Introduction
The Yellow River plateau in the Qinghai-Gansu border region in China (Southeastern Qinghai and Southern Gansu provinces) is a hotbed of cultural and linguistic contact. In the districts of Linxia in Gansu province, and of Tongren and Huangnan in Qinghai province, there are several languages descended from at least four different language taxa: Sinitic, Tibeto-Burman (Amdo Tibetan), Turkic, Mongolic.
Nine ethnic groups are located in the region: Chinese (Hàn 汉); Huí 回 (Muslims); Dōngxiāng 东乡 (Santa); Bǎoān 保安 (Baonan); Tǔzú 土族 (Monguor); Dōngbù Yùgù 东部裕固 (Eastern Yugur or Yellow Uighur); Xībù Yùgù 西部裕固 (Western Yugur); Sālā 撒拉 (Salar); Ānduō Zàngzú 安多藏族 (Amdo Tibetan).
They speak a dozen different languages: (i) Sinitic languages, more specifically Northwestern dialects such as Línxià 临夏 (or Hézhōu 河州), Tāngwāng 唐汪 (in the Gansu Dongxiang zone), Gāngōu 甘沟 (in the Minhe county 民和县 of the Qinghai province), Wǔtún 五屯 (in the Tongren district 同仁县 of the Qinghai province); (ii) Mongolic languages such as Santa, Baonan, Monguor, Eastern Yugur; not to mention (iii) Turkic languages such as Salar, Western Yugur; and (iv) Tibeto-Burman languages, namely, Amdo Tibetan.
These languages share many syntactic structural properties and it has been suggested that they could form a Sprachbund, or linguistic area, and also that some of these languages might be viewed as ‘mixed’.
This paper will examine one of the most characteristic syntactic properties, namely the case system, for the three following Sinitic languages: Línxià (or Hézhōu), Tāngwāng, and Gāngōu. To determine the number of cases in a given language is not an easy task. 1
The Case System in Línxià 临夏 (or Hézhōu 河州)
The Línxià (or Hézhōu) language is a verb-final language (SOV type) with supposedly head-final constructions. The lexicon is clearly Chinese, with three tones. 2 Strikingly, one of the most important characteristics of the Línxià Sinitic language is the fact that it seems to have a case system, unknown in Standard Mandarin, as several suffixes are found used on nouns to express this category. These suffixes, which are atonal, are -ha 哈, -xet’e, -tala, -la 啦 and -liangge 两个.
The Accusative, Dative and Beneficiary Suffix -ha 哈
The suffix -ha 哈 is attached to the noun in pre-verbal position and can express the accusative (bīngé 宾格, direct object function), the dative (与格 yǔgé), as well as some beneficiary cases (shòuyìgé 受益格, i.e. indirect objects). For example:
(1) 我这个人哈认不的
wŏ zhèige rén-ha rènbude (where -ha is an accusative suffix)
1sg this.cl person-ha not.know
‘I don’t know this person.’ 3
This accusative suffix can also be found after pronouns, as in:
(2) 你什么哈想著哩?
nǐ shénme-ha xiǎng zhuoli
2sg what-ha think prog
‘What are you thinking?’
In the three following examples, there are two -ha suffixes in the same sentence. One is accusative, and the other is dative:
(3) 我他们哈这个事情哈沒说
wǒ tāmen-ha zhèige shìqing-ha méi shuō
1sg 3pl-ha this.cl matter-ha neg tell
‘I did not tell them this matter.’
(4) 尕娃无常了, 无常的哈我他的阿妈哈沒说
gǎwā wúcháng le wúchángde-ha wǒ tāde āmā-ha méi shuō
child die perf dying.part-ha 1sg 3sg mother-ha neg tell
‘The child is dead, I have not told this dying (event) to his mother.’
(5) 他你哈信哈沒來吗?
tā nǐ-ha xìn-ha méi lái ma ?
3sg 2sg-ha letter-ha neg come part
‘He did not send you a letter?’
When there is a single suffix -ha in double-object constructions, it usually marks the dative and is attached to the indirect object, as in:
(6) 我他哈电影票给的了
wǒ tā-ha diànyǐng piào gěi de le
1sg 3sg-ha movie ticket give part perf
‘I gave him the movie ticket.’
