Morphosyntactic vagueness and directionality
-
Yueh Hsin Kuo
Abstract
This paper shows that morphosyntactic vagueness exists between several minor morphosyntactic categories in Mandarin (modal, protasis connective and types of classifier): these categories share formal distributional properties and functional similarities such that their morphosyntactic distinctions become neutralized in some contexts. Consequently, bidirectionality rather than unidirectionality may characterize these categories diachronically. This paper proposes that any direction of change that is motivated by morphosyntactic vagueness should be regarded as regular and systematic, even if it is not unidirectional, and identifies some directions for future research on morphosyntactic vagueness in diachrony.
1 Introduction
Vagueness is well known in lexical semantics (e.g. Tuggy 1993), where the meaning of a word is underspecified in some respects. For example, parent is vague with respect to its referent’s gender. Recently, Denison (2017) has extended the notion of vagueness to word class, or morphosyntactic category: the morphosyntactic category of a word may be underspecified as well. For example, expert in expert advice is vague between noun and adjective (morphosyntactic categories are referred to in small capitals). In cognitive linguistics, where there is no strict division between grammatical and lexical items, this is a welcome extension, as by hypothesis phenomena associated with lexical items are expected to be found elsewhere, such as in major morphosyntactic categories (e.g. noun and adjective). Denison (2017) suggests that contexts where morphosyntactic vagueness is observed are likely syntactic bridging/critical contexts (Heine 2002; Diewald 2002), which are contexts where the original morphosyntactic analysis of an item co-exists with a new one. For example, a noun may be developing into an adjective when it is bare and immediately precedes and modifies another noun (e.g. expert in expert opinion, but not the expert’s opinion). While in Present-Day English, change from noun to adjective may seem more common, Denison (2017) notes that expert, synchronically more noun-like, has had a longer history as an adjective.
Bringing together previous research, this paper demonstrates that morphosyntactic vagueness can be found between not just major, but also minor morphosyntactic categories, such as modal and protasis connective and types of classifier in Mandarin. Their functional properties are as follows (their morphosyntactic properties will be detailed later). modal “is a linguistic category referring to the factual status of a proposition” (Narrog 2012: 6); in English it is expressed by can, should and must, among others. protasis connective, such as if, is typically part of a bi-clausal conditional construction (if… then) that marks the proposition referred to by the apodosis (the then-marked clause) as contingent on the proposition referred to by the protasis (the if-marked clause); for a cognitive description, see Sweetser (1990). classifier is a category of expressions that are units used in counting and measuring, whose close equivalents in English include group in a group of students and bag in a bag of flour (Ahrens & Huang 2016; Chen 2017; Kuo 2020a).
Moreover, bidirectional changes between these minor morphosyntactic categories are also possible, which is not how grammatical change is conceptualized typically: the grammaticalization tradition suggests that the development of grammatical categories is largely unidirectional; i.e. they become more grammatical (e.g. Kuteva et al. 2019). It is proposed here that, first, change that is motivated by morphosyntactic vagueness leads to grammatically equivalent morphosyntactic categories. Second, any direction of change that is motivated by morphosyntactic vagueness should be considered as a regular kind of change, even if it is not unidirectional, as morphosyntactic vagueness is a systematic phenomenon.
This paper is structured as follows. Section 2 discusses morphosyntactic vagueness in more detail. Section 3 introduces modal and protasis connective in Mandarin, their morphosyntactic vagueness and bidirectional changes between them. Similarly, Section 4 demonstrates morphosyntactic vagueness and bidirectional changes between two types of classifier. Section 5 discusses grammaticalization, unidirectionality and implications of morphosyntactic vagueness for directionality in historical linguistics.
2 Vagueness
Vagueness and ambiguity are often contrasted. In Denison’s (2017: 293) words,
“Ambiguity is where AD/R [addressee/reader] (and linguist) cannot be sure which of two or more linguistic possibilities was intended by SP/W [speaker/writer] and something hangs on the choice. Vagueness is where a linguistic analysis is in some relevant respect underdetermined, at least for AD/R (perhaps for SP/W, too), but no further information is needed for interpretation.”
In other words, both vagueness and ambiguity are characterized by formal identity on a certain level, but they have different functional dimensions. Ambiguity is associated with more than one function or meaning, the distinction of which is at stake. While vagueness may also be associated with more than one function or meaning, their distinction is not at stake. Building on Denison’s work, it is proposed here that two conditions characterize morphosyntactic vagueness: shared formal distributional properties (i.e. a type of ‘formal identity’) and functional similarities.
For example, Denison (2017) notes that a noun in English may occur as a prenominal modifier (a fun story) or a predicative complement (it is fun) and so may an adjective. When it does, its morphosyntactic interpretation may be vague between noun and adjective, especially if there is no category-specific marker. The prenominal modifier and the predicative complement slots without any co-occurring category-specific marker, therefore, are where noun and adjective share formal distributional properties. This does not mean that there is no formal distributional distinction between noun and adjective, but that it is neutralized in some contexts. For example, such typically precedes a noun phrase, so fun in such fun is more noun-like; similarly, fun in a very fun story is rendered more adjective-like by very. The absence of a category-specific marker such as such or very in a fun story and it is fun therefore makes fun vague.
The potential vagueness between noun and adjective also does not mean that whenever a noun or adjective is used as a prenominal modifier or a predicative complement, its morphosyntactic category must be vague or will start to change. Functionally, for a noun to be interpreted as vague between noun and adjective, it has to be associated with semantic gradability, which renders it more adjective-like (Denison 2017: 305). For an adjective to be interpreted as vague between noun and adjective, presumably, it has to be construed referentially, i.e. more noun-like, too. In other words, the morphosyntactic categories, noun and adjective, are vague in contexts where their instances share formal distributional properties and functional similarities.
Denison (2017) focuses more on individual words and their morphosyntactic vagueness. Assuming morphosyntactic categories are generalizations over constructions (Croft 2001; Denison 2018: 139–140; Diessel 2019), all cases of morphosyntactic vagueness hypothetically could be traced back to constructions whose slots are morphosyntactically underdetermined, with varying degrees of lexical specificity. For example, see Denison (2018) with respect to the morphosyntactic vagueness of long from a constructional perspective. Sections 3 and 4 will identify constructional slots as origins of morphosyntactic vagueness between several minor morphosyntactic categories.[1]
3 Modal and protasis connective
Modal and protasis connective are traditionally considered as distinct morphosyntactic categories. However, Section 3.1 shows that they are morphosyntactically vague in Mandarin. No modal or protasis connective consistently appears in slots where they are distinguished and both are also known for their functional similarities. Section 3.2 identifies specific constructional slots where morphosyntactic vagueness occurs and enables bidirectional changes between instances of modal and protasis connective in Mandarin. Section 3.3 reviews both directions of development from a crosslinguistic perspective.
