Non-Wh Internal and External Topics in Classical Chinese and Their Similarity to Equivalent Constructions in Modern Mandarin Flipped Classroom: Its Effects on ESL Learners’ Critical Thinking and Reading Comprehension Levels

In this paper, I investigate Classical Chinese (5thc BC-3rdc AD), viz. the written form of Old Chinese ranging from the end of the Spring and Autumn era (770-476 BC) to the end of the Han dynasty (206 BC-220 AD), covering the Warring States period (475-221 BC). During the Archaic period, the official medium of written expression in China, Classical Chinese (or 文言 wenyan ‘Literary Chinese’), was formed. Classical Chinese had functioned as the main written language used in literary texts since the shanggu stage, but its use had progressively diminished from the beginning of the Medieval period during which the current Abstract In this paper I argue that non-wh internal and external topics in Classical Chinese can be analysed in parallel with their counterparts in modern Mandarin. I first scrutinise pronoun fronting in the context of negation in Classical Chinese, and propound that it is on a par with object preposing in modern Mandarin that is argued to be internal topics (Paul 2002). I then discuss non-interrogative elements base-generated and moved to the CP domain and illustrate their topical properties. Base-generated topics in Classical Chinese are Aboutness topics which permit DPs exclusively. In terms of moved external topics, they are constituted of preposed objects and preposed predicates: the former allow DPs and TPs, whereas the latter allow DPs and AdjPs. Moreover, moved topics can be further divided into Hanging Topics and Left Dislocation Topics, similar to those in modern Mandarin. Received: 04/04/2020

contemporary Chinese: although there is a small proportion of words from Ancient Chinese that is still attested in the contemporary language, there are numerous words that have disappeared from the modern language or have adopted disparate meanings (He andJiang 1980, Peyraube 2008).
Analogous to its modern counterpart, Classical Chinese is an SVO language, but a derived SOV order is also prevalent (Light 1979, Sun and Givón 1985, Peyraube 1994, Djamouri 2001, Aldridge 2010, Meisterernst 2010, Djamouri et al 2007, among many others). In Classical Chinese, when nominal and pronominal DPs function as objects, they can occur in their postverbal base positions, following the canonical V-DP order, as shown in (1) and (2) respectively. In the context of negation, pronominal, but not nominal, DPs may undergo preposing from their postverbal base positions to preverbal landing sites, generating a derived DP-V sequence without the support of any additional marking. Example (3a/b/c) illustrates the fronting of a first, second and third person pronoun respectively. It is worth mentioning that pronoun fronting in the context of negation is a robust feature of Classical Chinese (Wang 1958, Shi 1986, Yu 1989, Wei 1990, Feng 1996, Peyraube 1997, Djamouri 2000, Aldridge 2010 Although zero-marking preposing is restricted to pronominal objects and requires a negative environment, if a morpheme 之 zhī or 是 shì 1 (4a-b/c) is employed to mark the 1 之 ZHI and 是 SHI are referred to as contrastive markers in Peyraube (1996) and pretransitive/preverbal markers in Peyraube (1997). It is notable that only ZHI can function as the fronting marker following preposed wh-constituents, yet *wh-SHI is never attested, in that SHI always occurs below negation, yet wh never follows negation due to the Intervention Effect (Wang 2015). Apart from being a fronting marker, the same graph zhī may also function as a third person accusative pronoun (either animate or inanimate), a resumptive pronoun, a genitive marker, or a marker for explicit subordination (Wang 2013). As for the fronting marker SHI, it can additionally function as a demonstrative or determiner shì 'this', and it may originate from the demonstrative shì through a process of grammaticalisation (Peyraube 1997).