The Ablative Suffixes -ta and -xet’e
These two suffixes, for which no Chinese characters could be provided, both express the ablative case (cónggé 从格) in Línxià. Examples:
(7) 北京-ta 回來了
Běijīng-ta huílái le
Peking-ta return part
‘(He/She) is back from Peking.’
(8) 他-xet’e 借下的
tā-xet’e jièxià de
3sg-xet’e borrow part
‘(He/She) borrowed (it) from him.’
There are, however, two constraints for the use of the suffix -xet’e (which could have been borrowed in Línxià from the Western Yugur suffix -tan) as an ablative case. First of all, the NP to be suffixed needs to be a personal name or a personal pronoun. Moreover, the VP used in the construction should be a verb of ‘giving’ (including the ‘say’ verbs) or ‘receiving’, such as ‘to give’, ‘to buy’, ‘to transmit’, ‘to lend’, ‘to ask’, ‘to borrow’, etc.; i.e. the verbs called by PeyraubeReference Peyraube 4 the ‘lexical dative verbs’.
As for -ta, which is also used sometimes as a comparative suffix, it might have been borrowed from the Bao’an language -da (see Luo, Ma S in Refs 2 and 5). It could also be a direct borrowing of the locative/dative case -da, attested in West Middle Mongolian. 6
The Locative-allative suffix -tala
This suffix -tala is a kind of allative case (external locative or directional: zhǐgé 止格) expressing time, and meaning ‘until’. Example:
(9) 他晌午-tala 睡了
tā shǎngwu-tala shuì le
3sg midday-tala sleep perf
‘He slept until midday.’
The same marker, called a ‘terminative serial marker’ by both Janhunen et al. and SandmanReference Janhunen, Peltomaa, Sandman and Dongzhou 7 is also found in Wǔtún (where it also indicates the relationship between serialized verbs in describing complex events) as well as in Baonan (-tele). Moreover, it is well documented in Mongolic languages, from where it likely has its source. Example from Baonan:
(10) Ta xoro al-tele saaghe
2pl evening become-tele wait
‘You wait until (it) becomes evening.’ (Taken from Ref. 8, p. 339.)
The Instrumental, Manner and Comitative Suffixes -la 啦 and -liangge 两个
These two suffixes, -la 啦 and -liangge 两个, both express the instrumental-manner case (fāngshìgé-gōngjùgé- 方式格—工具格) and the comitative (bànsuígé 伴随格) case. The suffix -la is also used in Tāngwāng. Its origin is unclear, although it is also found with approximately the same functions in Salar, a Turkic language.Reference Shujun 5 Examples:
(11) 你普通话啦说!
nǐ pǔtōnghuà-la shuō
2sg common.language-la speak
‘Speak the common language!’ (-la is an instrumental/manner case)
(12) 你他们哈什么啦款待哩?
nǐ tāmen-ha shénme-la kuǎndài li
2sg 3pl-ha what-la entertain part
‘With what are you going to entertain them?’ (-la is an instrumental/manner case)
(13) 眼睛啦看, 嘴啦不说
yǎnjing-la kàn, zuǐ-la bù shuō
eye-la look mouth-la neg speak
‘Look with (your) eyes, don’t speak with (your) mouth.’ (-la is an instrumental case)
The suffix -liangge probably comes from the Standard Mandarin ‘Numeral + Classifier’ phrase: liǎng 兩 ‘two’ + ge 个 ‘cl’. The reanalysis process of ‘Numeral + Classifier’ phrase into a case suffix has also occurred in Baonan and in Santa. It could have later been borrowed into Línxià from one or other of these two Mongolic languages (Dwyer in Ref. 2). Examples:
(14) 我笔两个写去
wǒ bǐ-liangge xiě qù
1sg pen-liangge write dir
‘I am writing with a pen.’ (-liangge is an instrumental case)
(15) 我他两个不去
wǒ tā-liangge bù qù
1sg 3sg-liangge neg go
‘I don’t go with him.’ (-liangge is a comitative case)
Conclusion
All these case suffixes are obviously not Chinese, and it is reasonable to hypothesize that they have been borrowed from Altaic languages (I assume that both Mongolic and Turkic language sub-families belong to the Altaic family) spoken in the region and with which the Línxià Sinitic language has been in close contact for a long period of time, at least since the tenth century AD): Baonan, Santa, Monguor, Salar. It is, however, difficult to identify which Altaic language is the real source (the model language) of the transfer to the replica language (Línxià). 9 It is unlikely that Amdo Tibetan has exerted a stronger influence on Línxià than the Mongolic languages, as is suggested by Xie et al. in Ref. 2.