3.1 Morphosyntactic vagueness between modal and protasis connective
Section 3.1.1 focuses on shared formal distributional properties. Section 3.1.2 reviews functional similarities.
3.1.1 Shared formal distributional properties
The most typical construction that a modal appears in can be represented as [(subject) modal vp], as in (1). The subject can be null, which is frequent in a pro-drop language like Chinese (Li & Thompson 1976). A protasis connective precedes the predicate and can be either pre-subject or post-subject. The conditional construction that a protasis connective is typically part of can be represented as [(subject) protasis-connective vp…], as in (2) or [protasis-connective (subject) vp…], as in (3). The apodosis is not represented here.[2]
| (1) | 汝須自看 | ||||
| rǔ | xū | zì | kàn | ||
| you | must | self | look | ||
| ‘You must look at it yourself.’ | Zǔtángjí (10th century) | ||||
| (2) | 若你不管… | ||||
| ruò | nǐ | bù | guǎn | ||
| ‘If you do not care…’ | Hónglóumèng (1791) | ||||
| (3) | 你若不信…. | ||||
| nǐ | ruò | bù | xìn | ||
| you | if | not | believe | ||
| ‘If you do not believe it…’ | Hónglóumèng (1791) | ||||
The category of modal in Mandarin by itself is vague. Two subtypes are commonly proposed on the basis of distributional evidence other than [(subject) modal vp]: modal auxiliary and modal adverb. In what follows, Li & Thompson’s (1981: Ch. 5) distributional criteria for modal auxiliary will be examined because they are rather stringent (e.g. Peng 2007: Ch. 2). If stringent criteria are not effective at separating modal from protasis connective, it follows that less stringent ones (e.g. those for modal adverb) are even less so, and modal and protasis connective may therefore be taken to share distributional properties.
The criteria for modal (auxiliary) are organized into sets in (1).
| (4) | A modal | |
| (a) | can occur in the polar question construction and the negation construction; | |
| (b) | must take a following verbal complement (unless pragmatically recoverable); | |
| (c) | does not appear in any aspect, intensifying, nominalization or direct objection construction; | |
| (d) | cannot be pre-subject.[3] | |
(4a) is effective at distinguishing a modal from a protasis connective (as the latter occurs in neither polar question nor negation constructions), but not consistently so: no modal occurs only in these two constructions. Furthermore, (4a) does not uniquely define modal, as many kinds of verb occur in the two constructions (Li & Thompson 1981: Ch. 5). (4b) does not effectively distinguish between modal and protasis connective because a protasis connective can immediately precede the predicate and if it does not, the subject can be null, which would make it appear as if the protasis connective takes a following verbal complement. (4c) cannot be used to distinguish between modal and protasis connective. Like a modal, a protasis connective also does not occur in the constructions listed in (4c). (4d) is only effective at distinguishing a modal from a protasis connective when the latter is pre-subject. (4d) is not consistently reliable since a protasis connective can be post-subject and in null-subject contexts, frequent in Mandarin, any distinction between pre-subject and post-subject positions is neutralized.[4]
Eifring (1995: 54–55) proposes three distributional criteria that distinguish protasis connective and (non-modal) adverb from other morphosyntactic categories: they may immediately precede the sentence-final pause-marking particle ne, the verb of saying shuō and the copula shì. Nevertheless, they are not uniquely associated with protasis connective, as shuō and shì can be the complements of modals and ne typically follows nominal expressions. Moreover, Kuo (Forthcoming b) shows that in the Academia Sinica Balanced Corpora of Modern Chinese (Present-Day Mandarin), henceforth the Sinica Corpus, only one protasis connective, rúguǒ, precedes ne, and there is only one such instance. Rúguǒ and ne have a mutual information value of −1.84, which suggests that they rarely co-occur. In the Sinica Corpus of Early Mandarin, there is none. Collocation with ne therefore is panchronically rare.
One distributional context where a modal and a protasis connective are consistently distinguished is where they co-occur, i.e. in a modalized protasis, which can be represented as protasis-connective (subject) modal or (subject) protasis-connective modal (e.g. ‘if you can/must…’). Nevertheless, modalized protases are rare. A simple collexeme analysis (Stefanowitsch & Gries 2003) indicates that only 2 out of the 13 Present-Day Mandarin modals, as defined by Li & Thompson (1981), are significantly attracted to the protasis (p < .05, Fischer-Yates Exact Test), while 8 are significantly repelled, according to Kuo (2020b, Forthcoming b) with data from the Sinica Corpus. The two significantly attracted ones, néng ‘be able to; can’ and nénggòu ‘be able to; can’, show signs of univerbation with protasis connectives (Kuo 2021), which suggests that they are not purely modal inside protases. Similarly, according to Kuo (Forthcoming b), in the Sinica Corpus of Old Chinese, only 2 out of the 10 modals are significantly attracted, one of which is néng ‘be able to; can’ and the other is close to the significance threshold (p ≈ .049). This suggests that panchronically non-attraction is the norm and modalized protases are rare.
In sum, modal and protasis connective in Mandarin lack consistent distributional distinction. They share distributional properties in many contexts, particularly null-subject ones. This is rather different from English. A modal auxiliary in English participates in subject-auxiliary inversion, but a protasis connective does not; a modal is typically post-subject, unless inverted, and a protasis connective is pre-subject. English is typically not a pro-drop language (unlike Mandarin); therefore, the position of a modal or a protasis connective relative to the subject is more consistently marked. For example, if in if you… is consistently pre-subject, while must in you must is post-subject, unless inverted. A special modal that blurs the line between modal and protasis connective is should (see Section 3.3), but the distinction is in general clearer.
3.1.2 Functional similarities
Functional similarities between modal and protasis connective are well-known. The meaning of modal will be referred to as ‘modality’; that of the protasis connective, ‘conditionality’. Modality and conditionality present propositions as factually underdetermined and have similar uses, such as performative and epistemic ones (e.g. Sweetser 1990; Narrog 2012). In some formal semantic analyses, conditionality is basically modal in nature (e.g. Kratzer 2012).
One might think that modal is typically part of a monoclausal construction and protasis connective is typically part of a biclausal conditional construction, so modal and protasis connective can be consistently distinguished. This is not always true; functional motivations may often override this distinction. The apodosis of a conditional sentence may be elided, especially in speech, and the protasis along with the protasis connective may undergo ‘insubordination’ (Evans 2007) in the sense that they turn into a main clause utterance to serve different functions, including speech acts such as directives (Mauri & Sansò 2011), modality-marking (Cuenca 2020), among others (see Evans & Watanabe 2017). Modality is known to be able to go beyond a single clause and when it does, it may produce a conditional interpretation between multiple clauses (so-called ‘modal subordination’ in formal semantics; Roberts 1989). For example, the inference of conditionality in (5a) may be paraphrased by (5b).
| (5) | a. | An alien might have abducted the cow. He might have eaten it. |
| b. | An alien might have abducted the cow. If an alien abducted the cow, he might have eaten it. | |
| Faller (2017: 55–56; emphasis original) |
Modality and conditionality intersect in (5) because modality is typically relative to some conditions, which may or may not be explicitly expressed, while conditionality brings out more explicitly the condition that a modalized proposition is relative to. For example, the second might in (5a) is relative to the first (explicitly expressed) might, which is relative to the (not explicitly expressed) discourse context where it is possible for aliens to abduct cows. In (5b), if brings out the condition that the second might is relative to.
