West Coast Conference on Formal Linguistics (WCCFL) (38th : 2020)

“Noun-less” nominal expressions in Mandarin Chinese Huang, Nick 2020

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


73804-Huang_N_Nounless_ WCCFL38_2020_handout.pdf [ 192.16kB ]
JSON: 73804-1.0389857.json
JSON-LD: 73804-1.0389857-ld.json
RDF/XML (Pretty): 73804-1.0389857-rdf.xml
RDF/JSON: 73804-1.0389857-rdf.json
Turtle: 73804-1.0389857-turtle.txt
N-Triples: 73804-1.0389857-rdf-ntriples.txt
Original Record: 73804-1.0389857-source.json
Full Text

Full Text

“Noun-less” nominal expressions in Mandarin ChineseNick Huang (znhuang@nus.edu.sg). National University of Singapore / University of Connecticut∗West Coast Conference in Formal Linguistics 20201 IntroductionThis talk: nominal syntax and classifier semantics, with a Mandarin Chi-nese case study.(1) liangtwo*(ben)CLshubook‘two books’1 [Mandarin Chinese, Num Cl Noun]Case study: exceptions to the above generalization.(2) liangtwo{nianyear/ tianday/ piaovote/ yepage/ bustep/ huastroke/ bi}stroke‘two years / days / votes / pages / steps / strokes of the brush/pen’These bipartite expressions superficially consist of just a numeral and aclassifier — see Chao (1968) (“quasi-measures”), C.-C. J. Tang (2005)and S.-W. Tang (2012, 2013), among others.• Sidenote: will refer to (1) as “tripartite expressions.”I review existing analyses of bipartite expressions and add novel argu-ments to make the claim that they are syntactically bipartite:∗Special thanks to WCCFL organizers and referees and the NUS SynSem readinggroup for comments. All errors are mine alone.1Abbreviations: ADV: adverbial marker; BA: ba construction marker; CL: classifier;MOD: nominal modifier; PL: plural; PRF: perfective aspect; SG: singular.• No phonologically-null noun, NP ellipsis, etc.• Contra S.-W. Tang (2013) and similar proposals for Japanese byWatanabe (2012) and for English by Kayne (2007).If this “nounlessness” argument goes through, there are interesting impli-cations for syntax and semantics.Syntax:• Data pose a problem for the NP Hypothesis of nominal expressions(Bruening 2009; Bruening et al. 2018, also Chomsky 1970).• Indirect evidence in favor of the DP Hypothesis (Abney 1987; Sz-abolcsi 1994, etc.).Semantics: Cross-linguistic distribution of classifiers has been tied todeeper cross-linguistic variation in noun and numeral semantics (Chier-chia 1998a; Rothstein 2010; Krifka 1995; Wilhelm 2008; Bale & Coon2014; Sudo 2016b).• Core idea: classifiers are needed so that numerals and nouns cancombine with each other to express cardinality.• Presuppose presence of numerals and nouns.• Data pose problems for these proposals about numeral and noun se-mantics.• Discuss whether we can “rescue” these proposals.1Caveat: Many other aspects of the syntax and semantics of classifiers,whether in Chinese or in other languages, that I will not be able to addresshere for scope and time reasons – please see the references.2 Bipartite nominal expressions2.1 HypothesesThere are a number of hypotheses one could have about the representationof bipartite nominal expressions and lexical items like nian ‘year’ or hua‘(brush)stroke’; some of the simpler ones are set out in (3).To avoid pre-judging the issue, I will henceforth refer to these lexical itemsas post-numeral lexical items (henceforth, ‘PNLs’).I will argue in favour of the first hypothesis (3a): a PNL is a classifier (C.-C. J. Tang 2005, S.-W. Tang 2012, 2013) and there is no null NP followingit (cf. S.