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: 25 July 2021

Sentence-Final Particles in Chinesefree

Sentence-Final Particles in Chinesefree

  • Victor Junnan PanVictor Junnan PanChinese University of Hong Kong

Summary

Chinese has a rich system of Sentence-Final Particles (SFPs). Traditional grammar and descriptive linguistic studies attempt to capture the precise semantic interpretation and the discourse function of each particle. Much work related to this aspect tries to find out what the core semantic interpretation of a given SFP is, how the diverse interpretations of a given SFP are developed from its core interpretation, and in what context the use of a given SFP is licit. Linguists from different disciplines have made important observations and offered various explanations. On the other hand, diachronic studies trace the origin and the evolution of each SFP, which helps understand the core semantics of SFPs in modern Chinese. Studies on different Chinese dialects also help the understanding of the meaning and the function of SFPs from a comparative perspective. Under the generative framework, SFPs are analyzed as complementizers, which are located in the peripheral domain. Both traditional grammarians and generative syntacticians are interested in patterns like the rigid order that necessarily shows whenever SFPs co-occur. They attempt to establish the hierarchical order of SFPs and identify the general principle that regulates such an order. Recent studies show that such an order is regulated by a discourse constraint related to subjectivity, according to which the higher a functional projection is located, the more directly it is for such a projection to be linked to the speaker’s attitude, the more subjective the interpretation of such a projection becomes, and the less likely it is for such a projection to be embedded. This constraint offers an explanation to the question of why only some SFPs can appear in embedded clauses whereas the others demonstrate root properties. Syntacticians are also interested in the question of how to derive the final order of SFPs. Two analyses are available: disjunction analysis and complement-to-specifier raising analysis. A more recent finding is that under the minimalist framework, each SFP heads a phase and bears an EPP feature. Complement-to-specifier raising is required as a last resort to satisfy the Extended Projection Principle (EPP). The complement of an SFP is moved to the phase edge to postpone the transfer of the phrases that are embedded within the complement, which allows these phrases to be extracted later.

Subjects

  • Syntax

1. Introduction

Much descriptive work has been done since early grammar books on the rich system of Sentence-Final Particles (SFPs) in Chinese (cf. Chao, 1968; Li & Thompson, 1981; Zhu, 1982; among others). Over the last 50 years, scholars have attempted to describe the precise interpretation and the discourse function for each SFP, as well as the contexts in which the use of each SFP is licit. Diachronic studies help trace the origin and evolution of SFPs. Studies on different Chinese dialects also help understand the meaning and the function of SFPs in Mandarin from a comparative perspective. Although much progress has been made, there are still many SFPs whose core semantics and discourse functions are not explicit. Thus, future work from the descriptive perspectives is still needed. On the syntactic side, scholars are interested in questions like how to analyze SFPs, which are treated as functional heads in the generative tradition. Concretely, they are analyzed as complementizers, which head phrases equivalent to CP (Complementizer Phrase). On the other hand, Zhu (1982) observes that several SFPs could co-occur but require a fixed rigid order. Inspired by the split-CP hypothesis (cf. Rizzi, 1997), syntacticians attempt to establish a map as detailed as possible to determine the hierarchical order of SFPs and also try to find out the general principle that regulates such an order (see Li, 2006; Pan, 2015, 2019a; Pan & Paul, 2016; Paul, 2014, 2015; Paul & Pan, 2017; Tang, 2015, 2019, 2020; among others).

This article reviews some of these aspects of the research on SFPs in Mandarin Chinese. Section 2 addresses questions related to the (non-)optionality of SFPs; section 3 discusses the diachronic studies of some SFPs; section 4 presents the hierarchical order of SFPs; section 5 discusses root phenomena of some SFPs; section 6 addresses the head-finality of SFPs; section 7 presents the latest analysis of SFPs under the Minimalist Program; and section 8 concludes the article.

2. Nonoptionality

Although the presence or the absence of an SFP does not always affect the grammaticality of a given sentence, the presence of an SFP is not optional. This is because each particle conveys a specific meaning or has a specific discourse function; in other words, a specific semantic interpretation or a specific discourse function can only be obtained when the correct particle is used. In this sense, the presence of a particle is obligatory for the specific meaning associated with this SFP to be expressed. For instance, without any SFP, (1a) only states a fact. The particle ma (‎) transforms (1a) into a yes-no question, as shown in (1b). The confirmation question particle baconf (‎) in (1c) gives rise to a tag-question reading. The SFP neatt (‎) in (1d) serves to draw the attention of the co-speaker to the fact stated in the sentence, which has a function similar to hey, look, listen in English (see Chu, 2002; Jiang, 1986; Jin, 1996; Li, 2006; Qi, 2002a, 2002b, 2002c; Ren, 2017; and Wu, 2005; among others, for more detailed discussions on ne). The SFP baatt (‎) in (1e) expresses the speaker’s uncertainty about the fact stated in the sentence, which is translated as probably in English. The interjective particles such as a (‎) and la (‎) in (1f) express the mood of the speaker, which can be surprising, exciting, and so on (also see Chu, 2002). SFPs such as maatt (‎) in (1g) and bei (‎) in (1h) both emphasize the obviousness of the fact that the sentence states, but with different implications. See Cui (2019, 2020) for detailed discussions on the discourse function of maatt (‎).

(1)

The fact that a rising intonation applied to a declarative sentence sometimes gives rise to a yes-no question reading leads some scholars to suggest that the presence of the yes-no question particle ma is optional in a given sentence. However, a yes-no question with ma and a yes-no question with a rising intonation are not interchangeable (see Gunlogson, 2001, for detailed discussions on English yes-no questions; see Pan & Paul, 2016, for the discussion on Chinese ma). For instance, the indefinite reading of a wh-phrase in Chinese can be licensed in a yes-no question with ma only, as in (2a). (2b) shows that rising intonation cannot license the indefinite reading of the wh-phrase shénme ‘what’. Importantly, (2b) can only be interpreted as a root wh-question with the interrogative reading of shénme ‘what’.

