skip to primary navigationskip to content

Mental Representations of WH-words in non-native grammars of Chinese. Dr Boping Yuan

Mental Representations of WH-words in non-native grammars of Chinese (Dr Boping Yuan)

1. Behaviours of wh-words in English speakers’ L2 Chinese wh-questions

In second language (L2) research, Mandarin Chinese (hereafter Chinese) is generally treated as a “pure” wh-in-situ language, that is, wh-words remain in situ (i.e. without any movement). Many L2 studies have looked at L2 acquisition of wh-movement by speakers of a wh-in-situ language (Schachter, 1990; Martohardjono, 1993; Martohardjono and Gair, 1993; Klein, 1995; Hawkins and Chan, 1997; White and Juffs, 1998; Hawkins, 2004, Hawkins and Hattori, 2006), or vice versa (Kim, 2003; Yuan, forthcoming). Few, however, have done research investigating whether all wh-words behave similarly and develop uniformly in L2 acquisition of wh-questions. Furthermore, variability in L2 grammars has received increasing attention in second language research. However, most studies so far have concentrated on variability in L2 grammars of European languages, leaving variability in L2 East Asian languages, such as Chinese, unexplored. In this project, we will examine the behaviours of different wh-words in the development of English speakers’ L2 Chinese wh-questions.

Unlike English wh-questions, in which the wh-word has to move to the initial position of the sentence, as can be seen in the English translation of (1), the wh-word in Chinese wh-questions remains in situ, as shenme in (1). In Chinese, there are two question particles ma and ne, with ma being a yes-no-question particle (hereafter yes-no particle) and ne a wh-question particle (hereafter wh-particle). The yes-no particle is generally obligatory, as in (2), but the wh-particle is optional in Chinese wh-questions, as in (1). Another characteristic of Chinese wh-questions is that the in-situ wh-word can be located inside a complex NP (CNP) as in (3) or inside a sentential subject as in (4). The CNP and the sentential subject are often called “islands”. Obviously, wh-words cannot be extracted from these islands in English and this impossible extraction is generally argued to be due to the island effect.


(1) Ni  xiang chi shenme (ne)?

you want eat what       Q

‘What would you like to eat?’


(2) Ni xihuan ta   ma?

you like   him Q

‘Do you like him?’


(3) Ni xihuan [shei xie     de   shu] (ne)?

you  like    who write DE book  Q

‘*Whoi do you like the book [that ti wrote]?’


(4) [Shei qu Beijing] bijiao        heshi  (ne)?

who go Beijing relatively  suitable Q

‘*Whoi is [ti to go/goes to Beijing] more appropriate?’


In this project, we make a distinction between wh-nominals, wh-arguments and wh-adverbs. We assume by following Tsai (1994a, b, 1999) and Cheng and Rooryck (2002) that wh-adverbs in Chinese have an operator, but wh-nominals do not and that the operator has to undergo movement to Spec of CP for feature checking.


Will English-speaking learners of Chinese transfer wh-movement from their L1 English to their L2 Chinese? Will they be able to value (i.e. to specify) Chinese wh-questions with the wh-particle ne?  In English speakers’ L2 Chinese grammars, do wh-words develop in a uniform fashion? Do different wh-words behave differently both synchronically and developmentally?


2. Domain-Wide or Variable-Dependent? --- A Study of Semantics-Syntax Interface in L2 Chinese Grammars of Wh-Words Used as Existential Polarity


Recent years have witnessed an increasing number of second language (L2) studies of interfaces, and this kind of research has opened up a new frontier in L2 research, particularly in investigating possible causes of divergence between L2 grammars and the native grammar. However, there is disagreement among researchers about the extent to which L2 grammars are affected by interfaces. Some researchers (e.g. Hopp, 2004; Sorace, 2004; 2006; Tsimpli and Sorace, 2006) find that interfaces between the syntax domain and other domains, such as semantics, pragmatics and discourse, are particularly difficult for adult L2 learners and can result in delay in L2 development. Sorace (2006) and Sorace and Filiaci (2006) go a step further and suggest that interface properties involving the syntax domain and another cognitive domain may “never be completely acquired by L2 learners” (Sorace, 2005: 69). In contrast, other researchers (Borgononvo et al., 2005; 2006; Dekydtspotter, Sprounse and Thyre, 1999/2000; Dekydtspotter and Sprouse, 2001) demonstrate that native-like grammars can be attained at an interface between the syntax domain and other domains even when no clear positive evidence is readily available in the input.