-W. Tang 2012, 2013, Kayne 2007).(3) a. Num Cl b. Num NP + null Cl c. Num NP + head movementNum CLnian/huaNumCL/0NPnian/huaNumCL+Nnian/hua1NPt1d. Num Cl + elided NP e. Num Cl + covert NPNumCLnian/huaNP. . .NumCLnian/huaNP/02.2 PNLs are morphosyntactically classifiersFirst, reduplication (4): classifiers can be reduplicated to give a ‘one-by-one’ reading (following S.-W. Tang 2012, 2013).(4) ZhethisyionexiePLbaogao,reportta3SG[yionefenCLfen]CLdeADVkan.read‘These reports, he/she read them one by one.’PNLs can also be reduplicated to produce the same reading (5).(5) a. Ta3SG[yionehuastrokehua]strokedeADVbaBAzicharacterxiewritechu-lai.come-out‘He/she wrote out the character one stroke at a time.’b. Shijiantimejiuthenzheyanglike.this[yionetiandaytian]daydeADVguoqu.pass‘Time passes like this, day after day.’c. Womenweyaomust[yionepiaovotepiao]votedeADVqugozhengqufight.forxuanminvoterdeMODzhichi.support‘We must fight for voters’ support one vote at a time.’d. Ta3s[yioneyepageye]pagedeADVfan-kanturn.over-readmanhuacomicshu.book‘He/she read the comic book, turning over page after page.’e. Guojiacountryzhengnow[yionebustepbu]stepdeADVzouwalkxiangtowardmiewang.oblivion‘The country is now slowly slipping (lit. one step at a time)into oblivion.’Second, complementary distribution with “plural classifier” xie, which in-dicates a plurality of items denoted by the following NP (e.g. Li 1999).(6) a. yionebenCLshubook‘one book’2b. yionexiePLshubook/ *yionebenCLxiePLshubook/ *yionexiePLbenCLshubook‘some books’PNLs are also in complementary distribution with plural xie (7).(7) *yionexiePL(tianday/ huastroke/ yepage/ bu)step‘some days / strokes / pages / steps’Third, a “no bare classifiers” test, after an observation by Lü 1990 [1944](as cited by Zhang (2019)) and Yang 2001.A slight detour: three generalizations about Mandarin Chinese:• The numeral yi ‘one’ can be omitted (although interpreted) (8a).• NPs can be omitted through ellipsis (8b).• However, the omission of yi ‘one’ and an NP cannot happen together,i.e. classifiers cannot appear alone (8c).(8) a. Ta3SGmai-lebought(yi)onebenCLshu.book‘He bought one book.’b. Ta3SGmai-leboughtyionebenCLshu,bookwoIyealsomai-leboughtyionebenCL(shu).book‘He bought one book, and I also bought one (book).’c. Ta3SGmai-leboughtbenCLshu,bookwoIyealsomai-leboughtbenCL*(shu).book‘He bought one book, and I also bought one book.’ (Yang2001:76, ex. 33)Like classifiers, PNLs cannot appear alone (9).(9) a. ZhethisgeCLzicharactershao-lemissed*(yi)onehua.stroke‘This character is missing a stroke.’ (e.g. pointing out mis-takes in Chinese characters.)b. WoIzaiatzhethislüguanhotelzhu-lestayed*(yi)onetian.day‘I stayed at this hotel for a day.’c. LisiLisihuodereceive*(yi)onepiao.vote‘Lisi received a vote.’d. ZhethisfenCLbiaogeformlou-lemissed*(yi)oneye.page‘This form is missing a page.’e. WoItui-lemoved.back*(yi)onebu.step‘I took a step back.’2.3 PNLs are not nouns(3) c. Num NP + head movement of NP to Cl (cf. Simpson & Ngo2018 for Vietnamese)NumCL+Nnian/hua1NPt1First, as a rule, nouns can co-occur with the plural xie (6). But (7) showedthat PNLs cannot.(6) yionexiePLshubook‘some books’(7) *yionexiePL(tianday/ huastroke/ yepage/ bu)step‘some days / strokes / pages / steps’Second, nouns can appear bare in Mandarin Chinese (10). But PNLs can-not, as we saw in the discussion of the “no bare classifiers” test (9).3(10) WoImai-leboughtshu.book‘I bought books.’Third, nouns (and even overt pronouns) can be modified with relativeclauses (11).