(2)

3. Diachronic Studies

Although the semantic interpretation is clear for particles such as ma, such is not the case for all the SFPs. Linguists have attempted to give detailed descriptions of the semantics and the discourse function of each SFP. In this respect, diachronic research helps us trace the origin and the evolution of SFPs, to better understand their discourse functions in modern Chinese. In this section, we review the diachronic studies of the two most important SFPs: ma and ne.

3.1 Ma (‎)

One of the most studied Chinese SFPs is the yes-no particle ma, which turns a declarative sentence into a yes-no question. It is generally agreed that ma comes from negative words such as wu. A general grammaticalization path for the SFP ma is as follows.

According to Yang (2003), wu (‎) was used as a negative word, but also participated in the form [VP + NEG] to form a yes-no question, as shown in (3)(4).1

(3)

(4)

The negative wu (‎) or mo (磨/摩‎) was later written as me (‎) in Song dynasty, as in (5)(6).

(5)

(6)

Finally, the negative wu (‎) or mo (磨/摩‎) has been written as ma (‎) since the Qing dynasty, as in (7).

(7)

In modern Chinese, it is sometimes written as me (麽/么‎), as shown in (8).

(8)

Similar cases are found with modern Chinese. For example, in (9), (‎) is a common negative adverb located in a preverbal and postsubject position.

(9)

(‎) can also be used as an SFP to transform a declarative sentence into a yes-no question, as in (10).

(10)

Such phenomena are by no means isolated in modern Chinese. In fact, in a very early period, the negative word fǒu (written as ‎) has already been used as an SFP to indicate a yes-no question, as in (11).

(11)

It is convincing to many that the yes-no question particle ma and its variant me are related to the negative words in ancient Chinese. The reader can also refer to Ota (2003), Wang (1980), Yang (2003), and Zhong (1997) for more detailed discussions.

3.2 Ne (‎)

Generally, three kinds of ne have been identified in modern Chinese: the first indicates the progressive aspect, glossed as “neprog” (cf. (12)), the second can be used in interrogative sentences, glossed as “neint” (cf. (13)), and the third is used in exclamative sentences to express the speaker’s subjective opinion and attitude, glossed as “neatt” (cf. (14)).3

(12)

(13)

(14)

Historically, the grammaticalization path for the interrogative neint is clear:

Examples in (15)(16) are from Zutang ji during the Five Dynasties period, and ni (聻/尼‎) is used.

(15)

(16)

Example (17) is from the Song dynasty and na (‎) is used.

(17)

During and after the Yuan dynasty, li (‎) is used, as in (18)(19).

(18)

(19)

The grammaticalization path for the exclamative neatt is as follows:

(20)

Importantly, since the Yuan dynasty, li (‎) has been used both as an interrogative particle and as an interjective particle. Here are some examples.

(21)

(22)

(23)

Ne (‎) has appeared since the Qing dynasty. See Cao (1986), L. Jiang (1986), S. Jiang (2005), Ota (2003), Qi (2002a, 2002b, 2002c), Sun (1992), and Wang (1980) for detailed discussions and controversial issues concerning the origin and the evolution of the two ne particles.

4. Hierarchical Order and Co-occurrence

SFPs in Chinese can co-occur. Zhu (1982) identifies three classes of SFPs occurring with a fixed order; more recent work on the occurrence of SFPs with evidence from Chinese dialects can be found in Wang and Bi (2018). Under the generative framework, Lee (1986) analyzes the yes-no question particle ma as a complementizer (i.e., C head), which takes a tense phrase (TP) as its complement. Based on the split-CP hypothesis (cf. Rizzi, 1997), Paul (2014, 2015) extends this analysis to all the SFPs in Chinese and maps the SFPs from the three classes identified by Zhu (1982) onto three functional projections: low C < medium C (Force) < high C (Attitude). Pan (2015, 2019a, 2019b) proposes a more fine-grained architecture of the entire peripheral domain in Chinese, containing not only SFPs but also other peripheral functional projections.

(24)

Overt particles occupy four layers: S.AspP, OnlyP, iForceP and AttPs. Table 1 is extracted from Pan (2019a), which gives an overview of the distribution of SFPs in Chinese.

Table 1. Distribution of SFPs in Chinese from Pan (2019a)

Projections

Particles/Operators

Discourse Function

Root Phenomena?

S.AspP

(sentential aspect)

来着láizheAsp

Recent past

No

le

State changing

No

neprog

Progressive aspect

No

OnlyP

而已éryǐ

Sentential exclusive focus

No

iForceP

(illocutionary force)

ma

Standard yes-no question

Yes

baimp

Weak imperative

Yes

baconf

Confirmation yes-no question

Yes

AttitudeP (speaker’s attitude)

low layer

neatt

Speaker’s attitude,

subjective opinion, etc.

Yes

high

layer

a,哎ei,呗bei,啦la,嘞lei

na,呀ya,嘛ma,

来着laizheatt

baatt, etc.

Importantly, SFPs from different projections can co-occur but only with a rigid order, as indicated in (24). In (25), neprog is a sentential progressive aspect particle located at S.AspP and ma is a yes-no question particle located at iForceP. The fact that the entire sentence is interpreted as a root yes-no question suggests that ma takes a wide scope, which is coherent with the fact that ma is located in the highest position in this sentence.

(25)

In (26), the weak imperative particle baimp is located at iForceP and the interjective particle a is located at AttP. AttP-a takes scope over iForceP-baimp.

(26)

In (27), both neatt and maatt are interjective particles conveying the speaker’s subjective opinion and attitude; they occupy two different layers of AttP. The particle neatt is used to draw the attention of the co-speaker. The particle maatt is syntactically higher than neatt and has a wide scope and maatt gives rise to the implication “Please be patient!,” as indicated in the translation of the sentence. The reader can refer to Cui (2019, 2020) for the discussion on the discourse function of maatt in modern Chinese.