The Case System in Tāngwāng 唐汪
The Tāngwāng language is another Sinitic verb-final language (SOV type) spoken in a small town and some villages in the Dongxiang Autonomous County of Gansu province. It has obviously been influenced by the Mongolic Santa (or Dōngxiāng) language. 10 The lexicon is nevertheless clearly Chinese, and most of the syntactic structures are those found in the Sinitic languages and dialects of Northwestern China. 11
Tāngwāng also uses several suffixes on nouns to express case. Four cases are identified by Ibrahim, but seven by Xu Dan and Peyraube: 11 accusative, dative/beneficiary, ablative, locative, instrumental and comitative, comparative and locative-allative. The nominative is unmarked. Finally, the pronoun system clearly shows case distinctions as well.
The Accusative, Dative and Beneficiary Case -ha 哈
Examples:
(16) 你他哈打寮
Nǐ tā-ha dǎ-liao (-ha is an accusative suffix) 12
2sg 3sg-ha beat-per
‘You have beaten him.’
(17) 你啊个哈那本书哈给给寮?
nǐ āge-ha nèiběn shū-ha gěi-gei liao? (-ha is a dative suffix)
2sg who-ha that.cl book-ha give-to perf
‘To whom did you give that book?’
(18) 我你哈花哈给一个
wǒ nǐ-ha huā-ha gěi yīge
1sg 2sg-ha flower-ha give one.cl
‘I give you one flower.’
In this last example, the first -ha is a dative suffix, and the second one is an accusative suffix.
The Ablative Case -xie 些
(19) 北京些來下底人多得很
Běijīng-xie lái-xia-di rén duō-de-hěn
Peking-xie come-dir-part people many-part-very
‘People coming from Peking are many.’
(20) 我将学校里些來哩
wǒ jiāng xuéxiào-li-xie lái li
1sg just school-in-xie come part
‘I have just come (back) from school.’
This ablative suffix -xie has a very different form from the ablative suffixes used in Línxià (-ta or -xet’e). Its origin is unclear. Moreover, it is the same suffix -xie that is also used to express the comparative case, as seen below.
The Use of -xie 些 to Code the Comparative Case
(21) 这个那个(哈)些尕些
zhèige nèige(-ha)-xie gǎ xiē
this.cl that.cl(-ha)-xie small little
‘This (one) is slightly smaller than that (one).’
In the example above -xie is a comparative suffix equivalent to the comparative morpheme bǐ 比 in Standard Mandarin, and -ha an accusative suffix which is optional.
(22) 我你(哈)些胖些
wǒ nǐ(-ha)-xie pang xiē
1sg 2sg(-ha)-xie big little
‘I am a little bigger than you.’
The Instrumental/Comitative Case -la 啦
As in Línxià, a single suffix expresses both the instrumental and the comitative cases: -la 啦. Examples:
(23) 我钢笔啦写下底
wǒ gāngbǐ-la xiě-xia di (-la is instrumental)
1sg pen-la write-dir-part
‘I have written (it) with a pen.’
(24) 我刀子啦肉哈切咧
wǒ dāozi-la ròu-ha qiē lie (-la is instrumental, -ha is accusative)
1sg knife-la meat-ha cut part
‘I cut the meat with a knife.’
(25) Zubeta Muhamede 啦是兄妹是
Zubeta Muhamede-la shì xiōngmèi shì (-la is comitative)
Zubeta Mohamed-la be brother:sister be
‘Zubeta together with Mohamed are brother (and) sister.’
(26) 我哈三啦缸哈达坂些尕驴啦唐汪里生产队哈拉著來给寮
wǒ Hāsān-la gāng-ha Dábǎn-xie gǎlǘ-la Tángwāng-li
1sg Hasan-la jar-ha Daban-xie donkey-la Tangwang-in
shēngchǎnduì-ha lā-zhuo lái-gei-liao
production.unit-ha carry-prog come-caus-perf
‘Together with Hasan, I carried the jar with a small donkey from Daban to come to the unit of production of Tangwang.’ (See Djamouri in Ref. 11.)