(11) a. [RC huicanyouyongswimde]MODrenperson/ woI‘a person who can swim’ / ‘I, who can swim’b. [RC xiewritecuowrongde]MODbihuabrushstroke‘a stroke that was written incorrectly’PNLs cannot be modified directly by relative clauses (12).(To modify a PNL, one must first quantify it with a numeral to produce anominal expression (12b).)(12) a. *[RC tamentheyyinprintcuowrongde]MODyepageIntended: ‘a page that they printed incorrectly’b. [RC tamentheyyinprintcuowrongde]MODyioneyepage‘a page that they printed incorrectly’2.4 There is no null NP in bipartite expressions2.4.1 Ruling out NP ellipsis(3) d. Num Cl + elided NPNumCLnian/huaNP. . .NP ellipsis is usually unacceptable in out of the blue contexts.(13) Biangbiangzicharacteryouhas4242geCL[NP #(bihua)].brushstroke‘The character biang has 42 brushstrokes.Bipartite expressions are acceptable in out of the blue contexts.(14) Biangbiangzicharacteryouhas4242hua.stroke‘The character biang has 42 brushstrokes.’2.4.2 No covert noun(3) e. Num Cl + covert NP (cf. Kayne 2007, S.-W. Tang 2013)NumCLnian/huaNP/0For concreteness, suppose the hypothesis in (3d) works as follows:• There is a set of covert (phonologically null) nouns that mean ‘time,’‘vote,’ ‘(brush) stroke,’ etc.• PNLs are classifiers; the form of the PNL depends on the noun.• E.g., hua is used with the null noun that means ‘stroke.’2One might expect covert nouns to have the same properties as phonologi-cally overt nouns. This prediction is too strong.First, overt nouns can be modified with relative clauses. But these covertnouns (if they exist) cannot be, even though the PNL makes clear whichnoun is intended.2This assumption about null nouns is already somewhat problematic, as phonologicallynull elements are usually posited for closed-class items or functional heads, and less oftenfor open-class morphemes like nouns.4(15) a. (i) liangtwogeCL[RC xie-cuowrite-wrongde]MODbihuabrushstroke‘two brushstrokes that were incorrectly written’(ii) *liangtwohuaCLstroke[RC xie-cuowrite-wrongde]MOD/0strokeb. (i) shitenzhangCL[RC pingshenjudgetoucastde]MODxuanpiaoballot‘ten votes that judges (in a contest) cast’(ii) *shitenpiaoCLvote[RC pingshenjudgetoucastde]MOD/0votec. (i) [RC tamentheyyin-cuoprint-wrongde]MODyioneyepage‘a page that they printed incorrectly’(ii) *yioneyeCLpage[RC tamentheyyin-cuoprint-wrongde]MOD/0pageSecond, we would predict that the plural “classifier” xie can appear withcovert nouns. This is not the case (16), even there is a suitable context.(16) a. *Xiaohaichildxie-cuo-lewrite-wrong-PFVyionexiePL/0stroke.Intended: ‘The child wrote some strokes incorrectly.’ (Con-text: observing a child who is just starting to learn Chinesecharacters.)b. *ZhethisgeCLcansaizheparticipantjingranunexpectedlyhaistillnengcanyingdewinyionexiePL/0vote.Intended: ‘This participant still somehow managed to winsome votes.’ (Context: describing someone who performedpoorly in a contest and should have not won any votes fromthe judges)Third, an argument from ellipsis.• Generalization: ellipsis requires some degree of semantic identity be-tween elided constituent and antecedent (17a) (an influential hypoth-esis: e-GIVENness, Merchant 2001; see also Gengel 2007; Huang &Mendes 2019, among many others).3• If so, covert nouns should license ellipsis of overt nouns that are se-mantically identical (17b).• But ellipsis is not possible here.