(27)

Similarly, in (28), the particle baatt is interpreted as “probably” and it takes scope over the entire sentence.

(28)

(29) demonstrates a case where three SFPs co-occur in the same sentence. The sentential aspect SFP le takes a narrow scope, the exclusive focus SFP éryǐ which is interpreted as “it is just the case that. . .” takes intermediate scope and the attitude SFP baatt takes the widest scope.

(29)

Table 1 identifies two ne (neprog, neatt) and three ba (baimp, baconf, baatt), which are located in different layers. A sentence with a co-occurrence of [ne ba] is several ways ambiguous, as shown in (30). The possible combinations are indicated in table 2.

(30)

Table 2. Combination of ne and ba

S.AspP

iForceP

AttP1

AttP2

(30a)

ne-progressive

ba-confirmation question

(30b)

ne-progressive

ba-probability

(30c)

ne-attention drawing

ba-probability

When ba is analyzed as the confirmation question particle baconf located at iForceP, ne can only be analyzed as the sentential progressive particle neprog located at S.AspP, as shown in (30a). In this case, the predicate make joke is interpreted with a progressive aspect and baconf is interpreted as a tag question. When ba is analyzed as the attitude particle baatt conveying an uncertainty, which is located at the higher layer of AttP (i.e., AttP2), ne can be analyzed as either a progressive particle neprog at S.AspP or an attitude particle neatt, which is located at the lower layer of AttP (i.e., AttP1), as shown in (30b) and (30c), respectively. In both (30b) and (30c), the uncertainty particle baatt is translated as “probably,” which takes scope over the entire sentence. In (30b), neprog denotes a progressive aspectual reading and in (30c), neatt is translated as “look” which is used to draw the attention of the co-speaker.

The hierarchy proposed by Pan (2015, 2019a) has also been observed in archaic Chinese. The SFP (‎) is analyzed as an assertive particle in copular sentences, which can head a FiniteP à la Rizzi (1997), as shown in (31a). The particle (‎) is an interrogative particle and it transforms a declarative sentence into a yes-no question, as shown in (31b). (31b) and (31c) have the same word order; however, (31c) has a rhetorical question reading. This shows that behaves similarly to the yes-no question particle ma in modern Chinese. According to the system of Pan (2015, 2019a), a negative operator which heads a Special Question Phrase (SQP) takes scope over the entire question and gives rise to a strong assertion reading. The particle zāi (‎) is an interjective particle which expresses the speaker’s mood and attitude, which heads an AttP, as shown in (31d).

(31)

(32) is another example with the same order: TP < FiniteP () < iForceP () < SQP (¬) < AttP (zāi).

(32)

A partial hierarchy can be proposed for Old Chinese at this stage. More fine-grained analyses of the entire array of SFPs in old Chinese are still called for.

(33)

5. Embeddability

A very important question is what factors determine the rigid syntactic hierarchical order of functional projections in the left-periphery in Chinese. Pan (2015, 2019a) proposes that this order is correlated with a discourse constraint, the “Subjectivity Scale Constraint.”

(34)

This constraint provides us with a possible way to study the correlation between syntax and discourse. Higher particles are directly related to the subjective opinion and attitude of the speaker, and they can only be used in direct speech, which is why they show root properties. By contrast, lower particles are related to the sentence subject and they can be used in embedded clauses and thus can be used in indirect speech. For instance, (35) shows that when the final particle le takes scope over the negative predicate bù xué gāngqín ‘does not learn playing piano’, an implication such that ‘Zhangsan did learn playing piano before’ is available. The English translation of the pattern “neg < le” is “no longer/no more.”

(35)

Let us examine (36). The final particle le can either be parsed with the embedded predicate, as in (36a), or with the matrix predicate, as in (36b). In the former case, the no-longer reading is only available with the embedded predicate learns playing piano and in the latter case, such a reading is only available with the matrix predicate believe.

(36)

Illocutionary force particles, such as the yes-no question particle ma and the imperative particle baimp and the confirmation question particle baconf, are generally excluded from embedded clauses. According to (34), these particles are located in the relatively high position in the hierarchy of the left-periphery of Chinese (also cf. (24)). Here is an example with ma.

(37)

Attitude particles, such as neatt, which draws the attention of the co-speaker, are also excluded from embedded clauses. According to (34), these particles are located in the highest position in the hierarchy of the left-periphery of Chinese (also cf. (24)). They are directly linked to the speaker’s subjective attitude and opinion, which can only be conveyed in direct speech through a root clause.

(38)

Recall that two kinds of láizhe have been identified: the lower one located at S.AspP, which is related to sentential aspect, and the higher one located at AttP, which is related to the speaker’s opinion and attitude. (39) shows that the lower aspectual láizheAsp can be embedded, and (40) shows that the higher attitude láizheatt cannot be embedded. This contrast is captured by (34) in that the highest attitude-related particles express the speaker’s subjective opinion and attitude, which can only be expressed through root clause. By contrast, lower particles are closely related to the TP and are not linked to the speaker’s attitude, which is why they can appear in embedded clauses.

(39)

(40)

6. Head-Finality

Under the view of the existence of a head parameter, initial heads and final heads coexist. An initial head takes its complement on the right side, whereas a final head takes its complement on the left side. Languages like Japanese are consistent head-final languages. Chinese has both a head-initial order and a head-final order: VP and TP have initial heads, whereas NP and CP headed by the complementizer de have a final order. In (41), the matrix T takes the VP as its complement on the right side; V-know takes the complex NP as its complement on the right side. By contrast, the N head shìr ‘thing’ takes its complement clause CP headed by de on the left side and the complementizer de takes its complement TP also on the left side.