Flipped Classroom: Its Effects on ESL Learners' Critical Thinking and Reading Comprehension Levels
International Journal of Language and Literary Studies 4 fronting, then fronted DPs can be either nominal or pronominal (4a/b), and the context can be either negative or positive (4a-b/c).
(4) a. Furthermore, when a preposed nominal/pronominal object is followed by ZHI/SHI, it can be preceded by a morpheme 唯 wéi indicating 'only' (5). I state that 唯 wéi is to introduce foci, parallel to the copula shì in modern Mandarin (Cheung 2008, 2014, Paul and Whitman 2008, Hole 2012

TOPICS IN MANDARIN
As a topic-prominent language (Chao 1968: 69-72, Li andThompson 1976), modern Mandarin has attracted a significant amount of research interest on its topic-comment constructions for analysis of information-structural categories. According to traditional views, topic in modern Mandarin displays robust distribution and properties: it is definite or generic; it has discourse features; it occupies sentence-initial position of the first clause in a topic chain; it can be separated from the rest of the sentence by particles a/ya, ne, me or ba; it can extend its semantic domain to more than one clause; it is in control of the pronominalisation or deletion process of all coreferential NPs in a topic chain; except in clauses where it is also subject, it plays no role in such processes as true reflexivisation, Equi-NP deletion and imperativisation (Li and Thompson 1976, 1981, Tsao 1978, 1987, Li 1990, Jiang 1991, Qu 1994. Relational definition of topic is '[s' X [s ... Y ...]], where X is a major category and Y, possibly empty, is related to X' (Xu and Langendoen 1985: 20). Relational givennessnewness concerning the divide between topic and comment entails a partition of the semantic/conceptual representation of a sentence into two complementary parts, including what the sentence is about (the logical/psychological subject) and what is predicated about the other part (the logical/psychological predicate) (Gundel and Fretheim 2006). Based on relational givenness-newness, a topic is defined as the part of the sentence which is structurally marked as what the sentence is about or the frame in which the sentence holds, while a comment is marked as the predicate which holds for the topic (von Prince 2012). Shi (2000) defines topic based on its structural position, structural relationship with the verb and discourse function. Topic-comment construction is treated as a syntactic device derived from basic sentence structures via syntactic operations, so as to fulfill discourse functions. A topic 'represents an entity that has been mentioned in the previous discourse and is being discussed again in the current sentence, namely, topic is what the current sentence is set up to add new information to' (386). Topic is correlated with a position inside the comment and it depends on an element inside the comment for its thematic role. Given the fact that the thematic role is not independent, topic does not have syntactic function of its own.
A topic can be linked to a gap, a resumptive pronoun or an epithet. The coreferential relation between a topic and a gap inside the comment is subject to locality constraint which is based on the disparity between a gap and a resumptive pronoun coreferential with the topic. Although resumptive pronouns can be employed to replace gaps, they should be avoided whenever possible. However, when topics are related to elements inside islands such as relative and adverbial clauses, resumptive pronouns are obligatory and leaving gaps in the comments would generate infelicitous topic-comment constructions (Li and Thompson 1979, Huang 1982, Li 1990, Jiang 1991, Shyu 1995, Shi 2000, Del Gobbo and Badan 2007, Huang et al 2009, Badan and Del Gobbo 2011, Cheung 2014. Topics in Mandarin can be categorised into two types, viz. base-generated ones and moved ones. Base-generated topics denote Aboutness Topic (AT) structures without a real resumptive element or a pro or a trace in the sentence, and they can be further divided into 'whole-part' ATs and 'possessive' ATs. Moved topics are constituted of Hanging Topic (HT) and Left Dislocation Topic (LD). There are six distinctions between these two kinds of topics. First, HT must be a bare DP, yet LD can be a PP. Second, HT always requires a resumptive pronoun and only agrees with it in number and gender; LD only requires a resumptive pronoun (clitic) when it is correlated with a direct or partitive object, and LD agrees with the clitic in gender, number and Case. Third, HT does not need to agree with the resumptive pronoun inside the predicate in Case, yet LD agrees with the resumptive pronoun in Case. Fourth, HT can be resumed by a tonic pronoun or epithet, yet LD cannot. Fifth, it is permissible to have multiple LDs, but not HPs. Sixth, HT and LD can co-exist and the former always precedes the latter (Benincà andPoletto 2004, Badan andDel Gobbo, 2011).
In modern Mandarin, an object can undergo fronting from its postverbal base position to the 'low IP area' (dubbed by Paul (2005)) between TP and vP, preceding adverbs, negation and auxiliaries. Such object preposing is different from the lián…yě/dōu 'even' construction (7), as it requires no additional marking (8).
(7) a. Object preposing without marking in modern Mandarin is commonly accepted as a case of focalisation in the literature (Ernst and Wang 1995, Shyu 1995, Zhang 1997, Tsai 2000, among many others). Nevertheless, Paul (2002) invalidates the focalisation account on both syntactic and semantic grounds, and postulates that object preposing in Mandarin are internal topics between the subject and vP, occupying the specifier position of a functional projection. Parallel to external topics, internal objects can be derived either via movement or base-generation; disparate from external topics, internal topics cannot be clauses, and multiple topics are prohibited from the internal topic position.