In this sentence, the first suffix -la codes a comitative, the second -ha an accusative, the third -xie is an ablative, the fourth -la codes the instrumental, the fifth -li a locative-allative, while the sixth -ha marks the beneficiary.
The Allative Case
Xu DanReference Dan 13 has finally identified a seventh case, locative-allative (or terminative), which has the same form of the Línxià allative case: -tala.
(27) 你早上些这会儿-tala 做麻着?
nǐ zǎoshàng-xie zhèhuìr-thala zuò má zhe
2sg morning-xie this.moment-thala do what dur
‘What have you been doing from the morning until now?’
The Case System in Gāngōu 甘沟
The Gāngōu language is also a Sinitic verb-final language (SOV type) spoken in the Minhe Hui and Tu Autonomous County in Qinghai province.Reference Lide and Stuart 14 It has been heavily influenced by the Mongolic language of Monguor (Tǔ). In spite of having a good number of words borrowed from Monguor, the lexicon is basically Chinese and most of the syntactic structures are those of the Sinitic languages and dialects of Northwestern China. For more general characteristics of Gāngōu, see Feng and StuartReference Lide and Stuart 14 , Zhu et al., and SlaterReference Yongzhong, Chuluu and Stuart 15 for a study of contact situations between Gāngōu and Monguor.
As for Línxià and Tāngwāng, Gāngōu has several suffixes used on nouns to express cases. The five cases discussed below are: accusative, dative/beneficiary, ablative, instrumental and comitative, as well as comparative. The nominative is unmarked.
The Accusative, Dative and Beneficiary Case -ha 哈
In Gangou, the case suffix -ha is by far the most common. It can express the accusative, dative and beneficiary cases. Examples:
(28) 这个倆儿哈啊里-ma 远远之送掉
zhige liar-ha ali-ma yuanyuan-zhi song-diao (-ha is accusative)
this.cl two-ha where-ever far-gen send-res
‘Let us send these two far, far away.’Reference Yongzhong, Chuluu and Stuart 15 , 16
(29) 阿达这个花儿哈我倆儿哈折给
ada zhige huar-ha wo liar-ha zhe-gei
father this.cl flower-ha 1pl two-ha pick-caus
‘Father, pick this flower for us two.’
The first -ha in (29) is accusative, the second one beneficiary.
(30) 我你们哈还大的花儿兒折给去
wo nimu-ha han da-di huar zhe-gei qi (-ha is beneficiary)
1sg 2pl-ha more big-nom flower pick-caus go
‘I’ll go and pick up a bigger flower for you.’
The Ablative Case -sha 沙
It probably comes from the ablative suffix -sa in Monguor:
(31) 天窗沙个看时
tianchuang-sha ge kan-shi
skylight-sha one look-when/if
‘When looking from the skylight…’
The Comparative Case -sha 沙
The same suffix used for the ablative case, -sha, is also used for expressing the comparative, as in:
(32) 这个花儿好沒这个哈沙还大之折给給
Zhige huar hao mei zhige-ha-sha han da-zhi zhe-gei
this.cl flower good neg this.cl-ha-sha more big-part pick-caus
‘This flower is not good, (I will) pick a bigger one than this.’
In this last example, -ha is the accusative and -sha the comparative suffix. Ablative case markers and adpositions have been widely identified as sources for comparative markers in the relevant cross-linguistic literature.Reference Stassen 17
The Instrumental/Comitative Case -lia(r) 倆(儿)
Like other Sinitic languages of the region, the suffix expressing the instrumental case is the same as the one expressing the comitative case. Examples:
(33) 医院里狗倆儿试验执哩
yiyuan-li gou-liar shiyan zhi-li
hospital-in dog-liar perform:experiment prog
‘An experiment is being performed with dogs in the hospital.’
Comparative Analysis of the Four Languages
Comparing like with like, which is a basic task of the typologist, we can now summarize the situation in the three investigated languages as follows, while conflating the accusative and the dative/beneficiary markers into one single category since they are expressed by one and the same marker -ha 哈.
These cases can be ranked according to the following hierarchy: a language that does not have a given case at a certain point on this hierarchy will tend to not have any of the cases to the right of the missing case:
Nominative (unmarked) → Accusative → Dative/Beneficiary → Locative/Ablative/allative → Instrumental/Comitative → Others.