(41)

Under the split-CP hypothesis, some peripheral projections, such as TopicP, have an initial order, whereas the others, such as those headed by SFPs, have a final order. Adopting the head parameter, the final order is base-generated. Another possible view is that the final order is derived. This section discusses two existing approaches to derive the final order of SFPs.

6.1 Disjunction-Based Analyses

Diachronically, the yes-no question particle ma comes from the negative word wu in old Chinese. This leads some scholars to analyze the yes-no question particle as a disjunctive operator, which is the equivalent of or not in English (see Bailey, 2012; Tang, 2015; among others). The disjunctive head (i.e., or-not) takes two identical TPs in the specifier position and in the complement position, respectively. Then, the lower TP (in the complement position) is deleted, which gives rise to the apparent final position of the SFP.

(42)

It is somehow reasonable to treat the yes-no question particle ma as a disjunctive head based on the semantic consideration. However, it is rather difficult to uniformly treat all the SFPs, which bear different discourse functions, as disjunctive heads. For instance, an interjective particle, such as a, bei, or la, cannot be analyzed as a disjunctive head. Pan and Paul (2016) also point out that the real disjunctive word háishì in Chinese, which can only be used in disjunctive questions, does not exhibit syntactic properties of the yes-no question particle ma. Namely, háishì cannot stand in the sentence-final position. In (43), the second conjunct TP in a question with háishì ‘or’ cannot be deleted.

(43)

6.2 Comp-to-Spec Raising Analyses

Another possibility to derive an apparent final order of SFPs is to raise the complement TP to the specifier of the C that hosts an SFP (cf. Hsieh & Sybesma, 2011; Julien, 2002; Pan, forthcoming; Simpson & Wu, 2002; Sybesma, 1999; Takita, 2009; Tang, 1998; among others).

(44)

The above scholars generally agree with the idea of complement-to-specifier raising, but their analyses differ in the motivation for such a raising. For instance, analysis by Tang (1998) is based on the Linear Correspondence Axiom (LCA) (cf. Kayne, 1994). (45) is a simplified version of LCA.

(45)

After an SFP merges with its complement TP, the TP undergoes movement to a position asymmetrically c-commanding the SFP. As a result, the TP is pronounced preceding the SFP, which gives rise to the final order of SFP.

(46) is an example involving three SFPs. To derive the final order, we need to apply Kayne’s “roll-up” movement, as demonstrated in (47).

(46)

(47)

First, the TP she only resigned is moved from the complement of the S.Asp head le to the Spec of S.AspP. Since TP asymmetrically c-commands le, TP is pronounced preceding le, which gives rise to the order: TP < le. Second, the S.AspP is moved from the complement of the Only head éryǐ to the Spec of OnlyP to derive the order TP < le < éryǐ. Third, the OnlyP is moved from the complement of the Att head baatt to the Spec of AttP to derive the order TP < le < éryǐ < baatt.

Pan (2019a) discusses the advantages of the comp-to-spec raising analysis over disjunction analysis. Here is one advantage. Huang (1982) shows that the yes-no question particle ma triggers the existential closure at I’/T’ level in Chinese. In (48), the wh-object gets an existential reading in a yes-no question.

(48)

This phenomenon cannot be captured under the disjunction analysis of ma. The derivation goes as follows.

Step 1: The disjunctive head ma takes the TP1 as its complement. The particle ma triggers the ∃ quantifier at the level of T’ and ∃ c-commands the object wh-word shenme ‘what’ so that the latter obtains an ∃-reading something/anything.

(49)

Step 2: The identical TP2 is merged at the Spec of DisjP. Since ma does not c-command the TP2 located at the Spec of DisjP, ma cannot trigger the ∃ quantifier in TP2. Therefore, the object shenme ‘what’ in TP2 cannot get an ∃-reading.

(50)

Step 3: The lower TP1 in the complement position of DisjP is deleted.

(51)

At the end of the derivation, shenme ‘what’ in the TP2, which is located at the Spec of DisjP, fails to get an ∃-reading, contrary to the fact. This example constitutes an argument against the disjunction analysis of SFP. By contrast, the comp-to-spec raising analysis precisely predicts the indefinite reading of the wh-object. The derivation goes as follows.

Step 1: The C head ma takes the TP as its complement. The particle ma triggers the ∃ quantifier at the level of T’ and ∃ c-commands the object wh-word shenme ‘what’ so that the latter obtains an ∃-reading “something/anything”.

(52)

Step 2: The complement TP raises to the Spec of CP.

(53)

Since the ∃ quantifier has already been generated inside the TP before its raising, the ∃-reading of shenme ‘what’ is therefore guaranteed.5

7. A Minimalist Derivation

Pan (forthcoming) proposes an analysis which also adopts the idea of comp-to-spec raising of SFP, but the motivation of such raising and the technical details differ from the previous analyses. Under the minimalist framework, each SFP projects a phase and bears an EPP feature, which must be satisfied. The EPP of a phasal head C can be satisfied by externally merging an XP or a null operator in Spec CP, or, by internally merging an XP in the Spec under an Agree relation between the Probe-C and the Goal-XP. If there is no candidate to satisfy the EPP feature, the entire complement of the phase head C must raise to the Spec CP as a last resort to fulfill the requirement of the EPP.

The phasehood tests applied to SFPs by Pan are based on Chomsky (2000, 2001) and Citko (2014). Each phrase projected by an SFP is a derivational and transferable unit for the Conceptual-Intentional (C-I) interface and for the Articulatory-Perceptual (A-P) interface, which satisfies the basic criteria for phases. As any phase head, an SFP triggers Spell-Out and Transfer. The complement of an SFP is also a transferrable unit, which is known as an important property of a phasal domain. Both a phrase headed by an SFP and the complement of an SFP are phonological units, just like a phase and its phasal domain. Most importantly, an element moving out of a phase headed by an SFP can be interpreted at its edge. The complement of an SFP is moved to the edge in order to postpone the transfer of the phrases that are embedded within the complement, which allows these phrases to be extracted later. An important argument in support of this analysis is that when the concerned phase edge is occupied and unavailable for the moved complement, the phrases embedded within the complement will not be able to be extracted in a later stage after the complement is transferred to the interfaces.