INTERNAL TOPICS IN CLASSICAL CHINESE
There are two types of topics in Classical Chinese, namely internal topics in low IP area, as well as external topics in the CP area. In this section, I discuss internal topics.
In line with Paul's (2002) proposal on modern Mandarin, I suggest that preposed non-wh objects in Classical Chinese are also internal topics, parallel to their modern counterparts. Since object preposing in Classical Chinese takes the form of pronoun fronting to negation, preposed pronouns in the context of negation are internal topics.

Pronoun Fronting to Negation in Classical Chinese vs. Object Preposing in Mandarin
There is no denying the fact that object preposing in modern Mandarin and that in Classical Chinese are not completely consistent.
First, object preposing in standard Mandarin (but not in some dialects) forbids pronouns (Hou 1979, Paul 2002, whereas that in Classical Chinese are restricted to pronouns (9), and preposing of nominal DPs without any accompanying marker is not attested in corpora. Second, object preposing in modern Mandarin takes place in both positive and negative contexts (cf. (8a) and (8b)), whereas that in Classical Chinese only occurs in a negative context (see (9)).
Third, preposed objects in Mandarin target a landing site to the left of adverbs, negation and auxiliaries (Paul 2002), whereas those in Classical Chinese intervene between adverbials/negators and auxiliaries. Evidence demonstrates that pronouns base-generated postverbally move to a position following an aspecto-temporal adverb 將 jiāng and negators 不 bù or 未 wèi (9a-b) and preceding an auxiliary 能 néng. That is to say, the relative order between preposed objects and medial elements in Mandarin is Obj>Adv/Neg/Aux>V, yet that in Classical Chinese is Adv>Neg>Obj>Aux>V.
Despite the discrepancies, object preposing in modern Mandarin and that in Classical Chinese share three common properties.
First, both object preposing in Classical Chinese and that in modern Mandarin are derived from movement, rather than base-generation. As mentioned previously, the canonical order for both Classical Chinese and modern Mandarin is SVO, and the reverse SOV order is derived from movement from postverbal base positions to preverbal landing sites.
Second, neither preposed nominal/prenominal objects in Mandarin nor preposed pronominal DPs in Classical Chinese requires any marking. Pronouns in (9) and (10) (see below), for instance, undergo movement to a position following negation, and no marker is attached to them.
Third, analogous to modern Mandarin, Classical Chinese allows a preposed DP to be either [+human] or not, as exemplified by a third person accusative pronoun 之 zhī in (10a/bc).