This conforms to the general tendency that can be observed in the hierarchy of cases in other case languages, i.e.:
Nominative → Accusative or Ergative → Genitive → Dative → Locative → Ablative → Instrumental → Vocative → Others.
The fact that one and the same suffix -ha expresses the accusative as well as the dative/beneficiary in all four languages under investigation is rather striking, although not at all unusual. The case values identified (such as nominative, or accusative, or dative) may have different functions, recognized by the names used in traditional grammar such as: genitivus possessivus, genitivus partitivus, dativus possessivus, etc. It is a standard example of syncretism of case to reduce the number of suffixes used to express the values of this category. The same can be said for the suffixes -ta, -xie and -sha which express both the ablative and the comparative, respectively in Línxià, Tāngwāng and Gāngōu. It is likely that the ablative is the source morpheme, and the comparative, the target: ‘Ablative>Comparative’, as attested in Heine and KutevaReference Heine and Kuteva 18 for several typologically different languages, and in Stassen.Reference Stassen 17
As for the syncretism of the instrumental, manner and comitative cases, it is also quite well-known in general linguistics. One can assume that the comitative is the source. See again Heine and Kuteva (Ref. 18, pp. 82–87) who list several cases of ‘Comitative>Instrumental’, as in Ewe, Baka, Ngbaka Ma’Bo, Turkish, Latin, Albanian, Hungarian, Bulgarian, Imbabura Quechua, Yagua, and also of cases of ‘Comitative>Manner’ in German, Hausa, Ngbaka Ma’Bo, Hungarian, Tamil. See also Liu DanqingReference Danqing 19 and Wu Fuxiang. 9 Syncretism of the Comitative with Instrumental is however rare in Sinitic languages where the tendency for comitative adpositions is either to develop along a pathway to a comparative marker of equality, into a NP coordinative conjunction, or into a differential object marker via several oblique functions.Reference Chappell, Peyraube and Yunji 20 Similarly, the comparative marker of inequality is rarely based on an ablative adposition in Sinitic.Reference Chappell and Peyraube 21
Discussion
Case Suffixes or Postpositions?
As Corbett 1 pointed out, ‘the notion of “case” is one that is often taken as a given by the linguist’, and I have taken it for granted that the markers discussed above are bona fide suffixes rather than, say, postpositions or clitics, and that, being suffixes, they are case endings. This could be justified by certain properties, and by accepting a non-rigorous definition of case. In fact, cases suffixes are largely grammaticalized postpositions, and so it should not be surprising to find that cases and postpositions share a number of important characteristics, the development of case being in accordance with the principles of grammaticalization identified by Heine,Reference Heine 22 i.e. ‘desemanticization’, ‘decategorization’, ‘erosion’. This also explains why the use of prepositions is very limited in these languages.
ComrieReference Comrie 23 has also clearly demonstrated that the notion of case is far from straightforward and that even the traditional understanding actually encompasses a number of distinct, though interrelated, phenomena. He argued that we need to distinguish a morphological notion of case, the ‘case forms’ of a lexeme, from a functional notion, the ‘case’ that an NP is ‘in’ in a given construction. He pointed out (Ref. 23, p. 89) that there are important differences between agglutinating languages such as Turkish − where the ‘cases’ are immediately identified by their suffixes − and inflectional languages, where there is a clear need for a morphological case feature.
SpencerReference Spencer 24 also argues for a distinction between formal or morphological case and syntactic case, as, in most modern approaches to grammar, the term ‘case’ is systematically ambiguous between ‘(inflected) form of a nominal word’ and ‘property of a noun phrase’ (Ref. 24, p. 185). He convincingly shows that case markers in Hungarian (essive, allative, ablative, instrumental, translative, terminative) are best thought of as ‘fused postpositions’ and thus that the Hungarian system does not represent a canonical case system, such as the ones found in inflectional languages like Latin or Russian, i.e. a system ‘which must appeal to a morphosyntactic property, feature or attribute [Case] in order to generalize across systematically distinct forms.’
