Show Summary Details

Page of

Printed from Oxford Research Encyclopedias, Linguistics. Under the terms of the licence agreement, an individual user may print out a single article for personal use (for details see Privacy Policy and Legal Notice).

date: 29 January 2023

Chinese Reflexivesfree

Chinese Reflexivesfree

  • Yingying WangYingying WangHunan University
  •  and Haihua PanHaihua PanChinese University of Hong Kong


Among Chinese reflexives, simple reflexive ziji ‘self’ is best known not only for its licensing of long-distance binding that violates Binding Condition A in the canonical Binding Theory, but also for its special properties such as the asymmetry of the blocking effect. Different researchers have made great efforts to explain such phenomena from a syntactic or a semantic-pragmatic perspective, though up to now there is still no consensus on what the mechanism really is. Besides being used as an anaphor, ziji can also be used as a generic pronoun and an intensifier. Moreover, Chinese has other simple reflexives such as zishen ‘self-body’ and benren ‘person proper’, and complex ones like ta-ziji ‘himself’ and ziji-benshen ‘self-self’. These reflexives again indicate the complexity of the anaphoric system of Chinese, which calls for further investigation so that we can have a better understanding of the diversity of the binding patterns in natural languages.


  • Linguistic Theories

1. Reflexives and the Canonical Binding Theory

It is well known that the Canonical Binding Theory (CBT) developed by Chomsky (1981) provides a strict tripartite conception of noun phrases, according to which an anaphor is bound in its governing category (Binding Condition A), a pronominal is free in its governing category (Binding Condition B) and R-expressions are free everywhere (Binding Condition C).1 CBT takes a reflexive pronoun to be an anaphor subject to Binding Condition A (BCA). The complementary distribution between reflexives and pronouns with respect to the governing category is evident from the following examples in English.


In recent decades, work on binding has led to the discovery of an ever-increasing range of exceptions to BCA. The long-distance binding of reflexives is one of these exceptions. Long-Distance Reflexives (LDRs) that have attracted much attention in the literature include Icelandic sig (e.g., Reuland, 2011), Italian proprio (e.g., Giorgi, 2006, 2007), and Dutch zich (e.g., Rooryck & Wyngaerd, 2011) in European languages, as well as Chinese ziji (e.g., Pan, 1995, 1997, 2001), Japanese zibun (e.g., Kuno, 1987), and Korean caki (e.g., Han & Storoshenko, 2012) in Asian languages. As an illustration, the simple reflexive ziji in the Chinese version of (1a) can be bound by the matrix subject John that is outside its governing category, hence ‘long-distance bound’, though the embedded subject Bill is also a possible antecedent.


The reason why these LDRs cannot be straightforwardly accommodated is that, under CBT, such elements should count as pronominals, since they can be bound outside their governing categories. At the same time, it is clear that they are unlike standard pronominals in the sense that, except for very specialized ‘logophoric’ contexts which depict the thoughts, feelings, or point of view of a protagonist, they cannot point to entities in the larger discourse. With respect to these semantic properties, LDRs seem to pattern together with locally bound ones such that they cannot refer deictically. Hence, they are called Long-Distance Anaphors (LDAs), as their antecedents could be located outside their canonical governing categories.

The phenomenon of LDAs has prompted discussion on whether one should abandon or revise BCA, or rethink the assumption that LDAs are indeed anaphors that are directly bound long-distance (cf. Huang & Liu, 2001).

2. Chinese LDA ziji

Against this background, many researchers have investigated the reflexive ziji in Chinese. It has been found that LDA ziji has its special properties, which may not be observed in LDAs in other languages.

2.1 The properties of LDA ziji

Ziji is standardly known to be devoid of all features (person, gender, and number), thus allowing for a range of antecedents (Xu, 1993, 1994). In the literature, LDA ziji is often associated with three properties—that is, a de se interpretation, subject orientation, and the blocking effect (Huang & Liu, 2001; Pan, 1997, 2001, among others), though up to now the precise description of its properties has not been agreed on by researchers, which makes the interpretation of LDA ziji more mysterious.2

The idea that LDA ziji is an obligatory de se anaphor was first proposed by Pan (1995, 1997). He claims that LDA ziji corresponds to the quasi-indicator he* in English, as discussed in (Castañeda, 1966, 1968), viz. the third person pronoun he used in indirect speech that is the result of replacing I in direct speech. Based on this assumption, if (3) is uttered, the reported speech made by John must be ‘I am smart’. If what John actually said is ‘That guy is very smart’ (and unbeknownst to John, he is actually the person to whom that guy refers), one cannot report John’s speech by using the sentence in (3).


However, as observed by later researchers (Anand, 2006; Pollard & Xue, 2001), LDA ziji can also be used in an extensional context like (4), where the de se reading is not available.3


Recently, Wang and Pan (2014, 2015) further argue that even in an intensional context, LDA ziji can be interpreted as non-de se, or indirect de se, especially in indirect speech. According to them, in the situation where John said ‘That guy is very smart’ instead of ‘I am smart’, if the speaker stands in John’s shoes looking at this event—or using the notion of empathy (Kuno, 1987), if the speaker empathizes with John and attributes the relevant property to him—he or she can report John’s speech by uttering the sentence in (3). So, it seems that the de se requirement of LD ziji is not necessary.4

Next, (5) illustrates the property of subject orientation of LDA ziji.