Pronoun Fronting to Negation vs. ZHI/SHI Focus
In Classical Chinese, two types of SOV constructions need to be distinguished, i.e. pronoun fronting in the context of negation, and focus constructions marked by ZHI/SHI. The discrepancies between these two kinds of constructions lend support to the topical nature of pronoun fronting to negation.
First, pronoun fronting to negation and focus marked by ZHI/SHI permit distinct types of preposed elements. As is obvious from the term, pronoun fronting to negation denotes pronominal DPs exclusively. Focus marked by ZHI/SHI, however, is not restricted to pronominal DPs (see (14) below). ZHI/SHI focus also allows nominal DPs, clauses and whitems.
Examples in (12) show that clausal objects can be found in foci marked by ZHI/SHI, yet no data involving zero-marking contain any clausal objects. Second, pronoun fronting to negation and focus marked by ZHI/SHI display disparate relative orders with negation. To be more specific, when a pronoun raises independently, it always targets a position between negation and vP, as shown in (9-10). However, if a DP undergoes preposing accompanied by a marker ZHI/SHI, its landing site is above negation, as in (14a) (=(4b)) and (14b). That is to say, there exists a hierarchy among preposed items, markers and negators in the medial domain, viz. DP > ZHI/SHI > Neg > Pron.
Note that as can be seen from (14b), apart from its functional use as a fronting marker, the morpheme 是 shì can also function as a demonstrative pronoun. Third, constructions marked by ZHI/SHI are compatible with an 'even' focus interpretation, whereas pronoun fronting to negation is not. Under an 'even' circumstance, the speaker can presuppose that there are other elements sharing the same feature as that attributed to the quantified elements (Paris 1998). In (15a), based on a following rhetorical question and contextual information, the former clause should be assumed to imply an 'even' interpretation for the element being quantified that the person does not even adore his own parents, not to mention others. A similar interpretation is available in (15b) as well.

Pronoun Fronting to Negation vs. Cleft
In Classical Chinese, when a preposed DP is marked by ZHI/SHI, the construction can be preceded by a morpheme 唯 wéi indicating assertive modality and frequently translated as 'only' (Djamouri 2001) to form a cleft structure WEI…ZHI/SHI. Cleft contractions exhibit exhaustiveness and bipartition, yet neither phenomenon exists in pronoun fronting in negation.
First, exhaustiveness is absent from constructions involving pronoun fronting to negation, whereas it is present in WEI…ZHI/SHI clefts. Exhaustiveness specifies an exhaustive set for which a given proposition holds true, excluding all other possibilities (É. Kiss 1998, Cheung 2014. In (16a), exclusiveness is expressed, in that 唯 WEI 'only' excludes the DP 'mansion' in the former clause, rendering the clefted DP 'neighbour' the only available option. Moreover, the former clause indicates that the property of being chosen denoted by the presupposition is not held by the DP 'mansion' that is negated by a negator 非 fēi. The former clause in (16b) that involves pronoun fronting to negation, however, does not impose exhaustivity, and the following clause confirms that an alternative is true. Second, bipartition into focus and presupposition is absent from constructions involving pronoun fronting to negation, whereas it is present in constructions involving clefts. This asymmetry concerning bipartition is reflected by the fact that whole VPs in sentences involving pronoun fronting to negation can be negated/questioned, yet negation and interrogation do not apply to entire VPs in sentences involving WEI…ZHI/SHI. In (17a/b) that involves a raised pronoun, it is not only the pronoun that is negated/questioned; the verb is negated/questioned as well. Nevertheless, in (18), bipartition excludes the presupposition from the scope of negation/interrogation, and only the raised DP is affected. As for another property implied by bipartition, i.e. listing, the reason why it is not realised in WEI…ZHI/SHI structures is due to this cleft per se: WEI expresses an 'only' interpretation that is incompatible with list contexts. To summarise, fronted pronouns in the context of negation in Classical Chinese function as internal topics, similar to preposed objects in modern Mandarin, in that they both have zero marking and optional [+human] feature, and they are both derived from movement. In addition to the topical properties it demonstrates, pronoun fronting to negation also displays properties that are inconsistent with foci, including ZHI/SHI focus and WEI…ZHI/SHI cleft. For one thing, pronoun fronting to negation is disparate from ZHI/SHI focus, because they permit distinct preposed elements and display different relative orders with negation, and only the latter can adopt an 'even' interpretation. For another, pronoun fronting to negation is dissimilar to WEI…ZHI/SHI cleft, in that the latter, but not the former, exhibits exhaustiveness and bipartition.