(17) a. ZhethisgeCLzicharactershao-lemissedliangtwogeCLbihua,brushstrokenathatgeCLzicharacterzewhileshao-lemissedyionegeCLbihua.brushstroke‘This character is missing two strokes, while that characteris missing one (stroke).’(The first bihua licenses ellipsis of the second bihua)b. #ZhethisgeCLzicharactershao-lemissedliangtwohuastroke/0brushstroke, nathatgeCLzicharacterzewhileshao-lemissedyionegeCLbihua.brushstroke(Null /0brushstroke does not license ellipsis of bihua)(18) a. ZhethisgeCLhouxuanrencandidateyigonga.total.ofhuodewin40104010zhangCLxuanpiao,ballotduishourivalzewhilezhionlyhuodewin105105zhangCLxuanpiao.ballot‘This candidate won a total of 4010 votes, while his/her rivalonly won 105 (votes).’b. #ZhethisgeCLhouxuanrencandidateyigonga.total.ofhuodewin40104010piaovote/0vote,duishourivalzewhilezhionlyhuodewin105105zhangCLxuanpiao.ballot3There are arguments that semantic identity is insufficient. For instance, Merchant(2013) argues that structural parallelism is also necessary. But it is not clear that structuralparallelism is the issue here, unless we intend it to include lexical identity.52.5 Interim summary• PNLs are clearly classifiers (C.-C. J. Tang 2005, S.-W. Tang 2012,2013), not nouns.• No clear evidence that there is NP ellipsis or a noun.• Most parsimonious explanation: bipartite expressions are nounlesssyntactically.3 Consequences3.1 Syntax of nominal expressionsRecent work by Bruening 2009; Bruening et al. 2018 (reviving Chomsky1970) argue that nominal expressions are NPs – headed by nouns.• Bruening et al. (2018) apply this analysis to classifier languages likeKorean and Vietnamese (19).• This might be the case for Korean and Vietnamese (but see Simpson& Ngo 2018), but difficult to extend to noun-less bipartite expres-sions in Chinese.(19) NPD N’ClPNum ClN(Bruening et al. 2018, p. 2)In contrast, Mandarin bipartite expressions can be easily accommodatedunder a theory where nominals are headed by a functional head.• For instance, the DP hypothesis (Abney 1987; Szabolcsi 1994; asapplied to Chinese, see Cheng & Sybesma 1999; Huang et al. 2009;Zhang 2013, 2019, among others).Also possible: a hybrid theory: ‘bipartites’ are headed by functionalheads, ‘tripartites’ with overt nouns are headed by NPs.But misses a generalisation: both have the same distribution (setting asidethe data above about their internal structure).• Can appear as arguments of verbs.• Can co-occur with demonstratives (20a).• Can be fronted (20b).(20) a. Zhethis[bipartite yionehua]strokeheandnathat[tripartite yionegeCLbihua]brushstrokedouallxiewritecuowrongle.PRF‘This stroke and that brushstroke are all written incorrectly.’b. Ta3slianeven{[bipartite yionehua1]stroke/ [tripartite yionegeCLbihua1]}brushstrokedouallmeinotxie t1.write‘He didn’t even write a single stroke.’By Occam’s Razor, the data indirectly favour the DP Hypothesis / func-tional head hypothesis over the NP Hypothesis.3.2 On the semantics of classifiersDebate on the semantics of classifiers, nouns, and numerals.• Numerals need to pair up with nouns to express cardinality.• In certain languages, numerals or nouns have some semantic “defi-ciency” that prevents them from pairing up directly.• Classifiers are needed to remedy that deficiency.• Dispute over whether the deficiency lies in numerals or nouns (Bale& Coon 2014).6One perspective, “classifiers for nouns” (e.g. Chierchia 1998a,b, see alsoRothstein 2010).• Numerals can only combine with count nouns.• Mandarin only has mass nouns (some equivalent thereof), while En-glish has count and mass nouns.