It is clear that in the situation which is of interest to us, namely, the Northwestern Sinitic languages of China, we are not dealing with real case suffixes of the same type that are found in both Ancient Greek or Latin inflectional languages whose grammars are considered as uncontroversially needing to appeal to a feature of case. Can one nevertheless refer to case in Northwestern Sinitic? In my view, yes, if one adopts a broad definition of this phenomenon which privileges the functional notion to the detriment of the morphological one (see above the distinction made by ComrieReference Comrie 23 ). For example, the following definition may be adopted: ‘Case is a grammatical category associated with nominal phrases, for which it codes the syntactic function in the sentence or clause.’ Thus, ‘cases denote categories of meaning, not categories of forms’. And if cases are generally expressed by nominal affixes which vary with the nouns (and naturally it is just for such affixes, which together form the entire system of nominal inflection or the noun declensions, that the term ‘case’ is reserved), they may also be expressed by adpositions such as de or à in French. In other words, there is no reason to adhere to a strict definition of case.
Hence, we can accept that there is quite definitely a system of case in these Northwestern Sinitic languages in China. But are these cases expressed by suffixes or adpositions? It can certainly be considered that we are dealing with suffixes and not postpositions, on the understanding that these are derivational suffixes and not inflectional suffixes, We can even go further and claim that these Northwestern Sinitic languages are good examples of ‘case-increasing languages’ to take up the classification of Kulikov,Reference Kulikov 25 who distinguishes the following three main evolutionary types of languages: (i) case increasing languages, that is languages which undergo an augmentation in the number of cases, such as Uralic, New Indo-Aryan, Tocharian; (ii) case-reducing languages such as Germanic, Italic, Romance and Celtic; (iii) case-stable languages such as Armenian, Slavic, Baltic (Lithuanian), Turkic. See also Cao Guangshun and Yu Hsiao-jungReference Guangshun and Hsiao-jung 26 who already identified two accusative markers (gēndǐ 跟底and shàng/shàngtou 上/上头) in the vernacular Chinese of the thirteenth century).
Do we have a Qinghai-Gansu Linguistic Area?
Once again, it depends on which definition of linguistic area is adopted. Most areal typologists would agree that a linguistic area is the result of a diffusion of traits, normally understood as not being inherited, that occur in unrelated languages. 27 EmeneauReference Emmeneau 28 defines a linguistic area as ‘an area which includes languages belonging to more than one family but showing traits in common which are found not to belong to the other members of (at least) one of the families’. For Thomason,Reference Thomason 29 it is ‘a geographical region containing a group of three or more languages that share some structural features as a result of contact rather than as a result of accident or inheritance from a common ancestor.’ But the best definitions are probably the following ones by Aikhenvald and Dixon:Reference Aikhenvald and Dixon 30
A linguistic area (or Sprachbund) is generally taken to be a geographically delimited area including languages from two or more language families, sharing significant traits (which are not found in languages from these families spoken outside the area). (Ref. 30, p. 11)
and by Enfield:Reference Enfield 31
A linguistic area is defined as a geographical region in which neighboring languages belonging to different language families show a significant set of structural properties in common, where the commonality in structure is due to contact and where the shared structural properties are not found in languages immediately outside the area (ideally where these include languages belonging to the same families as those spoken inside the area). (Ref. 31, p. 190.)
In other words, to have a real ‘linguistic area’, we thus need (i) a geographical area, as the areal approach obviously privileges considerations of a geographical nature; (ii) a set of languages belonging to different families in this area; (iii) the situation that these various languages share some linguistic features; and (iv) the fact that these linguistic features are unknown to the other languages of the same family.
Such a restricted definition does not, however, up to now, allow the identification of many linguistic areas. A less restricted definition, such as the one given, for instance, by Thomason,Reference Thomason 29 has led some linguists to discuss the possibility of identifying large linguistic areas for East Asian and Southeast Asian languages, comprising Sino-Tibetan, Austronesian, Austroasiatic, Tai-Kadai and Hmong-Mien languages.Reference Matisoff 32 Chappell,Reference Chappell 33 in the context of areal linguistics of the Sinitic languages of China, has examined the feasibility of a classification into six micro-areas as a refinement of the traditional North–South division originally proposed.
Such an approach undoubtedly allows us to accept also the existence of a Qinghai-Gansu linguistic area, as several linguists have already claimed. Moreover, Chinese linguists use the term ‘ha-la-xie yuyan’ (ha-la-xie languages) quite freely with regard to the languages examined above.