EXTERNAL TOPICS
In Classical Chinese, a fronted pronoun can target a landing site between negation and vP in the low IP area, functioning as an internal topic. When a preposed object or predicate raises to a position above subject in the left periphery, it functions as an external topic. Additionally, external topics can be base-generated. That is to say, in Classical Chinese, internal topics are restricted to moved ones, whereas external topics can be moved ones and base-generated ones.

Topical Features
In Classical Chinese, objects and predicates can undergo movement from their base positions to a sentence-initial position in the CP domain.
As mentioned previously, objects are base-generated postverbally in this SVO language. Apart from staying in situ, objects have an option to raise to the left periphery ccommanding the subject. Preposed objects can be either DPs or TPs, as in (19a)  In terms of predicates, they are base-generated following the subject in the sentenceinternal domain. The canonical position of predicates is in (20) which also shows that a copula intervening between the subject and the predicate is optional. Apart from staying in situ, predicates can undergo optional movement from its base position following the subject to the CP area, and the preposed predicates display features of external topics.
zhì ér tǐ, jūnzǐ yě consolidate aspiration Conj realise gentleman PART '(Those who) consolidate and realise aspiration are gentlemen' Before discussing the topical properties of predicates moved to a sentence-initial position, I would like to present the types of preposed predicates. Different from fronted objects that are restricted to DPs, fronted predicates allow DPs and AdjPs, both of which can land in the CP domain preceding the subject. (21a/b) illustrates the movement of a nonwh/wh DP predicate from its base position below a subject to its landing site above the subject, ending up in the CP area. Examples in (21c-d) contain AdjPs. Coming to the topic-like features of moved and base-generated phrases in the CP domain, the first feature is that the DPs are always definite or generic, and data with indefinite DPs in the left periphery are not attested in corpora. This observation applies to both preposed and base-generated DPs. (23a) and (23b) exemplify definite and generic DPs base-generated in the sentence-initial position. (24a-b) and (24c) exemplify fronted definite and generic DPs functioning as objects. As for the raised predicate DPs 君子jūnzǐ (21a) and 小人 xiǎorén (25), they denote a pair of antonymous Confucian concepts representing ideal men and their opposite in general. So despite their translation, these DPs are regarded as being generic. The other property justifying the claimed topical nature of preposed objects and predicates is the existence of particles. After objects and predicates move to a sentence-initial position in the CP area, they can be marked by an optional particle, and thus being separated from the comment. As can be observed from (23-24), particles are not obligatory. Examples in (21) and (25) illustrate that a possible particle could be 哉 zāi. Additionally, 也 yě and 乎 hū may also be employed to mark fronted elements, as in (26a/b).