• A classifier ‘individuates [mass nouns] to a level suitable to count-ing.’ (Chierchia 1998a, p. 93)• In formal terms, following Bale & Coon 2014, pp. 696–697:(21) English [[two]] = Mandarin [[liang]]= λPλx. ATOMIC(P).{x :*P(x) & µ#(x) = 2}(22) a. [[brushstroke]] = {x : ATOM(x) & BRUSHSTROKE(x)}b. [[bihua]] = ∩BRUSHSTROKE (the “brushstroke kind”)(23) Classifier [[ge]] = ∪ (a function from kinds to sets of atoms)Another perspective: “classifiers for numerals”: variation in the semanticsof numerals.• All languages use a measure function to express cardinality.• English numerals have this measure function “built in” (Krifka1995).• Mandarin numerals do not, and so require a classifier to supply it.• Adapting Wilhelm 2008, p. 55:(24) a. [[two]] = λPλx.[P(x) & OU(x) = 2]OU = function that gives the cardinality of ‘object units’b. [[liang]] = 2(25) [[brushstroke]] = [[bihua]] = λx.[BRUSHSTROKE(x)](26) [[ge]] = λnλPλx.[P(x) & OUgeneral(x) = n]Both proposals presuppose the presence of numerals and nouns — currentdata pose a problem.One conclusion one might draw is that the cross-linguistic distribution ofclassifiers cannot be boiled down to cross-linguistic differences in nounsemantics (or numeral semantics).• Maybe there is nothing deep here. Languages just vary in whethertheir nominal expressions require classifiers.3.3 Salvaging the hypotheses?I would like to suggest another way of thinking about the hypotheses. Andif we do that, there might be an argument in favor of the “classifiers fornumerals” hypothesis.The two hypotheses can be re-framed as claims about classifier semantics:A central part of the “classifiers for numerals” proposal is that the denota-tion of classifiers contains a measure function. Classifiers take numeralsand nouns as arguments (26).So why is there no noun in these ‘nounless’ expressions?4• Hypothesis: the classifiers here (PNLs) have a rich enough semanticsin their own right, so no need for a noun argument.– For example, hua’s denotation contains the measure functionand a noun-like semantics that approximately means ‘brush-stroke.’In contrast, the intuition behind the “classifiers for nouns” hypothesis:numerals occur with a nominal if it has count noun semantics.4There are also instances where classifiers do not appear with an overt numeral inMandarin and other varieties of Chinese, as Bale & Coon (2014); Zhang (2019); Cheng &Sybesma (1999); Lü (1990 [1944]) observe. But one could argue on reasonable groundsthat there is a null numeral (Zhang 2019) or null (in)definiteness operator that the classifiertakes as an argument (Bale & Coon 2014).In addition, under the “classifiers for numerals” hypothesis, classifiers are usually as-sumed to form a constituent with numerals before merging with nouns. In the contextof Mandarin, this would mean that nominals are left-branching. However, I note that thehypothesis, as adapted in this section, does not make as strong a commitment.7• Classifiers are best thought of as operators so that a predicate can getcount noun semantics (such as Rothstein 2010; Bale & Coon 2014,implicitly in Chierchia 1998a).So, to account for the ‘nounless’ data:• Option 1: All classifiers, including PNLs, are operators. PNLs ap-pear with covert mass nouns (or something equivalent).– Difficult to reconcile with the data above.• Option 2:– Classifiers fall into two distinct semantic classes: most are op-erators; the rest (PNLs) have count noun semantics.