Let us start with simple cases. (54) involves two SFPs, each of which heads a phase. (55) is derived from (54) by moving the topic phrase that painting out of the TP to the Spec of TopP.

(54)

(55)

Pan (forthcoming) briefly illustrates the major steps of the minimalist derivation. The derivation is based on the second version of Phase Impenetrability Condition (PIC) proposed in Chomsky (2001).

(56)

In this version of PIC, the domain of the lower phase becomes inaccessible to further operations only after the next (higher) phasal head is merged. The major steps of the derivation of (55) are presented as follows.

Step 1: Since there is no candidate, which can be externally or internally merged with the S.Asp-le head to satisfy its EPP feature, the complement TP raises to the Spec of S.Asp-le to satisfy the EPP as a last resort.6

(57)

Step 2: The S.AspP raises to the Spec of iForce-ma to satisfy the EPP feature. Since the iForce is a phase head, the domain of the lower phase S.AspP, which is the lower copy of the TP, is transferred to the interfaces. Note that the higher copy of the TP is in fact at the edge of the phase iForceP, which is an escape hatch, therefore, it has not been transferred.

(58)

Step 3: Since the entire TP is at an escape hatch, its internal component is still accessible to further operations. This is why the topic phrase that painting can be extracted in the next phase cycle TopP.

(59)

By contrast, the situation is different for sentences in (60)(61).

(60)

(61)

In (61), TopP, NegQP and S.AspP are phases and their edges are escape hatches for Ā-movement. The idea is that the specifier of NegQP is occupied by the negative wh-word shenme ‘what’, and as a result, it is unavailable for any Ā-movement. Therefore, the topic phrase that painting cannot be extracted from the TP according to PIC. We continue the derivation from the step 1 of (57).

Step 2: The NegQ head is merged with the S.AspP and the negative wh-phrase shenme ‘what’ is merged in the specifier of the NegQP to satisfy the EPP feature. Once the EPP on the NegQ head is satisfied, its complement (i.e., S.AspP) no longer needs to raise to the Spec of NegQP. Since the NegQ head is a phase head, the domain of the lower phase S.AspP, which is the lower copy of the TP, is transferred to the interfaces. Note that at this moment, the higher copy of the TP is still available for further operations since it is located at the edge of the S.AspP, which is an escape hatch.

(62)

Step 3: When the next phasal head, Top, is merged with the NegQP, the domain of the NegQP (i.e., S.AspP) is transferred to the interfaces. The transferred S.AspP is no longer available for further operations. Note that at this stage, the higher copy of the TP has also been transferred and as a result, the topic phrase that painting can no longer be extracted, which is why the derivation crashes.

(63)

8. Conclusion

This article reviews the main findings concerning SFPs in Chinese. Diachronic studies concentrate on the origin and the evolution of each SFP, which helps us understand the core semantics and the discourse functions of SFPs in modern Mandarin. Traditional grammar tries to capture the core semantics as well as the diverse interpretations developed from the core semantics of each SFP. Syntactically, SFPs head different functional projections split from CP. Both traditional grammarians and generative grammarians are interested in the co-occurrence of different SFPs that display a rigid order. This article has reviewed the proposal that such an order is regulated by a discourse constraint related to subjectivity, according to which higher functional projections are directly linked to the speaker’s subjective attitude and are generally excluded from embedded clauses, whereas lower projections are more related to the sentence subject and are less subjective and can appear in embedded clauses. This constraint offers an explanation to the question of why only some SFPs can appear in embedded clauses whereas the others show root properties. Much work has also been done to account for the final order of SFPs. We compared two different derivations: the disjunction analysis and the complement-to-specifier raising analysis. Under the Minimalist Program, each SFP heads a phase and bears an EPP feature. Complement-to-specifier raising is required as a last resort to satisfy the EPP. The complement of an SFP is moved to the phase edge to postpone the transfer of the phrases that are embedded within the complement, which allows these phrases to be extracted later. Importantly, when the concerned phase edge is not available for the moved complement, phrases embedded within the complement can no longer be extracted in a later stage after the complement is transferred given the Phase Impenetrability Condition.