Coreferentiality
In Classical Chinese, a base-generated external topic is not related to any position in the rest of the sentence, because there is no gap or resumptive pronoun that is linked to this topic. A preposed object/predicate functioning as an external topic, however, can be linked to a gap or a resumptive pronoun inside the comment, and the coreferential relation between the topic and the gap/pronoun is subject to locality constraints.
When a DP is base-generated in the CP area as a topic, it does not link to a gap or resumptive pronoun within the comment. In (27a), (=(23a)), the DP in the sentence-initial position is base-generated in the left periphery, yet it is not related to any element in the medial domain, or even subcategorised by the verb. It is notable that within the comment, a DP undergoes preposing from a postverbal base position to a position above negation and it is followed by a marker ZHI, so it should be regarded as a focus construction. As for (27b), the sentence-initial DP in the CP domain is also a base-generated external topic, and the topic is not linked with any gap or resumptive pronoun. Similar to (27a), (27b) also contains a fronted object DP, but this DP is a pronoun fronted to negation without marking, so it is an internal object. That is to say, (27a) involves an external topic and a (internal) focus, whereas (27b) contains an external topic and an internal topic. When a DP/TP object moves from its postverbal base position to the CP area, this external topic can be linked to a resumptive pronoun in the medial domain. Example (28a) (=(24c)) contains a DP external topic base-generated postverbally as an object, and its coreferential resumptive pronoun additionally becomes an internal topic: in the context of negation, the pronoun undergoes movement from its postverbal base position to a position between the negator and the verb in the low IP area. In (28b), an object fronts to a sentenceinitial position and acts as an external topic, and its related resumptive pronoun inside the comment also undergoes movement. However, different from (28a) which involves pronoun fronting to negation, viz. an internal topic, (28b) contains a preposed DP followed by a marker ZHI, which I assume is a focus. As for (28c) (=(19b)), it exemplifies a preposed TP object corresponding to a resumptive pronoun inside the comment. As can be observed from examples that have been presented so far (and those below), the third person accusative pronoun 之 zhī is the most frequently attested resumptive pronoun, and the demonstrative pronoun 是 shì as in (28b-c) is less common. Alternatively, the topic-related resumptive pronoun can be null, leaving a gap inside the comment. Whenever a resumptive pronoun is replaced by a gap, the topic-related item is never within an island. (29a-b) and (29c) exemplify object and predicate DPs respectively, and none of them is base-generated in an island, so they can be linked to a gap, rather than a compulsory resumptive pronoun. Nonetheless, a resumptive pronoun is obligatory when the topic is coreferential with an element inside an island. In (30), the object DP 'sage' moves out of an island caused by coordination, so there is an overt resumptive pronoun linked to this topic. Data concerning DP movement out of islands without resumptive pronouns is never attested in corpora. Therefore, I posit that the coreferential relationship between an external topic realised by a preposed object/predicate and its resumptive pronoun is subject to locality constraints.

Base-Generated Topics vs. Moved Topics
In Classical Chinese, external topics can be categorised into two groups, namely, base-generated topics and moved topics. The former denotes AT, while the latter is constituted of HT and LD.
As discussed previously, there are DPs base-generated in the CP domain and they display no syntactic connection with the rest of the sentence, as shown in (31) (=(27)). These base-generated topics are ATs that can be further divided into 'possessive' and 'whole-part' ATs. (31a) describes a 'possessive' relationship between the person as an AT and his body as a resumptive element. By contrast, (31b) expresses a 'whole-part' relation between the topic and its resumptive element in the comment, in that they denote the same cohort. In terms of moved topics, they can be further divided into HTs and LDs. Both types of topics entail movement from the low IP area to the CP domain, as well as a related gap/resumptive pronoun inside the comment.
HTs in Classical Chinese display three properties: 1) they can only be DPs; 2) they require obligatory resumptive pronouns; and 3) they do not agree with resumptive pronouns in Case. As can be observed from examples that have been presented so far, there are preposed DPs in CP area linked to resumptive pronouns in the low IP area, and I state that these moved DPs are HTs, as in, for instance, (28)  It should be pointed out that HTs in modern Mandarin agree with their coreferential resumptive pronouns in number and gender (Badan and Del Gobbo 2011), so it seems (32b) serves as an example indicating that HTs in Classical Chinese are distinct from their modern counterparts. There is no denying the fact that the DP in (32b) is constituted of two coordinated parts, yet the accusative pronoun 之 zhī appears to be singular, so there seems to be a mismatch in number between the topic and its resumptive pronoun. Nevertheless, when the morpheme 之 zhī functions as a third person accusative pronoun in Classical Chinese, there is lack of morphological progress to mark singularity or plurality. Moreover, 之 zhī as a third person pronoun does not apply morphological change to distinguish genders. For instance, since the topic DP in (32a) is [+masculine], as its resumptive pronoun, zhī is presumed to be 'him'; in (33), however, zhī is employed to refer to a person of a [-masculine] feature. Besides, the same morpheme may function as a marker ZHI, as in (31a). The lack of morphological conversion concerning singularity/plurality, masculinity/femininity and even word class applies to the less common resumptive pronoun 是 shì as well, because the same morpheme shì can freely function as a singular/plural/masculine/feminine demonstrative pronoun, fronting marker (18) or determiner (33b). As a consequence, the interpretation of pronouns zhī and shì entirely depends on contextual information. Therefore, (32b) does not jeopardise the claimed parallel between HTs in Classical Chinese and those in modern Mandarin. The other type of moved topics is LDs which display three properties: 1) they can be phrases other than DPs; 2) they do not require obligatory resumptive pronouns; and 3) they are preceded by complementisers.
As mentioned previously, predicates fronted to the CP area are not restricted to DPs, because they permit AdjPs as well. Furthermore, preposed objects are not limited to DPs either, as they can be TPs too. Therefore, preposed AdjP predicates and TP objects should be postulated as LDs, as in (34a)  As regards the second property, LDs in Mandarin require resumptive pronouns when they correspond to objects (Badan and Del Gobbo 2011), yet LDs in Classical Chinese always have an option to leave resumptive pronouns empty, even if being linked to objects. For instance, in (35), although the DPs function as direct objects, when they move out of the low IP area to the left periphery, they do not correspond to any resumptive pronoun in the comment, so external topics like those in (35)  In terms of the third feature of LDs, it indicates that if a LD and a complementiser coexist, the complementiser always precedes the LD. As can be seen from Example (35b), COMP precedes the DP in the CP domain, so this external topic should be regarded as a LD. This statement coincides with the previous analysis of (35b) as a LD based on the second feature.
It is worth mentioning that the relative order between topics and foci in Classical Chinese shows that topics are located in structurally more prominent positions than foci. This observation coincides with that in modern Mandarin (Paul 2005, 2006, Hsu 2008, Badan and Del Gobbo 2011, Cheung 2013.