The fact remains that it is often difficult to distinguish, in terms of resemblances between languages, those that are the result of borrowing through contact from those that could be explained by the same source and parallel internal developments. In the absence of any historical data, as is often the case, common origin and diffusion are two hypotheses that can be envisaged.
While we have seen above that the case markers in Línxià, Tāngwāng and Gāngōu have probably been borrowed from non-Sinitic languages (predominantly Altaic languages), it could also be the case that some of them are native to Sinitic, notably the accusative/dative/beneficiary marker -ha.
Mixed Languages?
The last question that needs to be answered is this: can these Sinitic languages be characterized as being mixed languages, as has been assumed? The answer is No.
Although the concept of ‘mixed language’ is frequently used in historical linguistics over the last several decades, attested cases of language mixture are very rare. Again there is disagreement among linguists about the precise definition of a mixed language, but the basic criteria are: (i) a contact situation between two languages; (ii) a situation of bilingualism; (iii) a clear differentiation between a mixed language and languages that are mixed (every language is mixed to some extent).
A mixed language should be a language that arises through the fusion of usually two source languages, normally in situations of complete and total bilingualism, so that it is not possible to classify the resulting language as belonging to either of the language families that were its sources. As for the three languages that have been investigated in this paper, Línxià, Tāngwāng and Gāngōu, they all clearly belong to the Sinitic language family, and if they indeed combine the lexical stock of one language (Chinese) with some grammatical structures of another language (mainly from the Altaic family), these grammatical elements form neither the basic nor the entire grammatical system of the borrowed language.
We are simply dealing with classic cases of languages having a substratum and a superstratum, such as Irish English (with a substratum of Irish-Gaelic and a superstratum of Early Modern English), or Romanian (substratum: Dacian, superstratum: Vulgar Latin, and then later Slavic languages, Hungarian and German), or Paraguayan Spanish (substratum: Guaraní, superstratum: Spanish of the fifteenth century), or even Singaporean Mandarin (substratum: Southern Sinitic languages such as Southern Min, Teochew, Cantonese, Hainanese, superstratum: Standard Mandarin), languages for which no one would claim that they are mixed languages.
Conclusion
Having examined the nominal markers in the three following Sinitic languages: Hézhōu (or Línxià), Tāngwāng, Gāngōu, which have been sometimes viewed as mixed languages, I conclude that there are good reasons to consider that these languages do possess a case system and that the case markers are suffixes and no longer postpositions. It has also been shown that most of these case suffixes have been borrowed from Altaic languages (mainly Mongolic languages such as Bao’nan, Santa, Monguor, but also Turkic languages such as Salar or Western Yugur).
It has finally also been concluded that these languages, sharing several syntactic structural properties, form a kind of Sprachbund or linguistic area, the Qinghai-Gansu linguistic area, even if all the strict criteria defining what should be a linguistic area are not met. Finally, these four languages cannot be characterized as being mixed languages. They are most clearly Sinitic languages.
Abbreviations
1SG=1st personal pronoun singular
1PL=1st personal pronoun plural
2SG=2nd personal pronoun singular
2PL=2nd personal pronoun plural
3SG=3rd personal pronoun singular
3PL=3rd personal pronoun plural
CAUS=causative
CL=classifier
DIR=directional
DUR=durative
GEN=genitive
NEG=negation
PART=particle
PERF=perfective
PROG=progressive
RES=resultative
Acknowledgement
This paper is an abbreviated and modified version of Peyraube (2015).Reference Peyraube 34 The research has received funding from the Agence nationale de la recherche (Project Tysomin – Anr-11-ish2-001-01).
Alain Peyraube is currently Emeritus director of research at the Centre National de la Recherche Scientifique (CNRS, Paris, France) and Chair professor of Chinese Linguistics at the Ecole des Hautes Etudes en Sciences Sociales (EHESS, Paris). As a specialist in Chinese historical syntax and linguistic typology of Sinitic languages, Alain has authored five books (on Chinese locative constructions, on the historical evolution of the dative structures in Chinese, etc.) and around 200 articles on Chinese linguistics. His latest research has been carried out within a broadly functional and cognitive framework, using a cross-linguistic perspective and includes projects on grammatical change in Early Southern Min and language contact in Northwest China.