Further Reading

  • Alleton, V. (1981). Final particles and expression of modality in modern Chinese. Journal of Chinese linguistics, 9(1), 91–115.
  • Bayer, J. (1999). Final complementizers in hybrid languages. Journal of Linguistics, 35(2), 233–271.
  • Biberauer, T., Haegeman, L., & van Kemenade, A. (2014). Putting our heads together: Towards a syntax of particles. Studia Linguistica, 68(1), 1–15.
  • Cheng, L. L. S. (1991). On the typology of wh-questions [Unpublished doctoral dissertation]. MIT.
  • Deng, D. (2015). The syntacticization of illocutionary forces and the root vs. non-root distinction: Evidence from the sentence-final particle ba in Mandarin. Lingua, 162, 32–55.
  • Erlewine, M. Y. (2017). Low sentence-final particles in Mandarin Chinese and the final-over-final constraint. Journal of East Asian Linguistics, 26(1), 37–75.
  • Fang, M. (方梅‎). (1994). Běijīnghuà jùzhōng yǔqìcí de gōngńeng yánjiū (北京话句中语气词的功能研究‎) [Studies on the function of sentence-inner particles in Beijing dialect]. Zhongguo yuwen (中国语文‎), 2, 129–138.
  • Guo, X. (郭锡良‎). (1988). Xiānqín yǔqìcí xīntàn (先秦语气词新探‎) [Modal particles in Pre-Qin dynasty: A new perspective]. Guhanyu yanjiu (古汉语研究‎) [Research in Ancient Chinese Language], 49, 50–55, 74–82.
  • Haegeman, L. (2014). West Flemish verb-based discourse markers and the articulation of the speech act layer. Studia Linguistica, 68(1), 116–139.
  • Haegeman, L., & Hill, V. (2013). The syntacticization of discourse. In R. Folli, C. Sevdali, & R. Truswell (Eds.), Syntax and its limits (pp. 370–390). Oxford, UK: Oxford University Press.
  • Hu, M. (胡明扬‎). (1981). Běijīnghuà de yǔqì zhùcí he tàncí (北京话的语气助词和叹词‎ ) [On modal particles and interjections in Beijing dialect]. Zhongguo yuwen (中国语文‎), 5/6, 347–350, 416–423.
  • Huang, G. (黄国营‎). (1994). Jùmò yǔqìcí de céngcì dìwèi (句末语气词的层次地位‎) [The hierarchical distribution of sentence-final particles]. Yuyan yanjiu (语言研究‎) [Studies in Language and Linguistics], 1, 1–9.
  • Li, X. (李小军‎). (2010). Yǔqìcí “” “éryǐ” de xíngchéng fāzhǎn jí yǒuguān wèntí (语气词“已”“而已”的形成、发展及有关问题‎) [The engenderment and development of the modal particles yi (‎) and er-yi (而已‎), and the related problems]. Hanyushi xuebao (汉语史学报‎) [Journal of Chinese Language History], 9, 59–70.
  • Liu, X. (刘勋宁‎). (1985). Xiàndài hànyǔ jùwěi “le” de láiyuán (现代汉语句尾“了”的来源‎) [On the origin of the sentence-final particle le in modern Chinese]. Fangyan (方言‎) [Dialect], 2, 128–133.
  • Lu, J. (陆俭明‎). (1984). Guānyú xiàndài hànyǔ lǐ de yíwèn yǔqìcí (关于现代汉语里的疑问语气词‎) [On question particles in contemporary Chinese]. Zhongguo yuwen (中国语文‎), 5, 330–337.
  • Munaro, N., & Poletto, C. (2005). On the diachronic origin of sentential particles in North-Eastern Italian dialects. Nordic Journal of Linguistics, 28(2), 247–267.
  • Shao, J. (邵敬敏‎). (2012). Lùn yǔqìcí “a” zài yíwènjù zhōng de zuòyòng jì fāngfǎlùn de fǎnsī (论语气词“啊”在疑问句中的作用暨方法论的反思‎) [The function of modal particle “A (‎)” in interrogative sentences and the introspection of relevant methodology]. Yuyan kexue (语言科学‎) [Linguistic Sciences], 6, 596–603.
  • Shi, J. (史金生‎). (2000). Chuánxìn yǔqìcí “de” “le” “ne” de gòngxiàn shùnxù (传信语气词‎“‎”“‎”“‎”的共现顺序‎) [The co-occurrence sequence of the evidential modal particles de (‎), le (‎) and ne (‎)]. Hanyu xuexi (汉语学习‎) [Chinese Language Learning], 5, 32–35.
  • Simpson, A. (2014). Sentence-final particles. In C.-T. J. Huang, Y.-H. A. Li, & A. Simpson (Eds.), The handbook of Chinese linguistics (pp. 156–179). Hoboken, NJ: John Wiley & Sons.
  • Soh, H. L. (2009). Speaker presupposition and Mandarin Chinese sentence-final -le: A unified analysis of the “change of state” and the “contrary to expectation” reading. Natural Language & Linguistic Theory, 27(3), 623–657.
  • Sun, X. (孙锡信‎). (1999). Jìndài hànyǔ yǔqìcí—Hànyǔ yǔqìcí de lìshǐ kǎochá (近代汉语语气词‎: 汉语语气词的历史考察‎) [Modal particles in early modern Chinese: A historical study on modal particles in Chinese]. Beijing, China: Yuwen chubanshe.
  • Wu, F. (吴福祥‎). (1997). Cóng “VP-neg” shì fǎnfù wènjù de fēnhuà tán yǔqìcí “me” de chǎnshēng (‎“VP-neg”式反复问句的分化谈语气词‎“‎”的产生‎) [The origin of the modal particle me from the perspective of evolution of the ‘VP-neg’ construction]. Zhongguo yuwen (中国语文‎), 1, 44–54.
  • Yang, X., & Wiltschko, M. (2016). The conformational marker ha in Northern Mandarin. Journal of Pragmatics, 104, 67–82.
  • Yang, Y. (杨永龙‎). (2000). Xiānqín hànyǔ yǔqìcí tóngxiàn de jiégòu céngcì (先秦汉语语气词同现的结构层次‎) [The hierarchical structures of concurrences of Chinese modal particles used in the pre-Qin dynasty]. Guhanyu yanjiu (古汉语研究‎) [Research in Ancient Chinese Language], 4, 23–29.
  • Zhang, N. (2019). Sentence-final aspect particles as finite markers in Mandarin Chinese. Linguistics, 57(5), 967–1023.