In conclusion, Mandarin modal and protasis connective share functional similarities and distributional properties. This suggests that they may be morphosyntactically vague. Section 3.2 identifies slots that are vague between modal and protasis connective, thereby enabling bidirectional changes: a modal may turn into a protasis connective and vice versa.
3.2 Bidirectional changes between modal and protasis connective in Mandarin
Kuo (Forthcoming b) proposes that the modal and protasis connective slots, filled by xū ‘must’ and fēi ‘unless’ respectively, are morphosyntactically vague in the following constructions: [(subject)xū pthenq] (then represents a slot for a connective that, not unlike then, expresses conditionality and temporality) ‘must p then q’ and [fēi (subject)p, bù kě] ‘unless p, it is not possible/good’.[5]
First, formally, neither construction contains any category-specific marker that highlights any possible morphosyntactic distinction between modal and protasis connective: no polar question or negation construction indicating that xū or fēi must be a modal; no (non-null) subject indicating that xū or fēi must be a pre-subject protasis connective (assuming Li & Thompson’s criterion); finally, no collocation with ne indicating that xū or fēi must be a protasis connective (assuming Eifring’s criterion). Second, functionally both constructions are indirect speech act constructions in that they direct the addressee to do something via indirect means (for issues regarding indirect speech acts and construction grammar, see Stefanowitsch 2003). Their performative meanings can be understood modally and conditionally; that is, there is ‘performative equivalence’ between their modal and conditional readings.
For example, consider (6) and (7). Neither contains modal-specific or protasis connective-specific morphosyntax. Both are performative and whether xū in (6) or fēi (7) is interpreted as a modal or a protasis connective, the performative meaning is equivalent, in that either reading counts as an indirect act to get the addressee to do something.
| (6) | 須運釿始得 | ||||
| xū | yùn | jīn | shǐ | dé | |
| must | move | axe | only-then | possible | |
| Modal xū: | ‘You must use an axe; only then is it possible.’ | ||||
| Conditional xū: | ‘Only if you use an axe is it possible.’ | Zǔtángjí (10th century) | |||
| (7) | 若孟子詩書等,非讀不可 | |||||||
| ruò | mèngzǐ | shīshū | děng | fēi | dú | bù | kě | |
| like | Mèngzǐ | Shī Shū | so-on | unless | read | not | possible/good | |
| Modal fēi: | ‘Books such as Mèngzǐ, Shī and Shū, unless one reads them, it is not good.’ | |||||||
| Conditional fēi: | ‘You must read them, (or) it is not good.’ | |||||||
| Zhūzǐ Yǔlèi (1270) | ||||||||
The whole constructions therefore also represent critical contexts (Diewald 2002) where category change (i.e. modal > protasis connective and protasis connective > modal) is observed. (8) and (9) exemplify the outcomes of the developments. Xū is a pre-subject protasis connective in (8). Fēi is a monoclausal modal utterance in (9).
| (8) | 須你逐一去看,理會過方可 | ||||||||
| xū | nǐ | zhúyī | qù | kàn | lǐhuì | guòfāng | kě | ||
| if | you | one.by.one | go | look | understand | EXP | only.then | can | |
| ‘Only if you have looked at it one by one and understood it is it possible.’ | |||||||||
| Zhūzǐ Yǔlèi (1270) | |||||||||
| (9) | 我非去 | |||
| wǒ | fēi | qù | ||
| I | must | go | ||
| ‘I must go.’ | Wang (2008:109) | |||
Bidirectional changes (e.g. modal > protasis connective and protasis connective > modal) are also found elsewhere in Chinese. The diachronic source of the connective yào ‘if’ is disputed, but the consensus is that it is modal (Traugott 1985; Yu 1998; Hsu et al. 2015; see Kuo 2021 for a review). Kuo (Forthcoming a) proposes that the deontic modal meaning of bì ‘must’ in Old Chinese, as in (10), is the source of its conditional meaning ‘only if’, as in (11), and argues that a motivating factor is that both modal and conditional meanings perform the same indirect act (i.e. they are performatively equivalent). Kuo’s analysis is grounded in the Invited Inferencing Theory of Semantic Change (Traugott & Dasher 2002), but the syntactic critical context identified by Kuo, where bì precedes an apodosis connective, is amenable to a vagueness analysis: it contains neither modal-specific nor protasis connective-specific morphosyntax, just like its 10th century counterpart, [(subject)xū pthenq], ‘must p then q’, as exemplified in (6).[6]
| (10) | 必有以知天地之恒制，乃可以有天下之成利 | |||||||
| bì | yǒu | yǐ | zhī | tiāndì | zhī | héngzhì | nǎi | |
| must | have | loc | know | universe | poss | law | therefore | |
| kěyǐ | yǒu | tiānxià | zhī | chéng | lì | |||
| can | have | world | poss | pre-existing | benefit | |||
| Guóyǔ (ca. the 5th–4th centuries BC) | ||||||||
| Modal bì: | ‘You must know the universe’s rules, and then you can enjoy all the benefits already out there in the world.’ | |
| Conditional bì: | ‘Only if you know the rules can you enjoy all the benefits.’ |
| (11) | 必民樂其政也,而令乃行 | ||||||
| bì | mín | yuè | qí | zhèng | yě | ér | |
| only.if | people | like | 3ps | government | fp | and | |
| lìng | nǎi | xíng | |||||
| command | therefore | implement | |||||
| ‘Only if the people enjoy their commands be implemented.’ | |||||||
| Guǎnzǐ (ca. the 1st century BC) | |||||||
Similarly, chūfēi is typically categorized as a protasis connective ‘unless; only if’ (Wang et al. 2014), as in (12), but has also developed modal uses, not unlike fēi (Eifring 1995), as in (13). (12) is likely to be a critical context where chūfēi begins to be recategorized as a modal, due to its similarity to xū in [(subject)xū pthenq] (despite variation on then); see (6). In the Sinica Corpus of Early Mandarin, instances of chūfēi that are unambiguously modal are found after the 13th century. An example is (13), where the apodosis, containing chūfēi, suggests that chūfēi is a modal.