In this sentence, ziji can only have the matrix subject Zhangsan as its possible antecedent; it cannot be bound by the matrix object Lisi. However, this generalization is not without exceptions. As observed by many researchers, nonsubjects can also be accepted as the antecedents of LDA ziji in the following situations: (a) backward binding (Giorgi, 2006) occurs (e.g., (6)); (b) the predicates in question imply non-coreference between the subject and ziji (e.g., (7)); and (c) there is an animate feature conflict between the subject and ziji (e.g., (8));




It is worth noting that, although the inanimate possessed NP Zhangsan’s letter in (8b) and (8c) cannot bind ziji, the possessor Zhangsan can do so when the main verb is anshi ‘hint’, but not biaoming ‘indicate’. According to Pan (2001), this is because that Zhangsan’s letter hints at X entails that Zhangsan hints at X, though the same inference pattern does not hold for biaoming ‘indicate’. These examples seem to suggest that for the LDA ziji, the subject is its preferred antecedent, but not its only antecedent. Discourse factors may also play an important role in determining the possible antecedents of LDA ziji.

Finally, the so-called blocking effect shows that a first/second person pronouns can block ziji from extending its binding domain beyond the clause containing these pronouns. A case in point is (9):


Moreover, the blocking effect in Chinese has the following two special features, as first pointed out by Pan (1997, 2001): (a) it is asymmetrical, namely that first/second person pronouns can induce the blocking effect, but third person NPs do not necessarily block the long-distance binding of ziji by first/second person NPs (e.g., (10a) vs. (10b)); and (b) nonpotential binders may act as blockers, that is, first/second person pronouns in other grammatical functions, not just subjects or those contained in a subject, can also block LD binding of ziji (e.g., (11)).



In addition, it is also worth noting that there are several controversial cases in the literature. The first case involves first/second person pronouns at the non-subject positions in the matrix clause like (12). Pan (2001), Anand (2006), and others claim that there is no blocking effect involved, whereas Huang and Liu (2001) disagree with them.


The second case concerns sentences where first/second person pronouns in the embedded clauses do not intervene between the potential antecedent and ziji (e.g., (13)). According to Pan (2001), such a pronoun does not induce the blocking effect. However, others believe that first/second person pronouns can induce the blocking effect in any position of the embedded clause (Anand, 2006; Giorgi, 2006; Huang & Liu, 2001).


The third case is about ziji in BA and BEI sentences. Cole and Wang (1996) claim that the nominals right after ba and bei in the following sentences do not induce the blocking effect.


Cole et al. (2001) further add that the co-reference between Zhangsan and ziji in (14) is not so easily obtained, though it is possible, and conclude that there is a mild blocking effect involved. Hu (2019) also treats this kind of case as an exception, although others do not.

2.2 Theories on LDA ziji

A lot of theories have been proposed to account for the properties of LDRs. Most early analyses for long-distance reflexives are syntactic, and their basic strategies tend to show that BCA or its revised version still applies to LDAs. For instance, one of the influential proposals is the LF movement theory for LDRs, which was first suggested by Battistella (1989) and subsequently developed in other papers (see Cole et al., 1990, 2001; Cole & Sung, 1994; Cole & Wang, 1996; Huang & Tang, 1991; Li, 1993). The basic idea is that LDAs including ziji in Chinese are head-moved to the T positon at LF, getting their features from the subject through Spec-Head agreement, and thus LD binding is really a local phenomenon fully conforming to BCA. The reason why LDAs have to move is because they are somehow defective in terms of their feature specification. It is assumed that these LDRs must end up with an item in a local configuration which can provide them with the necessary features for their interpretation. Although the LF movement analysis looks appealing in the framework of Generative Grammar, it has encountered a number of problems. For instance, LDA ziji is not sensitive to island conditions (e.g., (4)); blocking effects can be triggered by nonsubjects which in general are not potential antecedents of LDA ziji (e.g., (11)).5 These problems have made people question the movement theory for LDRs, though the latest version, developed by Reuland (2011, 2017) using the Chain Condition, seems to account for the LDRs in European languages, though not for those in Asian languages. Another new syntactic account is provided by Giblin (2016), who argues that LDA ziji is a canonical SE anaphor when syntactic conditions are satisfied.6 However, Sperlich (2019) has recently reviewed major Minimalist theories of anaphora, finding that none of them is able to properly account for the distribution of LDA ziji, hence supporting a theory of ziji that is pragmatic in nature.

In fact, since Sells (1987) points out the relevance of discourse factors in the interpretation of LDAs, many researchers have attempted to account for them from a semantic-pragmatic perspective by assigning certain discourse roles to their antecedents (Anand, 2006; Huang & Liu, 2001; Oshima, 2004; Pan, 1997, 2001; Wang & Pan, 2014, 2015), though some others argue that thematic prominence or a combination of it with subject is the right condition to explain the long distance binding property of reflexives (Chou, 1992; Xu, 1993, 1994). For all of them, LDAs are essentially different from local ones: Although the latter are constrained by BCA, the former are not true anaphors in the sense of Binding Theory. Other proposals of a dual nature for ziji include Chen (1992), Xue et al. (1995), and Yu (1998), though they do not agree on where the boundary is between LDAs and local anaphors. Now the prevailing view is that the dividing line between these two is given by ‘governing category’ (cf. Huang & Liu, 2001). Contrary to this binary view, Hu and Pan (2002) and Hu (2019) believe that a unified account of reflexives can be given at some deeper level. It is claimed that the binding of reflexives, including both the simple and complex ones (see section 4), is actually constrained by the same condition, in term of a well-defined notion of prominence. The prominence of NPs can be computed through a series of hierarchies (e.g., the Empathy Hierarchy, the Animacy Hierarchy, etc.; see Hu, 2019, pp. 174–175). Since different reflexives in various languages may have different definitions of the most prominent NP, their binding domains as well as their possible antecedents turn out to be different from one another. Moreover, a feature characterization of anaphoric expressions is also proposed by Hu (2019) so as to classify anaphors typologically.

