and unstressed (weakened) forms. Cĭ 此 and shì 是 are categorized as the stressed (or emphatic) type, while zī 茲 and zhī 之, the “weakened” type. It is unclear whether his notion of “stressed vs. unstressed” is purely prosodic or contains additional semantic, syntactic, or pragmatic differences. Nevertheless, it has been suggested that Archaic Chinese may be a mora-sensitive (or quantity-sensitive) language (Feng 2013). It is possible that the vowel contrast here is a manifestation of such prosodic distinctions as stressed vs. weakened or heavy vs. light syllables. Although these four forms do not form a contemporary system, there may be semantic or pragmatic motivation that the newly emerged shì 是 in the Western Zhou and cĭ 此 in the early Spring and Autumn period are both the stressed type.
We find cases of anaphoric zhī occurring at the preverbal position in two inscriptions, Jicheng 7.4107 and Jicheng 9.4615. The exceptional sentence position of zhī is most likely due to rhyming (Ao 1983; Yin 1985).
Meisterernst (2010) also offers a detailed discussion of syntactic differences between the [NP s NP o shì VP] and the [NP s NP o zhī vp] constructions.
The demonstrative shí 時 often occurs adnominally in the SJ (11 tokens) and the SS (22 tokens).