| (12) | 除非首尾熟背得方得 | ||||||
| chūfēi | shǒuwěi | shú | bèi | dé | fāng | dé | |
| chūfēi | headtail | familiar | memorize | can | only.then | get | |
| Conditional: | ‘Only if you can commit to memory the whole thing from start to finish can you get it.’ | ||||||
| Modal: | ‘You Must commit it to memory; only then can you get it.’ | ||||||
| Zhūzǐ Yǔlèi (1270) | |||||||
| (13) | 要作遊,除非到南京去 | |||||||
| yào | zuò | yóu | chúfēi | dào | nánjīng | qù | ||
| want | do | journal | must | arrive | Nánjīng | go | ||
| ‘If you want go on a trip, you must go to Nánjīng.’ | ||||||||
| Rúlín wàishǐ (1750) | ||||||||
3.3 Bidirectional changes between modal and protasis connective in other languages
Bidirectional changes are also attested crosslinguistically, although not necessarily within the same languages simultaneously. It has not been claimed that in those languages modals and protasis connectives are morphosyntactically vague, but crosslinguistic parallels are reviewed here to suggest that neither direction of change is peculiar to Mandarin. In Germanic languages, some modals have developed conditional uses that are unlike protasis connectives (e.g. van der Auwera & Plungian 1998). For example, Trousdale (2012: 173) reports examples like (14), where the inflection on the main verb indicates that should heads a protasis.
| (14) | Should Hillary wins [sic] the nomination fairly she will be a very good president. |
Other similar examples include German sollte (Breitbarth 2019) and Dutch mogen ‘may’ and moeten ‘must’ (Byloo & Nuyts 2014), among others.
The other direction of change (protasis connective > modal) is found in various Asian languages, particularly Japanese and Korean (e.g. Akatsuka & Clancy 1993; Traugott & Dasher 2002: 150–151; Kuo 2020b: 150–155). Such modal expressions typically originate from a conditional sentence whose apodosis evaluates the proposition in the protasis and highlights it as deontically necessary, not unlike fēi in (7) (e.g. fēi p bù kě ‘unless p, it is not good’ > fēi p ‘must p’). For example, Fujii (2004) proposes that Japanese -nai to has developed into a protasis connective. Its development can be summarized as p-nai to, ikanai ‘if not p, it is bad’ > p-nai to ‘must p (literally, if not p)’.
| (15) | hayaku | ika-nai | to | |
| quickly | go-not | if | ||
| ‘You must go quickly.’ (literally, ‘If you don’t go quickly.’) | ||||
| Fujii (2004: 125) | ||||
In sum, bidirectional changes exist between modal and protasis connective in Mandarin and other languages.
4 Two types of classifier
Functionally, a classifier is a unit of counting (e.g. Croft 2001). Formally, it immediately follows a schematic numeral slot and precedes an NP slot in Mandarin. The canonical classifier construction can be represented as [numeral classifier np]. For example, yi ge rén, literally ‘one classifier person’, means ‘one person’ and if yi is replaced by sān ‘three’, it means ‘three people’. While ge can be used in counting almost anything in Mandarin, other classifiers are more specialized in that they are counting units for more specific types of things, or indicate specific measurements (cf. kilo in a kilo of water) or arrangements of things (cf. row in a row of trees). Classifiers that occur with a schematic numeral slot will be referred to as canonical classifier.
Approximate measure words (AMWs) are also a type of classifier, found immediately following an invariable numeral yi (Ahrens & Huang 2016). An AMW construction, [yiamw np], denotes an unspecific, approximate amount of NP. For example, diǎn has the lexical meaning of ‘dot; spot’ but as an AMW it means ‘(a) bit (of); some’, e.g. yi diǎn niúnǎi ‘a bit of milk; some milk’ (Ahrens & Huang 2016: 194). AMW constructions resemble the so-called binominal quantifier constructions such as [a lot ofnp] and [a bit ofnp] (e.g. Traugott 2008; Brems 2011). Yi and a may suggest that one singular unit of something is being counted, but in fact both kinds of constructions do not denote a precise amount. An AMW in [yiamw np] and lot in [a lot ofnp] are also ‘uncountable’: yi and a are fixed and cannot be replaced with a numeral, e.g. *sān diǎn yánsè ‘some color; lit. three dots of color’ (Ahrens & Huang 2016: 191); *three lots of NP ‘many/much NP’, unless one intends lot in the sense of ‘plots (of land)’ or ‘groups (of items or people)’.
Binominal quantifier constructions in English and their counterparts in languages such as Spanish (Verveckken 2012), Dutch (Brems 2011) and German (Neels & Hartmann 2018) have relatively transparent lexical sources. So do AMW constructions. For example, diǎn, originally a noun meaning ‘dot; spot’, has become an AMW after being used as a classifier in the canonical construction (Chen 2017; Kuo 2018); i.e. [yi diǎnnp] ‘one dot of NP’ has become [yi diǎnnp] ‘a bit of NP’. Most other AMWs have comparable lexical meanings and canonical classifier uses, so their histories are likely similar to diǎn. For example, AMW dūi has the lexical meaning of ‘heap; pile’ (Ahrens & Huang 2013); compare un montón de NP ‘a heap of NP’ in Spanish (Verveckken 2012). That is, in very general terms, [yiamw np] ‘an approximate amount of NP’ develops from [yiclassifier np] ‘one unit of NP’, which is not unlike binominal quantifier constructions, e.g. [a lot ofnp] ‘many/much NP’ < [a lot ofnp] ‘a parcel/plot of NP, typically land; a group of items or people’ (Traugott 2008: 230–232). One may therefore hypothesize that change from canonical classifier to AMW (and its crosslinguistic parallel ‘noun > quantifier’) is typical.
However, there is one counterexample. Xiē ‘some’ is typically an AMW. For some speakers, it can appear in a canonical classifier construction with a relatively more schematic numeral slot. That is, xiē has become ‘countable’. For example, it immediately follows liǎng ‘two’ and means ‘group; type’ in (16)–(18).