As to the semantic-pragmatic account of LDA ziji, Pan (1995, 1997) suggests treating LDA ziji as a de se anaphor based on the notion of de se ascription (Lewis, 1979). Li (1991) argues that it is required to have highly reflective contexts for unbound and LDA ziji, and ziji is constrained by Perspectivity. Later, inspired by the theory of logophoricity (Sells, 1987), Huang and Liu (2001) argue that LDA ziji is a logophor (i.e., its antecedent should be associated with the role of Source, Self, and Pivot).7 Pan (2001) discusses the differences between Logophoricty, Perspectivity, and Self-ascription, and argues that it is neither Logophoricity nor Perspectivity but Self-ascription that can account for the properties of LDA ziji (cf. Pan, 2001, pp. 290–293, 305–307). Recently, based on the theory of linguistic empathy (Kuno, 1987), Wang and Pan (2014, 2015) claim that LDA ziji is an anaphor with an empathy requirement, that is, the speaker is required to emphasize with the referent of ziji.8 However, these theories have their own problems when being used to explain the data concerning the properties of LDA ziji. For instance, Pan’s early work does not concern the occurrences of LDA ziji in extensional contexts where the de se reading is not available. Huang and Liu’s analysis suffers from the inadequacy of Sells’s extended notion of logophoricity which covers the speaker’s point of view (or Pivot; see Cole et al., 2005; Kuno, 1987; Oshima, 2007). One piece of evidence showing that logophoricity and empathy should be kept apart, as mentioned by Oshima (2007), is that in Japanese, logophoric zibun can co-occur with the first person pronoun boku ‘I’, while empathic zibun cannot (see (15)).


In view of the distinction between logophority and empathy, Wang and Pan’s analysis using the notion of empathy looks more promising, as empathic expressions are subject to a series of constraints such as The Ban on Conflicting Empathy Foci, Speech Act Empathy Hierarchy, Topic Empathy Hierarchy, and so on (cf. Kuno, 1987; Kuno & Kaburaki, 1977; Oshima, 2007), which may help explain the blocking effect and other properties of LDA ziji (cf. Wang & Pan, 2020). However, such an analysis is still confronted with the following sticky case where the second person pronoun can induce a blocking effect in a sentence with a first person matrix subject:


As noted by Chou (2012), the difficulty is that the strategy for the explanation of the blocking effect, that is, the prohibition of perspective conflicts (or The Ban on Conflicting Empathy Foci), wrongly predicts the grammatical long-distance construal of ziji in (16): when ziji is bound by the matrix wo ‘I’ that is anchored to the external speaker, there is no perspective conflict with the second person embedded subject ni ‘you’, since the latter is also anchored to the external speaker as a reference.

Recently, inspired by Rizzi (1997), Cinque (1999), Speas and Tenny (2003), and Speas (2004), who argue that there are designated syntactic positions in the left periphery of clauses which are devoted to expressing discourse notions, some researchers propose to enrich the syntactic representation in such a way that the relevant discourse elements like the coordinate of the attitude holder (Giorgi, 2006, 2007) and the speaker’s point of view (Nishigauchi, 2014) can be represented in certain syntactic positions, in order to account for the discourse properties of LDAs.

3. Other Uses of ziji

3.1 The Local Anaphor ziji

As mentioned earlier, ziji can be locally bound just like English reflexives (e.g., ziji in (2) can refer to Bill). However, the distributions of ziji in its local domain (i.e., its governing category) are still different from those of English reflexives. First, local ziji can be used as an attributive modifier in an argument NP (e.g., (17)) or in an adjunct (e.g., (18)). In the former case, it is bound by a co-argument of that NP, while in the latter case, it is bound by an argument outside the adjunct.



Second, local ziji can take a sub-commanding NP as its antecedent if the local subject is inanimate, though it cannot do so if the local subject is animate (e.g., (19a) vs. (19b)).9


According to Tang (1989), the reason why strict c-command is not required for Chinese reflexive ziji is that it is inherently [+animate]. However, Pan (1995, 1997) objects to this view, and shows that the feature of ziji can be [-animate] in cases as in (20):


It seems that the NP c-commanding ziji in (19a) can’t be a potential antecedent just because it is not compatible with ziji in meaning and thus skipped.

Moreover, local ziji has different properties from long-distance ones, which further supports their separate treatments (Huang & Liu, 2001; Pan, 1995, 1997, 2001; among others). For instance, the former does not exhibit any blocking effect (see (21)).