– Disjunctive: logically coherent, but arguably not aestheticallyideal.If this argument goes through, Mandarin would constitute another lan-guage that supports the ‘classifiers for numerals’ hypothesis (Dëne Su˛łiné,Wilhelm 2008; Mi’gmaq and Chol, Bale & Coon 2014; Japanese, Sudo2016a).4 Conclusion• Reviewed properties of Chinese ‘nounless’ nominal expressions.• Argued that these expressions do not contain a noun in the syntax:the lexical item that follows the numeral is a classifier.• Showed that they pose a problem for recent proposals that nominalexpressions are headed by N.• Showed that they pose a problem for certain theories about thecross-linguistic distribution of classifiers, but argued that the prob-lem might not be fatal for a ‘classifiers for numerals’ proposal.ReferencesAbney, Steven. 1987. The English noun phrase in its sentential aspect. Doctoraldissertation, MIT.Allan, Keith. 1977. Classifiers. Language 53:285–311.Bale, Alan, & Jessica Coon. 2014. Classifiers are for numerals, not for nouns:Consequences for the mass/count distinction. Linguistic Inquiry 45:695–707.Bruening, Benjamin. 2009. Selectional asymmetries between CP and DP suggestthat the DP hypothesis is wrong. University of Pennsylvania Working Papersin Linguistics 15.Bruening, Benjamin, Xuyen Dinh, & Lan Kim. 2018. Selection, idioms, and thestructure of nominal phrases with and without classifiers. Glossa: a journal ofgeneral linguistics 3.Chao, Yuan-Ren. 1968. A grammar of spoken Chinese. Berkeley: University ofCalifornia Press.Cheng, Lisa Lai-Shen, & Rint Sybesma. 1999. Bare and not-so-bare nouns andthe structure of NP. Linguistic Inquiry 30:509–542.Chierchia, Gennaro. 1998a. Plurality of mass nouns and the notion of “semanticparameter”. In Events and grammar, ed. Susan Rothstein, 53–104. Dordrecht:Kluwer Academic Publishers.Chierchia, Gennaro. 1998b. Reference to kinds across language. Natural Lan-guage Semantics 6:339–405.Chomsky, Noam. 1970. Remarks on nominalization. In Readings in Englishtransformational grammar, ed. Roderick Jacobs & Peter Rosenbaum, 184–221. Waltham, MA: Ginn & Co.Gengel, Kirsten. 2007. Focus and ellipsis: A generative analysis of pseudo-gapping and other elliptical structures. Doctoral dissertation, University ofStuttgart.Huang, C.-T. James, Y.-H. Audrey Li, & Yafei Li. 2009. The syntax of Chinese.Cambridge: Cambridge University Press.8Huang, C.-T. James, & Masao Ochi. 2014. Remarks on classifiers and nominalstructure in East Asian. In Peaches and plums, ed. C.-T. James Huang & Feng-Hsi Liu, 53–74. Taipei: Institute of Linguistics, Academia Sinica.Huang, Nick, & Gesoel Mendes. 2019. On pronominalization and ellipsis inclausal idioms. In Proceedings of NELS 49, ed. Maggie Baird & JonathanPesetsky, 123–130. Amherst, MA: GLSA.Kayne, Richard S. 2007. Silent years, silent hours. In Movement and silence.Oxford: Oxford University Press.Krifka, Manfred. 1995. Common nouns: A constrastive analysis of English andChinese. In Events and grammar, ed. Gregory N. Carlson & Francis JeffryPelletier, 398–411. Chicago: University of Chicago Press.Li, Yen-hui Audrey. 1999. Plurality in a classifier language. Journal of EastAsian Linguistics 8:75–99.Longobardi, Giuseppe. 1994. Reference and proper names: A theory of N-movement in syntax and logical form. Linguistic Inquiry 25:609–665.Lü, Shuxiang. 1990 [1944]. “Ge-zi de yingyong fanwei, fulun danweici qian yi-zide tuoluo [the uses of ge and omission of yi before classifiers]. In Lü Shux-iang Wenji [Collected works of Lü Shuxiang], volume 2, 144–175. Beijing:Commercial Press.Merchant, Jason. 