CONCLUSION
In this research, I explore non-wh constructions functioning as internal and external topics in Classical Chinese and suggest that topics in Classical Chinese and those in modern Mandarin can be analysed in line with each other.
In the low IP area, pronoun fronting in the context of negation is a robust feature of Classical Chinese. Pronoun fronting to negation in Classical Chinese is parallel to object preposing in modern Mandarin, in that both phenomena involve zero marking and [±human] DPs and both structures are derived from movement. Since preposed objects without marking in modern Mandarin are internal topics (Paul 2002), I hypothesise that pronouns fronted to negation should also be internal topics. To reinforce my argument, I compare pronoun fronting to negation with foci marked by ZHI/SHI and point out that these two constructions require distinct types of preposed elements and display disparate relative orders with negation; furthermore, the latter, but not the former, is compatible with an 'even' focus interpretation. Moreover, pronoun fronting to negation is in contrast to WEI…ZHI/SHI cleft, in that only the latter exhibits exhaustiveness and bipartition.
In the CP domain, both base-generated and moved topics marked by optional particles are attested from Classical Chinese corpora. Base-generated DPs in the left periphery are not linked to any element in the rest of the sentence, so they are presumed to be ATs. Additionally, these base-generated ATs can be further divided into two subcategories based on their 'possessive' or 'whole-part' interpretation. An object or predicate raising from its base position in the sentence-internal domain to the sentence-initial position preceding the subject, however, always corresponds to a gap or a resumptive pronoun in the rest of the sentence, and the coreferential relation between the preposed object/predicate and its resumptive pronoun/gap is subject to locality constraints. Due to these features, fronted objects and predicates are postulated to be moved external topics. Discrepant from internal topics that are exclusively pronouns and from base-generated external topics that are restricted to DPs, moved external topics permit object/predicate DPs, predicate AdjPs, as well as clausal objects. Furthermore, moved topics in the CP area can be categorised into HTs and LDs, depending on the types of phrases they allow, presence/absence of resumptive pronouns, as well as their relative orders with complementisers.
In this research, I have investigated Classical Chinese that has robust linguistic properties yet is less analysed than its modern counterpart. Through comparing and contrasting Classical and Contemporary Chinese within a historical syntactic framework and via corpus linguistic methodologies, the features and changes of the Chinese language in different periods have been revealed. Further research on the similarities and differences between Classical Chinese and modern Mandarin can be conducted in the future.