In addition, local ziji normally does not induce the de se reading, or the indirect de se reading (i.e., the speaker empathizes with the referent of ziji and helps the latter do self-ascription; cf. Wang & Pan, 2014), which is always associated with LDAs. Due to this, the occurrences of local ziji are often called syntactic anaphors.

Semantically, besides the normal reading which requires complete identity with its antecedent, simple reflexive ziji (as well as complex ta-ziji) also allows near identity with its antecedent, hence getting the so-called statue reading (Jackendoff, 1992), as in the following sentence (cf. Lidz, 1996; Liu, 2003):


However, according to Liu (2003), such a statue reading is not available for Chinese complex reflexives with the form ziji-benshen ‘self-self’ (see section 4 for more details).

3.2 The Generic Use of ziji

Besides anaphoric uses, ziji can also be used freely as a generic pronoun, just like one in English. For example, ziji , at the subject position of the sentence in (23), is a case in question.


At this point, it is worth noting that not every occurrence of sentence-free ziji is interpreted this way. In certain contexts, it may refer to the speaker (e.g., (24a)), the addressee (e.g., (24b)), or even a third party that is salient in discourse (e.g., (24c)), according to Pan (2001).


In the literature, such instances of ziji are often treated as LDAs, whose antecedents are assumed to be the external Source in Sells’s system (Huang & Liu, 2001) or the empathy locus in Kuno’s system (Wang & Pan, 2015).

3.3 The Intensifier Use of ziji

In Chinese, ziji also has emphatic and contrastive uses. Consider examples (25a,b):


In (25a), ziji is used to emphasize that it is speaker’s personal thought. It can be paraphrased as benren ‘this person’. The sentence in (25b) says that the agents do everything themselves, without delegating to others. In this case, ziji acts as an adverb modifying the verb phrase. As indicated by their translations, English reflexives also have this use. In the literature, these reflexives bearing the intensifying function are often called self-intensifiers to avoid impinging on the territory of degree adverbs like very which are called intensifiers (Gast & Siemund, 2006).

Cross-linguistically, self-intensifiers are generally divided into four types: adnominal intensifiers, adverbial exclusive intensifiers, adverbial inclusive intensifiers, and attributive intensifiers (König, 1991). Ziji, as a self-intensifier, can manifest all of the uses above, with the exception of the adverbial inclusive one.10 The adnominal and adverbial exclusive uses of ziji can be illustrated by examples (25a) and (25b), respectively; an example of the attributive use of ziji, which often requires an additional possessive modifier de, is given in (26):


According to Wang (2011), the distinction between the adnominal use (e.g., (25a)) and adverbial use (e.g., (25b)) of intensifier ziji is often hard to be detected from its syntactic position, as ziji, occurring at the same position, which is often the nonargument position adjacent to a nominal, may exhibit these two uses, unlike the self-intensifiers in English.11 Wang demonstrates that different types of predicates may play a role in determining this: Verbs indicating a state give rise to an adnominal use of ziji (e.g., (27a)), whereas verbs of action indicate an adverbial use (27b).


In some cases like (28) where the relevant verb (e.g., kai ‘open’) is ambiguous between a stative reading and a dynamic reading, intensifier ziji may be seen as an adnominal use or an adverbial use, accordingly.


A further test that can be used to distinguish the two uses of intensifier ziji is negation: Negation can be put in between adverbial ziji and its preceding NP, although this is not possible for adnominal ziji, given that ziji is the appositive of this NP. Moreover, these two uses of intensifiers behave differently in negative sentences (cf. Gast & Siemund, 2006). Concerning the sloppy reading of the possessive his, the sentence in (29) says that only Lisi’s lawn was mowed whereas Zhangsan’s lawn was not. In sentence (30), by contrast, both Zhangsan’s and Lisi’s lawns were mowed, though Zhangsan delegated the job to someone else.



Semantically, adnominal intensifiers express an identity function; that is, they do not contribute to the truth conditions of the relevant sentences. What they do is that they evoke a set of alternatives that have something to do with their referents so that the contrasts are established between them. As to the meaning of the adverbial use of ziji, besides the typical exclusive reading, which can be paraphrased by duzi ‘alone’ (e.g., (31a)) or bu xuyao bangzhu ‘without help’ (e.g., (31b)) just like its English counterparts, ziji has other possible readings.



According to Wang (2011), this reading is somehow weaker and used only in a special context; that is, a person with distinguished social status does something that should not have been done by him- or herself. In addition, Hole (2008, p. 288) mentions another reading of ziji in an adverbial position which stresses that there does not exist an external cause for the change-of-state event it modifies (e.g., (33a)). Similar readings can be observed in the cases with animate subjects who are not aware of such an external cause (e.g., (33b)).


Wang (2011) generalizes such meaning of ziji as ‘without outside force’.12 Moreover, the use of ziji can also mark the volitionality of the subject (or actor, agent), which can be paraphrased by the phrase of one’s accord in English (Tsai, 2002):


3.4 The Construction of ziji + verb/preposition ziji

Finally, it is worth mentioning an idiom involving a pair of ziji, that is, the construction of ‘ziji + verb/preposition ziji’. Such a construction functions as a complex predicate. Sentences with this kind of predicate, (35a) for instance, are often seen as a case of topicalization (see (35b)).