| (16) | 中國的文人們有兩「些」, 一些…「別的一些文們」… | |||||||
| zhōngguó | de | wénrénmen | yǒu | liǎng | xiē | yi | xiē… | |
| China | poss | scholar.pl | there.is | two | type | one | type | |
| Bié | de | yi | xiē | wénrénmen | ||||
| other | poss | one | type | scholar.pl | ||||
| ‘There are two types of Chinese scholars, one type… the other type of scholars…’ | ||||||||
| Hǎiyàn Monthly (Jan. 1936) | ||||||||
| (https://zh.m.wikisource.org/zh-hant/文人比较学) | ||||||||
| (Accessed 15 May 2021) | ||||||||
| (17) | 最近在团体遇两些人⋯有一些人⋯另一些人 | |||||||
| zuìjìn | zàituántǐ | yùdào | liǎng | xiē | rén… | yǒu | yi | |
| recently | in | groupmeet | two | type | person | there.is | one | |
| xiē | rén… | lìng | yi | xiē | rén | |||
| type | person | other | one | some | person | |||
| ‘Recently I met two types of people in group meetings there’s one type of people…the other type of people…’ | ||||||||
| (http://happyconan3.blogspot.com/2011/03/blog-post_20.html) (Accessed 15 May 2021) |
||||||||
| (18) | 我们的生命中有两些人⋯ 一些人另一些人 | |||||||
| womende | shēngmìng | zhòng | yǒu | liǎng | xiē | rén… | yi | |
| we.poss | life | inside | there.is | two | type | person | one | |
| xiē | rén | lìng | yi | xiē | rén | |||
| type | person | other | one | type | person | |||
| ‘In our lives there are two types of people… one type of people… the other type of people…’ | ||||||||
| (http://blog.sina.com.cn/s/blog_701373400100l1kw.html) (Accessed 15 May 2021) |
||||||||
Without any diachronic data, reconstructing the history of xiē on the basis of what we know about similar Mandarin and crosslinguistic expressions would lead one to propose xiē as a canonical classifier in origin, as in (16)–(18), while the AMW use ‘some’ is a later development. But this is incorrect. Diachronic data clearly indicate AMW xiē to be historically earlier, while the non-AMW use is later (Kuo 2020a, 2020b), i.e. chronologically, xiē ‘some’ > yi xiē ‘some’ > yi/liǎng xiē ‘one/two groups; types’. The history of [yi xiēnp] ‘some’ > ‘group’ therefore is not unlike the reversal of that of [a lot ofnp], ‘parcel; plot; group’ > ‘many/much’. Functional motivations behind the history of xiē have been proposed, by evoking analogy between the meaning of xiē and those of the network of related classifier constructions (Kuo 2020a, 2020b).
What has not been elaborated on is that amw and canonical classifier are actually morphosyntactically vague, and therefore bidirectional changes are not unexpected: both share distributional and functional properties. In constructional terms, the ‘X’ slot in [yi X np] is morphosyntactically vague between amw and canonical classifier. Distributionally, even though AMWs only appear after yi, canonical classifiers can follow it, too. Distributional tests have been proposed to distinguish types of classifier, yet none is consistently reliable and no classifier consistently appears in such distributional contexts (e.g. see Zhang 2013 for a review). Functionally, AMWs and their close canonical counterparts (classifiers after yi) are counting units and similar in their respective constructions (e.g. ‘an approximate amount of NP’ and ‘one unit of NP’). The co-existence of canonical and AMW uses further accentuates this functional similarity; all AMWs listed by Ahrens & Huang (2016) can be found as canonical classifiers. For example, depending on what fills in NP and context, [yi diannp] may be interpreted as ‘one dot of NP’, and/or ‘a bit of NP’, similar to [a lot ofnp], where lot could mean ‘plot’ and/or ‘much’ if the NP is land.[7]
5 Discussion
Section 5.1 introduces grammaticalization and unidirectionality. Sections 5.2 discusses how to interpret the case studies in terms of directionality. Section 5.3 considers grammaticalization in light of Section 5.2.
5.1 Grammaticalization and unidirectionality
Grammaticalization can be defined generally as “the genesis and development of grammatical forms” (Kuteva et al. 2019: 3), or more specifically as “the change whereby lexical items and constructions come in certain linguistic contexts to serve grammatical functions and, once grammaticalized, continue to develop new grammatical functions” (Hopper & Traugott 2003: xv). Grammaticalization has been so closely associated with the unidirectionality hypothesis that “it has been seen as the core property of grammaticalization in most of the literature” (Börjars & Vincent 2013: 164), even though unidirectionality is not part of the definitions cited above. The hypothesis states that grammatical change is predictable, as a grammatical item or construction typically originates from something lexical or less grammatical, and as it develops, becomes even more grammatical, but not less so (e.g. Hopper & Traugott 2003; Kuteva et al 2019). Even though unidirectionality has been conceptualized in slightly different ways,[8] it is defined here as the diachronic relationship between members of lexical and grammatical (i.e. major and minor morphosyntactic) categories (e.g. Norde 2009: Ch. 2; Börjars & Vincent 2013: 165–166). This relationship is very often represented in the literature as “clines” or “pathways” from one (less grammatical) to another (more grammatical) category, e.g. main verb > modal auxiliary.
Unidirectionality is a powerful idea because it suggests that the history of two similar and likely related items can be reconstructed, even when there is little historical data: the less grammatical one should be the diachronically older one. Unidirectionality has inspired many theories (see Narrog 2012 for a review) to account for the fact that languages across time and space seem to have undergone similar grammatical developments. Nevertheless, exceptions to unidirectionality do exist and are discussed under the rubric of ‘counterdirectionality’ or ‘degrammaticalization’, namely the process whereby a grammatical item or construction becomes ‘less grammatical’ (Norde 2009).
5.2 Morphosyntactic vagueness and directionality
The case studies in Sections 3 and 4 are instances of bidirectionality: given two morphosyntactic categories, diachronically an instance of one of them can become an instance of the other and vice versa. How is this bidirectionality, which is motivated by morphosyntactic vagueness, to be understood with respect to unidirectionality and counterdirectionality? One could view one of the morphosyntactic categories as more grammatical and the other as less grammatical. For example, in some hierarchical theories of clause structure, clause periphery is associated with higher grammaticality (e.g. Narrog 2012; Breitbarth 2019). Therefore, protasis connectives, typically being at clause periphery, are more ‘grammatical’ than modals. This suggests modal > protasis connective is unidirectional (see also Breitbarth 2019, who analyses it as grammaticalization), while protasis connective > modal is counterdirectional. Quantifiers are scalar and procedural (e.g. see Traugott & Trousdale 2013 for a lot of and references cited in Section 4) and less open-ended than nouns. Therefore, the quantifier-like AMW xiē may be more grammatical than the noun-like canonical classifier xiē. It then follows that AMW > canonical classifier is counterdirectional.