The current project examines a semantics-syntax interface in English and Japanese speakers’ L2 Chinese grammars of wh-words used as existential polarity words (EPWs hereafter). The study aims to investigate whether the semantics-syntax interface involving Chinese wh-words as EPWs can be acquired by English and Japanese speakers? If problems occur, will the interface domain be holistically affected in L2 Chinese grammars?


Wh-words in Chinese are ambiguous; they can be used as interrogative words, universal quantifiers as well as EPWs, as shown in (1).



a. Ni xiang mai shenme (ne)?             (shenme = an interrogative word)

You want buy what   (wh-Q)

‘What do you want to buy?’


b. Wo shenme dou xiang mai.                        (shenme = a universal quantifier)

I     what     each want buy

‘I want to buy everything.’


c.  Wo bu xiang mai shenme.              (shenme = existential polarity word)

I    not want  buy what

‘I don’t want to buy anything.’


The Chinese examples in (1) indicate that wh-words in Chinese should be considered as variables and that their force values depend on what licenses them. If they are licensed by the wh-particle ne in the head C0, as in (1a), they are identified as interrogative wh-words; if they are licensed by dou, as in (1b), they are interpreted as universal quantifiers; and if they are licensed by a c-commanding negator, as in (1c), they are EPWs.  This suggests that the wh-word itself does not have an inherent force and that its meaning is determined by some element that bears a certain structural relation with it. Huang (1982), Li (1992) and Lin (1998) argue that Chinese wh-words used as EPWs not only have to be licensed but they are also subject to both syntactic and semantic restrictions. Wh-words in Japanese can also be used as EPWs, but they are licensed by particles ka or mo. That is, they have to be combined with the particles in order to become EPWs. In contrast to Chinese and Japanese wh-words, wh-words in English are not ambiguous. They are intrinsically interrogative words with [+Q] and [+wh] features inherently embedded in them, and therefore they cannot be used as EPWs.


3. Study of wh-words used as universal quantifiers in English and Japanese speakers’ L2 Chinese


In English, wh-words, such as who, what, when, where, how and why, are mainly used as interrogatives in wh-questions. In contrast, Chinese wh-words, apart from being used as interrogatives, can also be used as universal quantifiers, as in (1a) and (2a). However, there are constraints on the use of a wh-word as a universal quantifier in Chinese (Cheng 1991, 1995; Lee 1986; Xu 1997). When used as a universal quantifier, the wh-word has to be licensed by dou, which is considered the head of a functional category in this paper. Without the use of dou, the sentence would be ungrammatical, as in (1b) and (2c). This is believed to be due to the fact that the wh-word in Chinese has an uninterpretable distributing feature and does not have any quantificational force, and that it is dou which values the wh-word and contributes the universal quantificational force to it. Another constraint is that the wh-word has to appear adjacent to the left of dou. This is because the head dou carries a movement-forcing interpretable feature, which forces the wh-word to move to its specifier for feature checking. This can be clearly seen from the contrast in (2a,b). The wh-word shenme “what” in (2) is the object of the verb chi “eat”; the wh-word has to raise to the left of dou for checking and valuing, as in (2a), and the wh-word staying in situ would lead to ungrammaticality, as in (2b).


(1)   a. Shei  dou xihuan ta.

who  all[1] like    her

‘Everyone likes her.’


b. * Shei  xihuan ta.

who   like     her

‘Everyone likes her.’


(2)  a.  Ta shenmei dou chi   ti.

he  what     all   eat

‘He eats everything.’


b.*Ta dou chi shenme.

he  all  eat  what

‘He eats everything.’


c.*Ta  chi shenme.

he  eat  what

‘He eats everything.’


Similar to wh-words in Chinese, Japanese wh-words can also be used as universal quantifiers. In this case they have to be combined with the particle –mo.  This can be seen in the list in (3) and example sentences in (4) and (5).[2]


(3) A list of Japanese wh-words used as interrogatives and universal quantifiers (adapted from Watanabe 1992: 48-49)

As interrogatives

As universal quantifiers (-mo)

dare “who”

dare-mo “everyone”

nani “what”

#nani-mo “everything”[3]

doko “where”

doko-mo “everywhere”

itsu “when”

itsu-mo “whenever”

naze “why”

*naze-mo ---------


(4)  dare-mo-ga         ringo-o    tabeta.

everyone-Nom  apple-Acc ate

‘Everyone ate an apple’                               (from Watanabe 2001: (26b))


(5)  Sensei-ga     dare-mo-o hometa.

 teacher-Nom everyone  praised
 ‘The teacher praised everyone.’    
(Chieko Kuribara, personal communication)