According to Tsai (2012), these two occurrences of ziji can be both analyzed as bound variables, which are bound by the same lambda operator as in (35c), hence getting the same interpretation as its corresponding sentence with a single anaphoric ziji, that is, Zhangsan zhihao anwei ziji ‘Zhangsan has to comfort himself’. This is consistent with the treatments of ziji by Pan (1997) and Huang and Liu (2001) as a variable bound by a Lambda operator. Tsai further notes that this doubling construal is subject to a syntactic constraint, that is, one occurrence of ziji must c-command the other. This constraint can be testified by using (36a) and (36b) below:


4. Other Reflexives in Chinese

Besides ziji, there are other simple reflexives such as zishen ‘self-body’ (e.g., (37a)) and benren ‘person proper’ (e.g., (37b)), though less research has been done on these items expect for Pan (1995, 1997), Liu (2003), and Wang (2011).


Generally speaking, these simple reflexives can be used as intensifiers as well (e.g., (38a–38b)), though the adverbial use is impossible (e.g., (38c)); that is, they are always adnominal intensifiers.


Moreover, there are also complex reflexives in Chinese. Cross-linguistically, the complex type of reflexives is often divided into two subcategories: one is of the form ‘pronoun + simple reflexive’ (e.g., Chinese ta-ziji, Japanese kare-zibun, and Korean ku-casin), and the other is formed by two simple reflexives (e.g., Chinese ziji-benshen, Japanese zibun-zisin, and Korean caki-casi).

As to the first type, the complex reflexives of the form X-ziji in Chinese have more feature specification than simple reflexives due to the contribution of the pronoun in question (e.g., ziji is specified for nothing, whereas ta-ziji is specified for the third person singular feature). It can be used as an anaphor (e.g., (39)) or an adnominal intensifier (e.g., (40)).



Note that the complex reflexives of this type occurring at the object position like (39) can only be treated as anaphors as a whole, but not pronouns plus adnominal intensifier ziji, for the pronoun X in X-ziji in these situations can’t be locally bound by the subject according to Binding Condition B. This is compatible with the generalization by Reuland (2011, 2017) that a complex anaphor is often used to express reflexivity in natural languages to avoid fully identical occurrences of expressions in a local domain (i.e., the so-called protection strategy). As to its adnominal intensifier use, since X-ziji is a combination of two free forms, X can be omitted without changing the meaning of the whole sentence in question, unlike English reflexives which can never be separated into two components.


According to Pan (1998), besides the possibility of being locally bound, the non-intensifier use of ta-ziji can also be bound outside its governing category when the embedded subject is inanimate (e.g., (42a)), though this cannot happen when the embedded subject is animate (e.g., (42b)).


Due to the above contrast, Pan claims that Chinese complex reflexive ta-ziji is still different from its English counterpart himself, contra the previous view that these two are basically the same (Huang, 1983; Huang & Tang, 1991; Tang, 1989).13 Moreover, to explain the binding properties of ta-ziji, Pan and Hu (2003) adopt an Optimality-theoretic (OT) (Prince & Smolensky, 1993) account of reflexivization, and argue that prominence and locality are the two important factors that regulate the interpretation of reflexives in different languages and their different rankings can account for the differences between English and Chinese in reflexive binding. More specifically, in Chinese, the prominence constraint is ranked higher than the locality constraint, whereas in English, the latter is ranked higher than the former. According to their analysis, since the inanimate NP naben shu ‘that book’ in (42a) is less prominent, ta-ziji can skip it and be bound by the more prominent NP John.

As to the second type, Yu (1996, 1998) and Liu (2003) investigate the item ziji-benshen, which can be used as an anaphor as well as a self-intensifier.



According to Liu (2003), compared to other reflexives, what is special for ziji-benshen is that it requires complete identity with its antecedent in the examples like (43). That is, the statue reading, as mentioned earlier in (22) is not available here. Liu argues that this semantic property of the complex anaphor ziji-benshen ‘self-self’ is due to the semantic composition of the near reflexive function of the morpheme ziji ‘self’, the focus function of the morpheme benshen ‘self’ and the operator status of ziji-benshen ‘self-self’. Actually, Pan (1995, 1997) devotes one whole chapter (Chapter 7) to discuss the properties of reflexives like benren ‘person proper’, benshen ‘self-body’, and their compound forms. He considers them different from ziji and its compound forms, as they are contrastive in nature and thus subject to discourse conditions like Prominence, or conditions for the intensifier use of object himself in English, as discussed in Baker (1995).

5. More Questions Ahead

Chinese reflexives reflect the diversity of the patterns for binding, intensification, and other relevant phenomena. Up to now, it seems that there are no definite answers to the following questions: (a) Is it possible to interpret Chinese LDAs like ziji and ta-ziji using some principles in Universal Grammar (UG) plus properties of the vocabulary of the language? The updated version of UG on binding such as the one given by Reuland (2011) rejects CBT as part of UG and presents a detailed proposal of how locality conditions on binding can be derived from more primitive notions. Based on these new insights, some scholars like Reuland himself as well as Giblin (2016) try to say that the answer is ‘yes’, though this answer is still doubtful for others, given the mysterious properties such as the blocking effects concerning simple reflexive ziji and the long-distance binding of complex reflexive ta-ziji. (b) Is there any more fundamental explanation for the contrast between English and Chinese reflexives? Cheng (1994) analyzes the formation of ziji from a diachronic perspective and proposes that zi ‘self’ and ji ‘self’—the two different free morphemes in ancient Chinese are responsible for the special properties of ziji in Modern Chinese. Recently, Jia (2020) has discussed the contrasts of the conceptual basis of English and Chinese reflexives and argues that the different relationships between the Subject (the locus of consciousness) and the Self (the objective aspect of a person) embodied in these two languages may explain their differences. (c) How are the different uses of Chinese reflexives, especially ziji, related to each other? Wu (2018) adopts a perspective from linguistic typology to understand forms and meanings of self including Chinese ziji and believes that by means of the inventory approach, one could reveal the interworkings between the forms and meanings of ziji with a hierarchical structure. Together with the research conducted from perspectives other than formal linguistics we may find a way to reveal the myth of Chinese reflexives.