2001. The syntax of silence. Oxford: Oxford University Press.Merchant, Jason. 2013. Voice and ellipsis. Linguistic Inquiry 44:77–108.Rothstein, Susan. 2010. Counting and the mass/count distinction. Journal ofSemantics 27:343–397.Simpson, Andrew, & Binh Ngo. 2018. Classifier syntax in Vietnamese. Journalof East Asian Linguistics 27:211–246.Sudo, Yasutada. 2016a. Countable nouns in Japanese. In 11th Workshop onAltaic Formal Linguistics (WAFL 11).Sudo, Yasutada. 2016b. The semantic role of classifiers in Japanese. BalticInternational Yearbook of Cognition, Logic and Communication 11:1–15.Szabolcsi, Anna. 1994. The noun phrase. In The Syntactic Structure of Hungar-ian, ed. Ferenc Kiefer & Katalin Kiss. San Diego, CA: Academic Press.Tang, C.-C. Jane. 2005. Nouns or classifiers: a non-movement analysis of clas-sifiers in Chinese. Language and Linguistics 6:431–472.Tang, Sze-Wing. 2012. Zai shuo “nian, yue, ri” [Chinese temporal nominals"year", "month", and "day": revisited]. Yuyan jiaoxue yu yanjiu [LanguageTeaching and Linguistic Studies] 2:39–43.Tang, Sze-Wing. 2013. ‘Nian, yue, ri’ deng ji qi jufa tedian [Chinese tempo-ral nominals “year,” “month,” and “day” etc. and their syntactic properties].In Zou xiang dangdai qianyan kexue de xiandai hanyu yufa yanjiu, ed. ShenYang, 61–69. Beijing: Commercial Press.Watanabe, Akira. 2012. Measure words as nouns: A perspective from silentyears. Studia Linguistica 66:181–205.Wilhelm, Andrea. 2008. Bare nouns and number in Dëne Su˛łiné. Natural Lan-guage Semantics 16:39–68.Yang, Rong. 2001. Common nouns, classifiers, and quantification in Chinese.Doctoral dissertation, Rutgers University.Zhang, Niina Ning. 2013. Classifier structures in Mandarin Chinese. Berlin:Walter de Gruyter.Zhang, Niina Ning. 2019. Complex indefinites and the projection of DP in Man-darin Chinese. Journal of East Asian Linguistics 28:179–210.5 Appendix5.1 PNLs with the classifier geSome time-interval PNLs can appear optionally with the classifier ge, e.g.xiaoshi / zhongtou ‘hour’, xingqi / libai ‘week.’ Some speakers reportsimilar alternations for tian ‘day’ and nian ‘year.’(27) liangtwo(ge)CLxiaoshihour‘two hours’9It is possible that these terms are syntactically ambiguous between nounand a classifier; when analyzed as a noun, the classifier ge appears (fol-lowing C.-C. J. Tang 2005).5.2 Distinguishing PNLs from classifier omissionMandarin allows classifiers to be omitted in formal registers (e.g. newsheadlines), producing Num Noun sequences.(28) [Num Liang]two[Noun du-fan]drug-addictjiechihold[Num wu]five[Noun renzhi]hostage‘Two drug addicts hold five people as hostages.’But PNLs are acceptable in non-formal registers, and so are not instancesof classifier omission.• Rules out an analysis like (3b).(3) b.NumCL/0NPnian/hua5.3 Relative clauses, xie, and covert nouns(15) and (16) show us that relative clauses and xie do not co-occur withcovert nouns (if they exist).Can we explain these examples by claiming that relative clauses and xiehave some idiosyncratic requirement for an overt head noun?• Not quite: relative clauses and xie can co-occur with elided NPs.(29) a. LisiLisimai-leboughtyionexiePLshu,bookwoIyealsomai-leboughtyionexiePLshu.book‘Lisi bought some books, and I also bought some (books).’b. [RC Niyouchangsingde]MODgesongbunotbithan[RC woIchangsingde]MODgesonghaogoodting.hear‘The song you sang is not as nice as the one I sang.’10


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items