As we can see from (4) and (5), wh-words in Japanese behave as universal quantifiers in combination with the quantificational element –mo.  There is no need to use something equivalent to the Chinese dou in a Japanese sentence with a wh-word used as a universal quantifier, and unlike the case in Chinese, there is no requirement for movement or raising in such a sentence.[4] In this paper, we will follow Nishigauchi (1990), Watanabe (1992, 2001) and Tsai (1994, 1999) in assuming that the wh-word is combined with the particle –mo in order to function as a universal quantifier in Japanese sentences like (4) and (5). In these sentences, the wh-word itself does not have a fixed semantic force, and only when a relation is established between the wh-word and the particle –mo and when the former is licensed by the latter will the wh-word be identified as a universal quantifier. In this sense, wh-words in Japanese are variables and they have to be bound by their licensers.


It has been proposed by some second language (L2) researchers (e.g. Smith and Tsimpli 1995, Hawkins and Chan 1997) that features that are not used in L2 learners’ first language (L1) but are required in the L2 will become inaccessible in their L2 acquisition after a critical period. This has been referred to as the Representational Deficit Hypothesis (cf. Hawkins 2003). More recently, this group of researchers (Tsimpli 2003; Tsimpli and Dimitrakopoulou 2007; Hawkins 2004; Hawkins and Hattori 2006) propose that in adult L2 acquisition, uninterpretable features, but not interpretable features, will be inaccessible if they are not selected in the construction of the L1. This has been called the Interpretability Hypothesis. According to this hypothesis, the inaccessibility of the uninterpretable featureunderlying L2 grammar.0EuroSla and Hawkins and Hattori 2006) ction of thiee features will lead to a representational deficit in underlying L2 grammars although L2 learners’ performance may still appear native-like (cf. Hawkins and Hattori 2006).  As neither the functional projection headed by dou nor the uninterpretable distributing feature of the Chinese wh-word is available in English and Japanese, both the Representational Deficit Hypothesis and the Interpretability Hypothesis would predict a deficit in this aspect of English and Japanese speakers’ L2 Chinese grammars although it would not rule out the possibility that some behaviours of the wh-word and dou may appear native-like in the learners’ L2 Chinese.


This project examines whether dou in English- and Japanese-speaking learners’ L2 Chinese grammars can project a functional category with a movement-forcing interpretable distributing feature and whether the uninterpretable distributing feature of the wh-word used as a universal quantifier can be acquired in adult L2 acquisition of Chinese.


4. Is Chinese “daodi” “the hell” in English Speakers’ L2 Acquisition of Chinese daodi… wh… Questions?


English phrases such as what the hell, who on earth, what the dickens etc. are widely considered to belong to the same category and are generically called wh-the-hell phrases (cf. Dikken & Giannakidou, 2002; Huang & Ochi 2004; Chou, 2006, 2007). These English phrases are generally translated into Chinese as daodi…wh-word…. or jiujing…wh-word…(cf. Concise English-Chinese Chinese-English Dictionary, 1992; A Chinese-English Dictionary, 1997; Oxford Chinese Beginner’s Dictionary, 2001). As these two phrases behave the same in Chinese wh-questions, I will use daodi…wh… as a representative of the two in this paper.[5] The English wh-the-hell and Chinese daodi…wh share some syntactic properties, but they are also different from each other semantically, pragmatically as well as syntactically.


To examine English speakers’ L2 Chinese daodi…wh-questions, an empirical study will be conducted with a focus on the syntax-semantics and syntax-discourse interfaces involved in L2 Chinese questions of this type. We will ask the following research questions:

  1. Syntactically, is it possible for English speakers to establish the requirements in their L2 Chinese daodi…wh questions that daodi…wh is c-commanded by a CP[+Q], the wh-word is required in the c-commanding domain of daodi and the discontinuous form daodi…wh is used?
  1. At the syntax-semantics interface, is daodi…wh allowed in the complement clause of the affirmative form of veridical verbs in English speakers’ L2 Chinese, in contrast with wh-the-hell in their L1 English?
  2. At the syntax-discourse interface, (i) do English speakers transfer the aggressively non-D-linked property of the-hell from their L1 to daodi, which is not aggressively non-D-linked in the target language Chinese?  (ii) can daodi…wh be linked to a discourse-familiar referent in English speakers’ L2 Chinese, unlike its counterpart wh-the-hell in their L1 English? (iii) will English speakers be able to provide genuine information to the Chinese daodi…wh…modal question, instead of providing a negative rhetoric answer, as in the case in the wh-the-hell…modal question in their L1 English?
  3. If any L1 transfer occurs, will English speakers’ L2 Chinese grammars be able to recover from the cross-linguistic influence and acquire the native-like competence in the aspects concerned?


