Contact-induced Grammaticalization

Contact-induced grammaticalization may be a possible mechanism by which the TSM pro-V and PR-V YONG got transferred into TM. It is defined in Heine and Kuteva (2003) as following:

Contact-induced grammaticalization is a grammaticalization process that is due to the influence of one language on another. There are two main types of contact-induced grammaticalization depending on whether or not there exists already a model for the process in the model language to be replicated. If no such model exists we will refer to the process as ordinary grammaticalization; if there is a model that is transferred to the replica language we will refer to it as replica grammaticalization.

According to Heine and Kuteva (2003), in the process of contact-induced grammaticalization, speakers of the target language develop a grammatical category which is equivalent to that of the source language by using material available in their own target language. Applying this model to the issue concerned in the current study, the potential situation may be that the speakers of TM develop the categories which are equivalent to the pro-V YONG and PR-V YONG in TSM by using the existing function of Instrument main verb YONG³ (I-V YONG) in TM. If the pro-V and PR-V YONG in TM are derived from I-V YONG in TM, a crucial question should be answer first: Is this derivation a process of grammaticalization?

To examine whether this derivation is a process of grammaticalization or not, two criteria proposed by Brinton and Traugott (2005) could be profitably utilized:

decategorization, and metaphorization and metonymization. First, Hopper (1991) defined decategorization as a process by which forms “lose or neutralize the morphological markers and syntactic privileges characteristic of the full categories Noun and Verb, and…assume

3 The term “Instrument main verb YONG” here refers to the prototypical function of YONG in the following

attributes characteristic of secondary categories such as Adjective, Participle, Preposition, etc.”. Following this notion, Brinton and Traugott (2005) further claim that

“Decategorization is a defining characteristic of grammaticalization since it [is] the mechanism by which lexical items become functional”. Based on this criterion, the degree of grammaticality of the three functions of YONG, I-V, pro-V and PR-V, can be identified by examining whether they can be construed with Mandarin verbal aspectual markers.

Typically, Mandarin verbs can be affixed with the four aspectual markers 了 le (perfective marker), 過 guo (experiential perfective marker), 在 zai (progressive marker), and 著 zhe (stative imperfective marker) (cf. Smith, 1991). As a main verb, I-V YONG can be construed with all the four aspectual markers, as shown in the following examples:

(25) a. 今天，我用了很多錢。 [Google, 2012/7/20]

jintian wo yongle henduo qian today I use-PF very-much money

‘I use a lot of money today.’

b. 我用過不少面膜， [Google, 2012/7/20]

wo yongguo bushao mianmo I use-EPF many mask

“I had used many masks.’

c. 你還在用免洗筷子嗎？ [Google, 2013/6/22]

ni hai zai yong mianxikuaizi ma

you still PRG use disposable-chopsticks

‘Are you still using disposable chopsticks?’

d. 他們拿著筆記本電腦，用著蘋果手機、 [Google, 2013/3/13]

tamen nazhe bijibendiannao, yongzhe pingguo shouji they take-SI laptop use-SI Apple smart-phone

‘They takes laptops, uses Apple smart phones.’

As mentioned in chapter 3 and 4, PR-V YONG can be affixed with all the four aspectual markers but pro-V YONG cannot. Pro-V YONG can only be construed with le, guo, zai, but not with zhe. A comparison among I-V, PR-V and pro-V YONG indicates that I-V and PR-V YONG are equally verb-like in their grammatical status while pro-V YONG is less verb-like and thus more “grammatical”. Hence, the degree of grammaticality of the three functions of YONG would be: pro-V YONG >> PR-V ≒ I-V. So far, four hypothetical processes of the development of the novel uses of YONG in TM can be proposed:

(26) a. I-V > PR-V b. I-V > pro-V

c. I-V > PR-V > pro-V d. I-V > pro-V > PR-V

As to (26a), considering the equally verb-like grammatical statuses of these two functions of YONG, if the PR-V YONG is derived from I-V YONG, this process will not be a process of grammaticalization. That is to say, this is a contact-induced process, but without further grammaticalization. As for (26b), if the pro-V YONG is derived from I-V YONG, this derivation will be a process of grammaticalization. However, this process would be unlikely to take place since I-V is not directly related to pro-V YONG semantically.

According to Brinton and Traugott (2005), grammaticalization involves metaphoric or metonymic semantic change in the process. The output target of grammaticalization is the result of semantic change from the source input. This implies that the input source and the output target in the process of grammaticalization must be associated with semantic extension. From the observation in the data, the semantic linking between I-V YONG and pro-V YONG is hardly to identify. Thus, due to the lack of semantic association between these two functions of YONG, process (26b) would be almost impossible to happen.

Besides, taking semantic relation into consideration, it can be figured out from the data that PR-V YONG shares more semantic similarities with pro-V YONG than I-V YONG does.

Thus, process (26c) seems more plausible than process (26b) in that the path _PR-V >

pro-V_ is a process based on semantic change and toward the more grammatical direction.

However, the process _PR-V > pro-V_ is a process of grammaticalization which had already completed in TSM. Namely, _PR-V > pro-V_ is not an internal development in the system of the target language TM. Therefore, (26c) is not exactly a process of

“contact-induced grammaticalization” as defined in the literature. Rather, it is a

“contact-induced” constructionalization, without a clear trace of grammaticalization within the target language. Finally, since unidirectionality is widely acknowledged as a crucial property of grammaticalization (cf. Haspelmath, 1999, 2004; Hopper and Traugott, 2003;

Heine and Kuteva, 2005 etc.), the possibility of process (26d) could be excluded due to the fact that the path _pro-V > PR-V_ is essentially a process of “degrammaticalization” (i.e.

from more to less grammatical, or from grammatical to lexical), which is then known to be against the general principle of unidirectinality of grammaticalization.

Taking all the hypothetical processes into consideration and examining them with criteria of (contact-induced) grammaticalization, it is clear that the emergence of the novel usage of YONG as pro-V and PR-V in TM is contact-induced, but not the result of contact-induced grammaticalization as defined in the literature. Nevertheless, it is also true that these novel functions in TM are transferred from TSM. Then, one question should be further asked: by what mechanism did these novel functions of YONG come into being in TM? To answer this question, another possible hypothesis will be proposed in the next section: relexification.