References

  • Bailey, L. (2012). The syntax of question particles [Unpublished doctoral dissertation]. Newcastle upon Tyne, UK: Newcastle University.
  • Cao, G. (曹广顺‎). (1986). Zǔtáng Jí zhōng yǔ yǔqì zhùcí “ne” yǒuguān de jǐgè zhùcí (祖堂集》中与语气助词‎“‎”有关的几个助词‎) [Several particles connected with the modal particle ne in Zutang ji]. Yuyan yanjiu (语言研究‎) [Studies in Language and Linguistics], 2, 115–122.
  • Chao, Y. R. (1968). A grammar of spoken Chinese. Berkeley, CA and Los Angeles, CA: University of California Press.
  • Chomsky, N. (2000). Minimalist inquiries: The framework. In R. Martin, D. Michaels, & J. Uriagereka (Eds.), Step by step: Essays on minimalist syntax in honor of Howard Lasnik (pp. 89–155). Cambridge, MA: MIT Press.
  • Chomsky, N. (2001). Derivation by phase. In M. Kenstowicz (Ed.), Ken Hale: A life in language (pp. 1–52). Cambridge, MA: MIT Press.
  • Chu, C. C. (2002). Relevance theory, discourse markers and the Mandarin utterance-final particle a/ya. Journal of the Chinese Language Teachers Association, 37(1), 1–42.
  • Citko, B. (2014). Phase theory: An introduction. New York, NY: Cambridge University Press.
  • Cui, X. (崔希亮‎). (2019). Hànyǔ yǔqìcí “~ma” de qíngtài yìyì (汉语语气词‎“~嘛‎”的情态意义‎) [The modal meanings of the Chinese modal particle ma]. Yuyan jiaoxue yu yanjiu (语言教学与研究‎) [Language Teaching and Linguistic Studies], 4, 60–68.
  • Cui, X. (崔希亮‎). (2020). Yǔqìcí yǔ yánzhě tàidù (语气词与言者态度‎) [Modal particles and the speaker’s attitude]. Yuyan jiaoxue yu yanjiu (语言教学与研究‎) [Language Teaching and Linguistic Studies], 3, 50–59.
  • Djamouri, R., & Paul, W. (2019). Disharmony in harmony with diachronic stability: The case of Chinese. In A. Breitbarth, M. Bouzouita, L. Danckaert, & M. Farasyn (Eds.), The determinants of diachronic stability (pp. 101–130). Amsterdam, Netherlands: John Benjamins.
  • Gunlogson, C. (2001). True to form: Rising and falling declaratives as questions in English [Unpublished doctoral dissertation]. Santa Cruz, CA: University of California at Santa Cruz.
  • Hsieh, F.-F., & Sybesma, R. (2011). On the linearization of Chinese sentence-final particles: Max spell out and why CP moves. Korea Journal of Chinese Language and Literature, 49(3), 53–90.
  • Huang, C.-T. J. (1982). Logical relations in Chinese and the theory of grammar [Unpublished doctoral dissertation]. MIT.
  • Jiang, L. (江蓝生‎). (1986). Yíwèn yǔqìcí “ne” de láiyuán (疑问语气词‎“‎”的来源‎) [On the origin of the interrogative particle ne]. Yuwen yanjiu (语文研究‎) [Linguistic Research], 2, 17–26.
  • Jiang, S. (蒋绍愚‎). (2005). Jìndài hànyǔ yánjiū gàiyào (近代汉语研究概要‎) [An overview of the studies of early modern Chinese]. Beijing, China: Peking University Press.
  • Jin, L. (金立鑫‎). (1996). Guānyú yíwènjù zhōng de “ne” (关于疑问句中的‎“‎”) [On the particle ne in questions]. Yuyan jiaoxue yu yanjiu (语言教学与研究‎) [Language Teaching and Linguistic Studies], 4, 43–49.
  • Julien, M. (2002). Syntactic heads and word formation. Oxford, UK: Oxford University Press.
  • Kayne, R. (1994). The antisymmetry of syntax. Cambridge, MA: MIT Press.
  • Lee, T. H.-t. (1986). Studies on quantification in Chinese [Unpublished doctoral dissertation]. University of California at Los Angeles.
  • Li, B. (2006). Chinese final particles and the syntax of the periphery [Unpublished doctoral dissertation]. Leiden University.
  • Li, C. N., & Thompson, S. (1981). Mandarin Chinese: A functional reference grammar. Los Angeles, CA: University of California Press.
  • Ota, T. (太田辰夫‎). (2003[1958]). Zhongguoyu Lishi Wenfa (中国语历史文法‎) [A historical grammar of modern Chinese] (2nd ed.) (S. Jiang & C. Xu, Trans.). Beijing, China: Peking University Press.
  • Pan, V. J. (2015). Mandarin peripheral construals at the syntax-discourse interface. The Linguistic Review, 32(4), 819–868.
  • Pan, V. J. (2019a). Architecture of the periphery in Chinese: Cartography and minimalism [Routledge Studies on Chinese Linguistics]. London, UK & New York, NY: Routledge.
  • Pan, V. J. (2019b). Derivation of the apparent narrow scope of sentence-final particles in Chinese: A reply to Erlewine (2017). Studies in Chinese Linguistics, 39(2), 99–126.
  • Pan, V. J. (forthcoming). Deriving head-final order in the peripheral domain of Chinese. Linguistic Inquiry.
  • Pan, V. J., & Paul, W. (2016). Why Chinese SFPs are neither optional nor disjunctors. Lingua, 170, 23–34.
  • Paul, W. (2014). Why particles are not particular: Sentence-final particles in Chinese as heads of a split CP. Studia Linguistica, 68(1), 77–115.
  • Paul, W. (2015). New perspectives on Chinese syntax. Berlin, Germany: De Gruyter Mouton.
  • Paul, W., & Pan, V. J. (2017). What you see is what you get: Chinese sentence-final particles as head-final complementizers. In J. Bayer & V. Struckmeier (Eds.), Discourse particles—Formal approaches to their syntax and semantics (pp. 49–77). Berlin, Germany: De Gruyter Mouton.
  • Qi, H. (齐沪扬‎). (2002a). Lùn xiàndài hànyǔ yǔqì xìtǒng de jiànlì (论现代汉语语气系统的建立‎) [On the establishment of mood system in modern Chinese]. Hanyu xuexi (汉语学习‎) [Chinese Language Learning], 2, 1–12.
  • Qi, H. (齐沪扬‎). (2002b). “Ne” de yìyì fēnxī hé lìshǐ yǎnbiàn (“‎”的意义分析和历史演‎) [Analyses of the meaning of ne and its historical evolution]. Shanghai shifan daxue xuebao (上海师范大学学报‎ (哲学社会科学版‎)) [Journal of Shanghai Normal University (Philosophy and Social Sciences)], 1, 34–45.
  • Qi, H. (齐沪扬‎). (2002c). Yǔqìcí yǔ yǔqì xìtǒng (语气词与语气系统‎) [Modal particles and the mood system]. Hefei, China: Anhui jiaoyu chubanshe.
  • Ren, Y. (任鹰‎). (2017). Yǔqìcí “ne” de gōngnéng jí láiyuán zàiyì (语气词‎“‎”的功能及来源再议‎) [Revisiting the function and the origin of the modal particle Ne]. Yuyan jiaoxue yu yanjiu (语言教学与研究‎) [Language Teaching and Linguistic Studies], 5, 70–80.
  • Rizzi, L. (1997). The fine structure of the left periphery. In L. Haegeman (Ed.), Elements of grammar (pp. 281–337). Dordrecht, Netherlands: Kluwer.
  • Simpson, A., & Wu, Z. (2002). IP-raising, tone sandhi and the creation of S-final particles: Evidence for cyclic spell-out. Journal of East Asian Linguistics, 11(1), 67–99.
  • Sun, X. (孙锡信‎). (1992). Yǔqìcí “ne” “li” kǎoyuán bǔshù (语气词‎“‎”“‎”考源补述‎) [A supplementary survey on the origin of the modal particles ne and li]. Hubei daxue xuebao (湖北大学学报‎ (哲学社会科学版‎)) [Journal of Hubei University (Philosophy and Social Sciences)], 6, 69–74, 82.
  • Sybesma, R. (1999). The Mandarin VP. Dordrecht, Netherlands: Kluwer Academic.
  • Takita, K. (2009). If Chinese is head-initial, Japanese cannot be. Journal of East Asian Linguistics, 18(1), 41–61.
  • Tang, S.-W. (1998). Parametrization of features in syntax [Unpublished doctoral dissertation]. University of California at Irvine.
  • Tang, S.-W. (2015). A generalized syntactic schema for utterance particles in Chinese. Lingua Sinica, 1(3), 1–23.
  • Tang, S.-W. (邓思颖‎). (2019). Cíhuì céngcì de jùmò zhùcí (词汇层次的句末助词‎) [Sentence-final particles in the lexical layer]. Yuyan jiaoxue yu yanjiu (语言教学与研究‎) [Language Teaching and Linguistic Studies], 3, 38–45.
  • Tang, S.-W. (2020). Cartographic syntax of performative projections: Evidence from Cantonese. Journal of East Asian Linguistics, 29(1), 1–30.
  • Wang, J., & Bi, Y. (王珏‎, 毕燕娟‎). (2018). Yǔqìcí jùmò diéyòng shùnxù yánjiū (语气词句末迭用顺序研究‎) [On the successive order sequence of final particles at the sentence-end]. Yuyan jiaoxue yu yanjiu (语言教学与研究‎) [Language Teaching and Linguistic Studies], 1, 89–100.
  • Wang, L. (王力‎). (1980). Hànyǔ shǐ gǎo (汉语史稿‎) [An outline of the history of Chinese]. Beijing, China: Zhonghua Book Company (中华书局‎).
  • Wu, G. (2005). The discourse function of the Chinese particle ne in statements. Journal of the Chinese Language Teachers Association, 40(1), 47–82.
  • Yang, Y. (杨永龙‎). (2003). Jùwěi yǔqìcí “ma” de yǔfǎhuà guòchéng (句尾语气词‎“‎”的语法化过程‎) [On the grammaticalization of the Mandarin interrogative particle ma]. Yuyan kexue (语言科学‎) [Linguistic Sciences], 1, 29–38.
  • Zhong, Z. (钟兆华‎). (1997). Lùn yíwèn yǔqìcí “ma” de xíngchéng yǔ fāzhǎn (论疑问语气词‎“‎”的形成与发展‎) [On the origin and the development of the interrogative particle ma]. Yuwen yanjiu (语文研究‎) [Linguistic Research], 62(1), 1–8.
  • Zhu, D. (1982). Yǔfǎ jiǎngyì (语法讲义‎) [A course lecture on grammar]. Beijing, China: Commercial Press.