5. The effect of increased computation demands on the L2 syntax-pragmatics interface


Chinese daodi…wh-questions are considered approximate counterparts of English wh-questions with phrases such as what the hell, who on earth, what the dickens (cf. Huang 2010, Huang and Ochi 2004, Chou 2006, 2012, Yuan 2011, to appear).[6] English questions of this type are generically called wh-the-hell questions (cf. Dikken and Giannakidou 2002). In a Chinese daodi…wh-question, daodi can co-exist with a wh-adjunct such as weishenme “why”, zenme “how”, as in (1c) and (1d), as well as a wh-nominal, such as shenme “what”, shei “who”, as in (1a) and (1b).



a. Ni   daodi   xiang chi shenme?

you the hell want eat what

“What the hell would you like to eat?”


b. Ni    daodi  xiang  jian  shei?

you the hell want meet who

“Who the hell would you like to meet?”


c. Ni   daodi  weishenme mei   qu Beijing?

you the hell why        didn’t  go Beijing

“Why the hell did you not go to Beijing?”


d. Ni      daodi zenme qu Beijing?

You the hell how    go Beijing

“How the hell would you go to Beijing?”


e. *Ni    daodi   zenme mei   qu Beijing?

You the hell how  didn’t go Beijing

*“How come the hell you didn’t go to Beijing?


However, daodi cannot co-exist with the wh-word zenme in the wh-question in (1e), and this forms a striking contrast with the wh-questions in (1a-d), where daodi can co-exist with the other types of wh-words in Chinese wh-questions. The unacceptability of the sentence in (1e) is due to the fact that the wh-question bears two attitudes in it, an attitude of impatience borne by daodi and another attitude of counter-expectation carried by the wh-word zenme “how come”. The examples in (1) illustrate that the Chinese wh-question can accommodate both daodi and a wh-word in it as long as the wh-word does not carry an attitude with it. The attitude is considered to be in the domain of pragmatics, and the sentences in (1) demonstrate that the grammaticality of Chinese wh-questions with daodi involves the syntax-pragmatics interface and that they are governed by a pragmatic factor of attitude.


Great efforts have been made in second language (L2) research to identify why, unlike children acquiring their mother tongues, adult L2 learners rarely succeed in achieving native-like competence in their acquisition of the target language. In recent years, an increasing number of researchers have turned their attention to interfaces for accounting for the failure in adult L2 acquisition, particularly since the Interface Hypothesis (IH) proposed by Sorace and Filiaci (2006). “The IH originally proposed that language structures involving an interface between syntax and other cognitive domains are less likely to be acquired completely than structures that do not involve this interface” (Sorace 2011, p.1). In more recent work by Sorace and her colleagues, the syntax-pragmatics/discourse interface has been identified as being particularly vulnerable, which is exemplified by optionality, indeterminacy and instability (see Sorace 2011 and work cited within). Given that the grammaticality of Chinese wh-questions with daodi involves the syntax-pragmatics interface, it serves as a good candidate for testing the latest reformulation of the IH. In this project, we will examine whether English speakers’ [7]acquisition of L2 Chinese attitude-bearing questions will provide evidence in support of the IH.


[1] Strictly speaking, it is inaccurate to translate dou as “all”. As we will see below, dou is the head of a functional category in Chinese while “all” can be a floating quantifier in English although both are very close in their meanings.

[2] According to Nichigauchi (1990), a wh-word combined with the particle –mo is a universal quantifier only when it is used in a non-negative environment. If it is used in a negative environment, it is a negative existential polarity word.

[3] According to Watanabe (1992), this combination is only possible in a fixed expression like (i).


(i)     Nani-mo-ga      iya-ni        natta.

-Nom hateful-be became

‘Everything has become hateful/ I have become sick with everything.’


[4] The fact that the objects ringo-o (=apple) and dare-mo-o (=everybody) in (4) and (5) are in a preverbal position is not a result of movement. This is due to the fact that Japanese has a head final VP.

[5] Apart from being approximate counterparts of the English the hell and on earth, the Chinese daodi and jiujing also have other meanings. Daodi can also mean “finally” and “after all”, and jiujing can mean “after all” and “in the final analysis”. However, daodi or jiujing cannot be used in wh-questions in any of these readings (cf. Lü, 1981).

[6] Another approximate counterpart in Chinese is jiujing…wh-.  As both daodi…wh- and jiujing…wh- behave the same in this aspect of Chinese grammar, I will use daodi…wh… as a representative of the two in this article.