Further Reading

  • Charnavel, I., Huang, C.-T. J., Cole, P., & Hermon, G. (2017). Long-distance Anaphora. In M. Everaert& H. van Riemsdijk (Eds.), The Wiley Blackwell companion to syntax, vol. IV (2nd ed., pp. 2321–2402). Hoboken, NJ: John Wiley & Sons.
  • Reuland, E. (2017a). Logophoricity. In M. Everaert& H. van Riemsdijk (Eds.), The Wiley Blackwell companion to syntax, vol. IV (pp. 2249–2269). Hoboken, NJ: John Wiley & Sons.
  • Reuland, E. (2017b). Long-distance binding in Germanic languages. In M. Everaert& H. van Riemsdijk (Eds.), The Wiley Blackwell companion to syntax, vol. IV (2nd ed., pp. 2403–2432). Hoboken, NJ: John Wiley & Sons.


  • Anand, P. (2006). De se (Unpublished doctoral dissertation). MIT, Cambridge, MA.
  • Baker, C. L. (1995). Contrast, discourse prominence, and intensification, with special reference to locally free reflexives in British English. Language, 71(1), 63–101.
  • Battistella, E. (1989). Chinese reflexivization: A movement to INFL approach. Linguistics, 27(6), 987–1012.
  • Castañeda, H.-N. (1966). “He”: A study in the logic of self-consciousness. Ratio, 8(2), 130–157.
  • Castañeda, H.-N. (1968). On the logic of attributions of self-knowledge to others. Journal of Philosophy, 65(15), 439–456.
  • Chen, P. (1992). The reflexive ziji in Chinese: Functional vs. formalist approaches. In T. Lee (Ed.), Research on Chinese linguistics in Hong Kong (pp. 1–36). Hong Kong: The Linguistic Society of Hong Kong.
  • Cheng, G. (1994). Remarks on Neo-Gricean pragmatic theory of anaphora—A case study of Chinese ziji from a diachronic perspective. Journal of PLA University of Foreign Language, 71, 10–22.
  • Chomsky, N. (1981). Lectures on government and binding. Dordrecht, The Netherlands: Foris.
  • Chou, C.-T. T. (2012). Syntax-pragmatics interface: Mandarin Chinese wh-the-hell and point-of-view operator. Syntax, 15(1), 1–24.
  • Chou, X.-L. (1992). An alternative approach to Chinese reflexives. MA report, the University of Texas, Austin.
  • Cinque, G. (1999). Adverbs and functional heads: A cross-linguistic perspective. Oxford, UK: Oxford University Press.
  • Cole, P., Hermon, G., & Huang, C.-T. J. (2005). Long distance anaphors: An Asian perspective. In M. Everaert& H. van Riemsdijk (Eds.), The Blackwell companion to syntax (pp. 21–84). Malden, MA: Blackwell.
  • Cole, P., Hermon, G., & Lee, C.-L. (2001). Grammatical and discourse conditions on long-distance reflexives in two Chinese dialects. In P. Cole, G. Hermon, & C.-T. J. Huang (Eds.), Long-distance reflexives (pp. 1–46). New York, NY: Academic Press.
  • Cole, P., Hermon, G., & Sung, L.-M. (1990). Principles and parameters of long-distance reflexives. Linguistic Inquiry, 21(1), 1–22.
  • Cole, P., & Sung, L.-M. (1994). Head movement and long-distance reflexives. Linguistic Inquiry, 25(3), 355–406.
  • Cole, P., & Wang, C. (1996). Antecedents and blockers of long-distance reflexives: The case of Chinese ziji. Linguistic Inquiry, 27(3), 357–390.
  • Gast, V., & Siemund, P. (2006). Rethinking the relationship between SELF-intensifiers and reflexives. Linguistics, 44(2), 343–381.
  • Giblin, I. (2016). Agreement restrictions in Mandarin long-distance binding (Unpublished doctoral dissertation). MIT, Cambridge, MA.
  • Giorgi, A. (2006). From temporal anchoring to long distance anaphors. Natural Language and Linguistic Theory, 24(4), 1009–1047.
  • Giorgi, A. (2007). On the nature of long-distance anaphors. Linguistic Inquiry, 38(2), 321–342.
  • Han, C.-H., & Storoshenko, D. R. (2012). Semantic binding of long-distance anaphor caki in Korean. Language, 88(4), 764–790.
  • Hole, D. (2008). Focus on identity—The dark side of zijĭ. The Linguistic Review, 25(3–4), 267–295.
  • Hu, J. (2019). Prominence and lociality in grammar: The syntax and semantics of Wh-questions and reflexives. New York: Routledge.
  • Hu, J., & Pan, H. (2002). NP prominence and the Chinese reflexive ziji. Contemporary Linguistics, 4(1), 46–60.
  • Huang, C.-T. J. (1983). A note on the binding theory. Linguistic Inquiry, 14, 554–561.
  • Huang, C.-T. J., & Liu, C.-S. L. (2001). Logophoricity, attitudes, and ziji at the interface. In P. Cole, G. Hermon, & C.-T. J. Huang (Eds.), Long-distance reflexives (pp. 141–195). New York, NY: Academic Press.
  • Huang, C.-T. J., & Tang, C.-C. J. (1991). The local nature of the long-distance reflexive in Chinese. In J. Koster& E. Reuland (Eds.), Long-distance anaphora (pp. 263–282). Cambridge, UK: Cambridge University Press.
  • Jackendoff, R. (1992). Madame Tussaud meets the binding theory. Natural Language and Linguistic Theory, 10(1), 1–31.
  • Jia, G. (2020). A contrastive study of the conceptual basis for reflexives in English and Chinese. Foreign Languages and their Teaching, 311, 60–68.
  • König, E. (1991). The meaning of focus particles: A comparative perspective. London, UK: Routledge.
  • Kuno, S. (1987). Functional syntax: Anaphora, discourse and empathy. Chicago, IL: University of Chicago Press.
  • Kuno, S., & Kaburaki, E. (1977). Empathy and syntax. Linguistic Inquiry, 8(4), 627–672.