However, it is proposed here that unidirectionality should not be proposed for any of the case studies. This is because they are motivated by morphosyntactic vagueness, which, by definition, indicates equivalence (or symmetry) between morphosyntactic categories. That is, distributional and functional similarities exist between morphosyntactic categories to the extent that their distinction is underdetermined and not at issue in some contexts. If we take morphosyntactic vagueness seriously, we should not propose unidirectionality between morphosyntactically vague categories, as such a proposal presupposes grammatical non-equivalence (or “asymmetry”; Börjars & Vincent 2013: 174–175). It then follows that neither case study should be construed as counterdirectional, as counterdirectionality also entails non-equivalence (i.e. the source is more grammatical than the outcome). Furthermore, counterdirectionality is characterized as unusual, as unidirectionality is the norm crosslinguistically (Kuteva et al. 2019). Consequently, construing any of the case studies here as counterdirectional does not do justice to the fact that morphosyntactic vagueness is a ‘systemic’ condition: it is a linguistic fact that some categories share distributional and functional similarities.[9]
One might argue that despite the ‘systematicity’ of morphosyntactic vagueness, the sole example of AMW > canonical classifier and its lack of any crosslinguistic parallel suggest that it is likely counterdirectional. Nevertheless, the supposed scarcity of AMW > canonical classifier (or its crosslinguistic parallel, quantifier > noun) could be due to a bias for unidirectionality in the literature. Moreover, whether a particular direction is more frequently attested or not crosslinguistically, if morphosyntactic vagueness motivates both directions of change, unidirectionality, at least given its definition here, is not warranted. This may seem like a strong statement, but follows naturally from the definition of vagueness: “Vagueness is where a linguistic analysis is in some relevant respect underdetermined, at least for AD/R (perhaps for SP/W, too), but no further information is needed for interpretation” (Denison 2017: 293).[10]
In short, morphosyntactic vagueness can motivate bidirectionality, which should not be understood as unidirectional or counterdirectional, as morphosyntactic vagueness describes grammatical equivalence, rather than non-equivalence, which unidirectionality and counterdirectionality entail.
5.3 Morphosyntactic vagueness and grammaticalization
Is it possible to consider the case studies in terms of grammaticalization? Yes, if grammaticalization is not defined strictly as unidirectional; the definitions cited in Section 5.1 actually do not mention unidirectionality. Therefore, the evolutions of grammatically equivalent categories can be considered as grammaticalization. This is not a novel idea. For example, Company (2018: 369–371) proposes that intra-category grammatical changes may be a kind of grammaticalization that involves no change in grammatical status, whose directionality therefore is “neither down nor up” (Company 2018: 361).[11] Summarizing the development of temporal mientras ‘while’ > conditional mientras ‘if’ in Spanish, Company (2018: 371) notes that “Mientras retains the same category, a conjunction of subordination, but with the distribution and meaning changed.”
However, it may be simply a definitional issue whether one labels vagueness-motivated change as grammaticalization or not. It is grammaticalization if grammatically equivalent categories are defined as possible sources and outcomes in grammaticalization, or if grammaticalization is defined broadly as the development of any grammatical form that may or may not be unidirectional. It is not grammaticalization if unidirectionality is built into the definition, such that the outcome must be more grammatical than the source in grammaticalization. For definitional issues regarding grammaticalization, see von Mengden & Simon (2014).
In short, whether change that is motivated by morphosyntactic vagueness can be considered as grammaticalization is open to interpretation, which will not be attempted here. It is assumed here that one of a diachronic cognitive linguist’s tasks is to provide cognitively plausible accounts of motivations and mechanisms of change, and morphosyntactic vagueness seems to be a plausible motivation for the case studies, even if only partially; for more detailed constructional accounts, see Kuo (2020a, 2020b, Forthcoming a). Therefore, for a diachronic cognitive linguist interested in diachronic issues broader than what constitutes grammaticalization, questions raised by morphosyntactic vagueness and vagueness-motivated directionality are of primary significance, and therefore can and probably should be dealt with independently of grammaticalization first.
6 Conclusion
In closing, some research questions are suggested. Why is it that some categories are morphosyntactically vague in some contexts in one language, but not to the same extent in another, even when such categories supposedly share crosslinguistic functional similarities? For example, although modality and conditionality share much in common, modal and protasis connective are morphosyntactically more distinct in English than Mandarin (Section 3.1.1). To put it another way, it may be proposed that morphosyntactic vagueness is an expression of iconicity, in that what is functionally similar is morphosyntactically similar (i.e. less morphosyntactically differentiated). If this is true, why is it that such iconic associations between functional and morphosyntactic properties are prominent in some languages (e.g. Mandarin), but not others (e.g. English)?[12] What diachronic developments have led to the formal distributional convergence or divergence of functionally similar categories across different languages? How do morphosyntactic categories become more or less vague? Moreover, so far, only ‘two-way’ morphosyntactic vagueness and its diachronic manifestation, bidirectionality, have been discussed. But all else being equal, ‘X-way’ vagueness could hypothetically motivate ‘X-way’ directionality, which begs the questions: What is the limit on ‘X’? What constrains morphosyntactic vagueness and vagueness-motivated directionality? How should it be accounted for?
To answer these questions, future research may need to uncover more morphosyntactically vague categories, by carefully observing under what distributional and functional conditions morphosyntactic distinctions become neutralized. Once such categories are identified, more diachronic investigation into the interaction of morphosyntactic vagueness and directionality may be carried out.
Abbreviations
3ps = third person; exp = experiential; fp = final particle; loc = locative; pl = plural; poss = possessive
Acknowledgements
I would like to thank the editor Stefan Hartmann for his invitation and insightful comments, which have greatly improved the paper. I am also grateful to Nadine Dietrich for our discussion about grammaticalization and unidirectionality. All errors remain solely my own.