Notes

  • 1. The major dynasties are listed here: Spring and Autumn (770 bc–476 bc); Qin dynasty (221 bc–207 bc); Han dynasty (202 bc– 220 ad); Tang dynasty (618 ad–907 ad); Five Dynasties and Ten Kingdoms period (907 ad– 979 ad); Song dynasty (960 ad–1279 ad); Jin (1115 ad–1234 ad); Yuan dynasty (1271 ad–1368 ad); Ming dynasty (1368 ad– 1644 ad); Qing dynasty (1636 ad–1912 ad).

  • 3. It is still controversial whether all these three ne can really be distinguished from each other. Especially, it has been recognized that neint is only compatible with an interrogative sentence, but it does not have any inherent interrogative force, which is different from a real interrogative particle such as the yes-no question particle ma (see Li, 2006; Pan & Paul, 2016).

  • 5. In fact, the existential operator cannot be generated adjoining to the whole Disjunction Phrase in Chinese. This is because the existential closure cannot apply in a position higher than T’ level in Chinese since a wh-subject can never have an indefinite reading. This has been extensively argued in Pan (2019b). For instance, in (i), the wh-subject shéi ‘who’ cannot get an indefinite reading ‘anyone’. This is because ∃ is triggered at T’ level by the SFP ma; however, the wh-subject shéi ‘who’ is outside the scope of ∃ and therefore, ∃ cannot bind shéi ‘who’ as a variable.

    (i)

  • 6. An SFP does not function as a probe and it does not agree with any particular goal.