  • Lewis, D. (1979). Attitude de dicto and de se. The Philosophical Review, 88(4), 513–543.
  • Li, N. C. (1991). Perspective-taking in Mandarin Discourse (Unpublished Ph.D dissertation). State University of New York, Buffalo.
  • Li, Y. (1993). What makes long-distance reflexives possible? Journal of East Asian Linguistics, 2(2), 135–166.
  • Lidz, J. L. (1996). Dimensions of reflexivity (Unpublished doctoral dissertation). University of Delaware, Newark.
  • Liu, C.-S. L. (2003). Pure reflexivity, pure identity, focus and Chinese ziji-benshen. Journal of East Asian Linguistics, 12(1), 19–58.
  • Nishigauchi, T. (2014). Reflexive binding: Awareness and empathy from a syntactic point of view. Journal of East Asian Linguistics, 23(2), 157–206.
  • Oshima, D. Y. (2004). Zibun revisited: Empathy, logophoricity, and binding. University of Washington Working Papers in Linguistics, 23, 175–190.
  • Oshima, D. Y. (2007). On empathic and logophoric binding. Research on Language and Computation, 5(1), 19–35.
  • Pan, H. (1995). Locality, self-ascription, discourse prominence, and Mandarin reflexives (Unpublished doctoral dissertation). University of Texas, Austin.
  • Pan, H. (1997). Constraints on reflexivization in Mandarin Chinese. New York, NY: Garland.
  • Pan, H. (1998). Closeness, prominence, and binding theory. Natural Language & Linguistic Theory, 16, 771–815.
  • Pan, H. (2001). Why the blocking effect? In P. Cole, G. Hermon, & C.-T. J. Huang (Eds.), Long-distance reflexives (pp. 279–316). New York, NY: Academic Press.
  • Pan, H., & Hu, J. (2003). Prominence and locality in the binding of Mandarin complex reflexive ta-ziji ‘s/he-self’ (Edited by A. B. Bodomo & K. K. Luke, Lexical-Functional Grammar Analysis of Chinese). Monograph Series of Journal of Chinese Linguistics, 19, 152–170.
  • Pollard, C., & Xue, P. (2001). Syntactic and non-syntactic constraints on long-distance binding. In P. Cole, G. Hermon, & C.-T. J. Huang (Eds.), Long-distance reflexives (pp. 317–342). New York, NY: Academic Press.
  • Prince, A., & Smolensky, P. (1993). Optimality theory: Constraint interaction in generative grammar. Malden, MA: Blackwell.
  • Reuland, E. (2011). Anaphora and language design. Cambridge, MA: MIT Press.
  • Reuland, E. (2017). Grammar of binding in the languages of the world: Unity versus diversity. Cognition, 168(2), 370–379.
  • Rizzi, L. (1997). The fine structure of the left periphery. In L. Haegeman (Ed.), Elements of grammar: Handbook in generative syntax (pp. 281–337). Dordrecht, The Netherlands: Kluwer.
  • Rooryck, J., & Wyngaerd, G. V. (2011). Dissolving binding theory. Oxford, UK: Oxford University Press.
  • Schäfer, F. (2007). By itself. Unpublished manuscript, University of Stuttgart, Germany.
  • Sells, P. (1987). Aspects of logophoricity. Linguistic Inquiry, 18(3), 445–479.
  • Siemund, P. (2000). Intensifiers: A comparison of English and German. London, UK: Routledge.
  • Speas, M. (2004). Evidentiality, logophoricity and the syntactic representation of pragmatic features. Lingua, 114(3), 255–276.
  • Speas, P., & Tenny, C. (2003). Configurational properties and point-of-view roles. In A. M. Di Sciullo (Ed.), Asymmetry in grammar: Volume 1: Syntax and semantics (pp. 344–351). Amsterdam, The Netherlands: John Benjamins.
  • Sperlich, D. (2019). Pragmatic and syntactic theories of Chinese reflexives. Lingua, 221, 22–36.
  • Tang, C.-C. J. (1989). Chinese reflexives. Natural Language and Linguistic Theory, 7, 93–121.
  • Tsai, W.-T. D. (2002). Self, selfness and nature: On reflexive adverbials in Chinese. Studies of the Chinese Language, 289, 357–362.
  • Tsai, W.-T. D. (2012). On reflexive doubling. Studies of the Chinese Language, 349, 329–337.
  • Wang, L. (2011). Intensifiers and reflexive pronouns in English and Mandarin Chinese (Unpublished doctoral dissertation). Freie Universitaet Berlin, Germany.
  • Wang, Y., & Pan, H. (2014). A note on the interpretation of non-de se attitude report. Language, 90(3), 746–754.
  • Wang, Y., & Pan, H. (2015). Empathy and Chinese long-distance reflexive ziji: Remarks on Giorgi (2006, 2007). Natural Language and Linguistic Theory, 33(1), 307–322.
  • Wang, Y., & Pan, H. (2020). The blocking effect revisited. Journal of Foreign Languages, 43(2), 9–19.
  • Wu, J. (2018). Forms and meanings of self: Perspectives from linguistic typology. Journal of Foreign Languages, 41(3), 4–10.
  • Xu, L. J. (1993). The long distance binding of ziji. Journal of Chinese Linguistics, 21, 123–141.
  • Xu, L. J. (1994). The antecedent of ziji. Journal of Chinese Linguistics, 22, 115–137.
  • Xue, P., Pollard, C. J., & Sag, I. A. (1995). A new perspective on Chinese reflexive ziji. In R. Aranovich (Ed.), Proceedings of the 13th West Coast Conference on Formal Linguistics (pp. 432–447). Stanford, CA: CLSI.
  • Yu, X.-F. W. (1996). A study of Chinese Reflexives (Unpublished doctoral dissertation). University of London, UK.
  • Yu, X.-F. W. (1998). Theta role assignment, internal structures of reflexives and reflexive binding. The Linguistic Review, 15(4), 389–405.