References
Ahrens, Kathleen & Chu-Ren Huang. 2016. Classifiers. In Chu-Ren Huang & Dingxu Shi (eds.), A reference grammar of Chinese, 169–198. Cambridge: Cambridge University Press.10.1017/CBO9781139028462.008Search in Google Scholar
Akatsuka, Noriko & Patricia M. Clancy. 1993. Conditionality and deontic modality in Japanese and Korean. In Patricia M. Clancy (ed.), Japanese/Korean Linguistics, vol. 2, 177–192. Stanford: Center for the Study of Language and Information.Search in Google Scholar
Berg, Thomas. 2014. Boundary permeability: A parameter for linguistic typology. Linguistic Typology 18(3). 489–531.10.1515/lingty-2014-0020Search in Google Scholar
Börjars, Kersti & Nigel Vincent. 2011. Grammaticalization and directionality. In Heiko Narrog & Bernd Heine (eds.), The Oxford handbook of grammaticalization, 163–176. Oxford: Oxford University Press.10.1093/oxfordhb/9780199586783.013.0014Search in Google Scholar
Breitbarth, Anne. 2019. Should a conditional marker arise… The diachronic development of conditional sollte in German. Glossa: A Journal of General Linguistics 4(1). 21.10.5334/gjgl.599Search in Google Scholar
Brems, Lieselotte. 2011. Layering of Size and Type Noun Constructions in English. Berlin: Mouton.10.1515/9783110252927Search in Google Scholar
Byloo, Pieter & Jan Nuyts. 2014. Meaning change in the Dutch core modals: (Inter)subjectification in a grammatical paradigm. Acta Linguistica Hafniensia 46(1). 85–116.10.1080/03740463.2014.955978Search in Google Scholar
Chen, I-Hsuan. 2017. The polysemy network of Chinese ‘one’-phrases in a diachronic constructional perspective. Constructions and Frames 9(1). 70–100.10.1075/cf.9.1.03cheSearch in Google Scholar
Company, Concepción Company. 2018. Four directionalities for grammaticalization: Evidence for new diachronic paths. Journal of Historical Linguistics 8(3). 356–387.10.1075/jhl.17032.comSearch in Google Scholar
Croft, William. 2001. Radical Construction Grammar: Syntactic theory in typological perspective. Cambridge: Cambridge University Press.10.1093/acprof:oso/9780198299554.001.0001Search in Google Scholar
Cuenca, Maria Josep. 2020. Defective connective constructions: Some cases in Catalan and Spanish. Corpus Pragmatics 4. 423–448.10.1007/s41701-020-00083-zSearch in Google Scholar
Denison, David. 2010. Category change in English with and without structural change. In Elizabeth C. Traugott & Graeme Trousdale (eds.), Gradualness, gradience, and grammaticalization, 105–128. Amsterdam: Benjamins.10.1075/tsl.90.07denSearch in Google Scholar
Denison, David. 2017. Ambiguity and vagueness in historical change. In Marianne Hundt, Sandra Mollin & Simone. E. Pfenninger (eds.), The changing English language: Psycholinguistic perspectives, 292–318. Cambridge: Cambridge University Press.10.1017/9781316091746.013Search in Google Scholar
Denison, David. 2018. Why would anyone take long? Word classes and Construction Grammar in the history of long. In Kristel Van Goethem, Muriel Norde, Evie Coussé & Gudrun Vanderbauwhede (eds.), Category change from a constructional perspective, 119–148. Amsterdam: Benjamins.10.1075/cal.20.05denSearch in Google Scholar
Diessel, Holger. 2019. The Grammar network. Cambridge: Cambridge University Press.10.1017/9781108671040Search in Google Scholar
Diewald, Gabriele. 2002. A model for relevant types of contexts in grammaticalization. In Ilse Wischer & Gabriele Diewald (eds.), New reflections on grammaticalization, 103–120. Amsterdam: Benjamins.10.1075/tsl.49.09dieSearch in Google Scholar
Eifring, Halvor. 1995. Clause combination in Chinse. Leiden: Brill.10.1163/9789004482739Search in Google Scholar
Evans, Nicholas. 2007. Insubordination and its uses. In Irina Nicholaeva (ed.), Finiteness: Theoretical and empirical foundations, 366–431. Oxford: Oxford University Press.10.1093/oso/9780199213733.003.0011Search in Google Scholar
Evans, Nicholas & Honoré Watanabe (eds.). 2016. Insubordination. Amsterdam: Benjamins.10.1075/tsl.115Search in Google Scholar
Faller, Martina. 2017. Reportative evidentials and modal subordination. Lingua 186/187. 55–67.10.1016/j.lingua.2014.10.004Search in Google Scholar
Fujii, Seiko. 2004. Lexically (un)filled constructional schemes and construction types: The case of Japanese modal conditional constructions. In Mirjam Fried & Jan-Ola Östman (eds.), Construction grammar in a cross-language perspective, 121–155. Amsterdam: Benjamins.10.1075/cal.2.04fujSearch in Google Scholar
Heine, Bernd. 2002. On the role of context in grammaticalization. In Ilse Wischer & Gabriele Diewald (eds.), New reflections on grammaticalization, 83–101. Amsterdam: Benjamins.10.1075/tsl.49.08heiSearch in Google Scholar
Hong, Bo & Dong Zhengcun. 2004. “Fēi X bùkě” géshì de lìshǐ yǎnhuà hé yǔfǎ huà [The historical evolution and grammaticalization of the expression “fei X bu ke”]. Zhongguo Yuwen 3. 253–261.Search in Google Scholar
Hopper, Paul J. & Elizabeth C. Traugott. 2003. Grammaticalization, 2nd edn. Cambridge: Cambridge University Press.10.1017/CBO9781139165525Search in Google Scholar
Hsu, Hsun-Ming, Yu-Fang Wang & Kai-Ming Hu. 2015. Direct and indirect conditionals: A corpus-based study of Chinese yaoshi and yaobushi in spoken and written discourse. Taiwan Journal of Linguistics. 13(2). 31–77.Search in Google Scholar
Kratzer, Angelika. 2012. Modals and conditionals. Oxford: Oxford University Press.10.1093/acprof:oso/9780199234684.003.0004Search in Google Scholar
Kuo, Yueh Hsin. 2018. The development of three classifiers into degree modifier constructions in Chinese. In Sylvie Hancil, Tine Breban & José Vicente Lozano (eds.), New trends on grammaticalization and language change, 315–331. Amsterdam: John Benjamins.10.1075/slcs.202.13kuoSearch in Google Scholar
Kuo, Yueh Hsin. 2020a. Reinforcement by realignment in diachronic construction grammar: The case of classifier xiē in Mandarin Chinese. Constructions and Frames 12(2). 206–238.10.1075/cf.00040.kuoSearch in Google Scholar
Kuo, Yueh Hsin. 2020b. Late-stage grammatical change in Chinese: A constructional account. Edinburgh: University of Edinburgh PhD Thesis.Search in Google Scholar
Kuo, Yueh Hsin. 2021. From dynamic modal to conditional protasis connective: Evidence from Chinese néng ‘be able to’. Functions of Language (online first).10.1075/fol.20030.kuoSearch in Google Scholar
Kuo, Yueh Hsin. Forthcoming a. From deontic modality to conditionality: A diachronic investigation into bi in Classical Chinese. Journal of Historical Pragmatics 22(2).10.1075/jhp.19002.kuoSearch in Google Scholar
Kuo, Yueh Hsin. Forthcoming b. Bidirectionality between modality and conditionality in Chinese: A constructionalization account. Diachronica.Search in Google Scholar
Kuteva, Tania, Bernd Heine, Bo Hong, Haiping Long, Heiko Narrog & Seongha Rhee. 2019. World lexicon of grammaticalization, 2nd edn. Cambridge: Cambridge University Press.10.1017/9781316479704Search in Google Scholar