  • 1. Governing category is defined by Chomsky (1981, p. 211) as follows: β‎ is a governing category for α‎ iff β‎ is the minimal category containing α‎, a governor of α‎, and a SUBJECT accessible to α‎.

  • 2. That is, the referent of the antecedent of LDA ziji must be aware that the sentence in question is a description of an event in which he himself is a protagonist.

  • 3. Abbreviations used in glosses include: ACC: accusative case, ASP: aspect marker, BA: particle in the BA-construction, BEI: passive marker bei, CL: classifier, DE: prenominal modification marker or postverbal resultative marker de, PRF: perfective marker, Q: question particle, TOP: topic marker, NOM: nominative case, PAST: past tense, PRES: present tense, ZHE: durative aspect marker, PRT: particle.

  • 4. See Pan (2001, pp. 305–306) for a similar view.

  • 5. In the various LF movement accounts, only c-commanding and sub-commanding subjects can induce the blocking effect (cf. Battistella, 1989; Cole & Sung, 1994; Huang & Tang, 1991).

  • 6 According to Reuland (2011, p. 47), a SE anaphor is a non-clitic pronominal that is underspecified in Φ‎ features.

  • 7. According to Sells (1987), the Source is the individual in a given situation who makes the report, the Self represents the one whose mental state or consciousness is being reported, and the Pivot is the one from whose standpoint the report is made.

  • 8. In an intensional context, besides being an empathy locus, LDA ziji may be a de se anaphor at the same time. In this respect, it is different from LDA zibun in Japanese, as the latter is a pure de se anaphor (or logophor) in such a context, as noted by Oshima (2007).

  • 9. Roughly, this means that ziji can be bound by the specifier of the local subject. Sub-command relaxes the c-command requirement between the reflexive and its antecedent by allowing the specifier of a c-commanding nominal to function as a possible antecedent when the head nominal is not animate.

  • 10. Adverbial inclusive intensifiers can be exemplified by English reflexives as follows:


    How can Jim complain about Jack’s snoring if he snores himself? = … if he snores, too

    But English does not use reflexives as attributive intensifiers. It uses the word own instead.

  • 11. In English, the adverbial use tends to follow the VP (e.g., I will do that myself), while the adnominal use may precede the VP (e.g., The president himself opened the meeting) (cf. Siemund, 2000). Besides, a piece of strong evidence supporting the division of adnominal intensifiers and adverbial intensifiers is that in many languages, there are different lexical items to express these two types of self-intensifiers (e.g., in Tetelcingo Nahuatl, sie + pronoun is used as an adnominal intensifier, while sa-siel is used as an actor-oriented intensifier). See Gast and Siemund (2006).

  • 12. In English, such a meaning can be expressed by the phrase “by itself” when it has an inanimate antecedent. A case in point given by Schäfer (2007) is in (ii):


    300 million years ago the climate became already warmer by itself and without human intervention. Why should this time humans be the cause?

    According to Schäfer, by itself used in such cases denies that there exists a cause for the change-of-state event it modifies.

  • 13. Baker (1995) pointed out that reflexives in English sometimes can be locally free as well. However, such a use is best analyzed as intensified non-nominative pronouns, subjects to certain conditions that regulate English intensive NPs generally.