Li, Charles N. & Sandra A. Thompson. 1976. Subject and topic: A new typology of language. In Charles N. Li & Sandra A. Thompson (eds.), Subject and topic, 457–490. New York: Academic Press.Search in Google Scholar
Li, Charles N. & Sandra A. Thompson. 1981. Mandarin Chinese: A functional reference grammar. Berkeley: University of California Press.10.1525/9780520352858Search in Google Scholar
Mauri, Caterina & Andrea Sansó. 2011. How directive constructions emerge: grammaticalization, constructionalization, cooptation. Journal of Pragmatics 43(14). 3489–3521.10.1016/j.pragma.2011.08.001Search in Google Scholar
Narrog, Heiko. 2012. Modality, subjectivity, and semantic change: A cross-linguistic perspective. Oxford: Oxford University Press.10.1093/acprof:oso/9780199694372.001.0001Search in Google Scholar
Neels, Jakob & Stefan Hartmann. 2018. Reduction or expansion? A bit of both: A case study on the development of German degree modifiers. In Evie Coussé, Peter Andersson & Joel Olofsson (eds.), Grammaticalization meets Construction Grammar, 137–168. Amsterdam: Benjamins.10.1075/cal.21.c6Search in Google Scholar
Norde, Muriel. 2009. Degrammaticalization. Oxford: Oxford University Press.10.1093/acprof:oso/9780199207923.001.0001Search in Google Scholar
Peng, Lizhen. 2007. Xiandai Hanyu Qingtai Yanjiu [A study of Modern Chinese modals]. Beijing: China Social Sciences Press.Search in Google Scholar
Roberts, Craige. 1989. Modal subordination and pronominal anaphora in discourse. Linguistics and Philosophy 12(6). 683–721.10.1007/BF00632602Search in Google Scholar
Searle, John R. 1979. Expression and meaning: Studies in the theory of Speech Act. Cambridge: Cambridge University Press.10.1017/CBO9780511609213Search in Google Scholar
Stefanowitsch, Anatol. 2003. A construction-based approach to indirect speech acts. In Klaus-Uwe Panther & Linda L. Thornburg (eds.), Metonymy and pragmatic inferencing, 105–126. Amsterdam: Benjamins.10.1075/pbns.113.09steSearch in Google Scholar
Stefanowitsch, Anatol & Stefan Th. Gries. 2003. Collostructions: Investigating the interaction of words and constructions. International Journal of Corpus Linguistics 8(2). 209–243.10.1075/ijcl.8.2.03steSearch in Google Scholar
Sweetser, Eve. 1990. From etymology to pragmatics: Metaphorical and cultural aspects of semantic structure. Cambridge: Cambridge University Press.10.1017/CBO9780511620904Search in Google Scholar
Traugott, Elizabeth C. 1985. Conditional markers. In John Haiman (ed.), Iconicity in syntax, 289–307. Amsterdam: Benjamins.10.1075/tsl.6.14cloSearch in Google Scholar
Traugott, Elizabeth C. 2008. Grammaticalization, constructions and the incremental development of language: Suggestions from the development of degree modifiers in English. In Regine Eckardt, Gerhard Jäger & Tonjes Veenstra (eds.), Variation, selection, development, 219–250. Berlin: Mouton.10.1515/9783110205398.3.219Search in Google Scholar
Traugott, Elizabeth C. & Graeme Trousdale (eds.). 2010. Gradience, gradualness and grammaticalization. Amsterdam: Benjamins.10.1075/tsl.90Search in Google Scholar
Traugott, Elizabeth C. & Graeme Trousdale. 2013. Constructionalization and constructional changes. Oxford: Oxford University Press.10.1093/acprof:oso/9780199679898.001.0001Search in Google Scholar
Traugott, Elizabeth C. & Richard Dasher. 2002. Regularity in semantic change. Cambridge: Cambridge University Press.10.1017/CBO9780511486500Search in Google Scholar
Trousdale, Graeme. 2012. Grammaticalization, constructions and the grammaticalization of constructions. In Kristin Davidse, Tine Breban, Lieselotte Brems & Tanja Mortelmans (eds.), Grammaticalization and language change: New reflections, 167–198. Amsterdam: Benjamins.10.1075/slcs.130.07troSearch in Google Scholar
Tsao, Feng-fu. 1996. Hànyǔ de tíshēng dòngcí [Raising verbs in Chinese]. Zhongguo Yuwen 3. 172–182.Search in Google Scholar
Tuggy, David. 1993. Ambiguity, polysemy, and vagueness. Cognitive Linguistics 4(3). 273–290.10.1515/cogl.1993.4.3.273Search in Google Scholar
van der Auwera, Johan & Vladimir A. Plungian. 1998. Modality’s semantic map. Linguistic Typology 2(1). 79–124.10.1515/lity.1998.2.1.79Search in Google Scholar
von Mengden, Ferdinand & Horst J. Simon. 2014. What is it then, this Grammaticalization? Folia Linguistica 48(2). 347–360.10.1515/flin.2014.012Search in Google Scholar
Verveckken, Katrien. 2012. Towards a constructional account of high and low frequency binominal quantifiers in Spanish. Cognitive Linguistics 23(2). 421–478.10.1515/cog-2012-0013Search in Google Scholar
Wang, Canlong. 2008. “Fēi VP bùkě” jù shì zhōng “bùkě” de yǐnxiàn: Jiān tán “fēi” de xū huà [On the occurrence and non-occurrence of buke (不可) and the grammaticalization of fei (非) in the construction “fei + VP + buke”]. Zhongguo Yuwen 2. 109–119.Search in Google Scholar
Wang, Yu-Fang, Jyun-gwan Chen, David Teanor & Hsun-Ming Hsu. 2014. Exclusivity, contingency, exceptionality and (un)desirability: A corpus-based study of Chinese chufei (‘unless’) in spoken and written discourse. Language & Communication 37. 40–59.10.1016/j.langcom.2014.04.002Search in Google Scholar
Yu, Hsueh-ying. 1998. Some speculations on the semantic change of Chinese modal verb ‘Yao’. Wenshan Review 1(2). 161–175.Search in Google Scholar
Zhang, Nina Ning. 2013. Classifier structures in Mandarin Chinese. Berlin: Mouton.10.1515/9783110304992Search in Google Scholar
Ziegeler, Debra. 2004. Redefining unidirectionality: Is there life after modality? In Olga Fischer, Muriel Norde & Harry Perridon (eds.), Up and down the cline: The nature of grammaticalization, 115–135. Amsterdam: Benjamins.10.1075/tsl.59.07zieSearch in Google Scholar
©2021 Walter de Gruyter GmbH, Berlin/Boston
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
Articles in the same Issue
- Frontmatter
- Diachronic Cognitive Linguistics
- Variation and change in idiolects: Tracing a lexical substitution inside a psychiatric hospital around 1900
- The syntax of metaphor
- Of travels and travails: The role of semantic typology, argument structure constructions, and language contact in semantic change
- Morphosyntactic vagueness and directionality
- What does solid spelling reveal about cognition? Evidence from Middle Low German
- The emergence of light verb constructions
- The diachronic development of the verbal bracket construction in German
- The problem of universalism in (diachronic) cognitive linguistics
Articles in the same Issue
- Frontmatter
- Diachronic Cognitive Linguistics
- Variation and change in idiolects: Tracing a lexical substitution inside a psychiatric hospital around 1900
- The syntax of metaphor
- Of travels and travails: The role of semantic typology, argument structure constructions, and language contact in semantic change
- Morphosyntactic vagueness and directionality
- What does solid spelling reveal about cognition? Evidence from Middle Low German
- The emergence of light verb constructions
- The diachronic development of the verbal bracket construction in German
- The problem of universalism in (diachronic) cognitive linguistics
