But Chinese IS an SVO language.
Ask Chris: I’m a native speaker of English and currently learning Mandarin but the word order seems to be different from typical SVO languages like English. For example,
‘把书本拿给我’ (give me the books) – SVO should be “拿书本给我”
‘从家里出去’ (leaving from the house) – How about “出去从家里”
‘他在家里睡觉’ (He sleeps at home) – Why is it not “他睡觉在家里”
Many of these sentences in Mandarin put the verb at the end after the object, instead of putting it before the object. So I really wonder if Mandarin Chinese is a real SVO language.
(Attention: This article contains Chinese text. Without proper rendering support, you may see question marks, boxes, or other symbols instead of Chinese characters.)
Chris answers: … Well I really did not expect that someone whose native language is not Chinese would like to ask me a question. Let’s talk about the issues in a straightforward way, just to make everything clearer. Basically you got two problems in your descriptions – yes, those are your problems. In your first sentence, you mix up the ba structure and the unmarked order of Standard Mandarin Chinese. That is quite common, even for a beginner syntactician who has little or limited knowledge of Standard Chinese; so that is not your fault, and just take it easy. But your second problem is that you do not – or maybe just fail to – differentiate objects and prepositional phrases; that will be a horrible problem. Considering that last year I received a question about the word order of Archaic Chinese and this time I got yours, I would like to provide a somewhat up-to-date summary of the history of word order of Chinese, set in an earlier Chomskyan view called ‘Principles and Parameter’”, and drawing from my intuition as a Chinese native speaker.
The most common, unmarked order of the Chinese language, including both Standard Mandarin Chinese and Archaic Chinese, is SVO; that is a historical issue I would like to address before we move to the analysis of Standard Mandarin Chinese. Once there was an argument about the SVO or SOV order of Archaic Chinese. Li and Thompson (1974), for example, hold the opinion that Archaic Chinese should have SOV order, since quite a few of the Sino-Tibetan languages now have that order. Little evidence, however, is discovered for that argument; most of the literary records of Archaic Chinese, especially the pre-Qin documents, demonstrate the features of an SVO language rather than an SOV one. Granted, Archaic Chinese has some structures that follow SOV order, but as far as I can tell, none of the structures is fully ‘unmarked’. For example, an unmarked sentence favours nominal structures as its subject and direct object, and no additional particle should be present in the sentence, but those sentences with SOV order either include a particular particle that triggers the SOV structure, or contain a pronoun as its direct object. In a typical unmarked sentence, such as ‘孟子(S)见(V)梁惠王(O)’ (Mencius met (or visited) King Hui of Liang), you can see that both the subject and the object are nouns (proper names), and the order is definitely SVO. The feature of SVO has been inherited by Middle Chinese, and then Standard Mandarin Chinese today.
As for Standard Chinese, an unmarked sentence, like ‘我今天吃了个苹果’ (I today eat-PAR CLA apple), is also definitely SVO; even for sentences containing a pronoun as the object, like “我揍了他” (I punch-PAR him), the word order is still SVO. Without the presence of a word like ‘把’ (ba) and ‘被’ (bei), and without the movement of the topic, the word order is always SVO; you can never find a sentence presenting an SOV order. This is the common view shared by syntacticians in Europe (Continental and UK); maybe those horrible Americans have some other ideas that I don’t know of.
So that argument answers your final question: yes, Standard Mandarin Chinese is an SVO language.
So how can we interpret your sentences in that way? Before I go further into the field of Chinese, I would like to stray a bit and talk about another language: German.
German main clauses demonstrate a variety of word orders, including SVO, VSO, OVS, SAOV, and even OASV if you pay enough attention to the language (A stands for an auxiliary verb, like ‘did’ or ‘have’ in English). But most commonly it is known as having V2-SOV order; that is, in a main clause, the verb will always be in the second position, while it will be at the end of a subordinate clause. I mention the structure of German, mainly to point out that it is possible for a language to have numerous word orders besides its unmarked order; the world is not made up of English, which only allows a limited number of word orders.
Let’s move back to Chinese then. Example 1 in the question is quite a complicated one involving both a double-object construction and the presence of ba; so it will take me longer and more examples to illustrate the variations in Standard Mandarin Chinese. Firstly, I would like to discuss the double-object construction, which means that the main verb is ditransitive, has three theta-roles, and requires three nouns (or nominal constructions) in a grammatical sentence; the three nouns will be the subject (S), direct object (DO), and indirect object (IO) respectively. For a typical ditransitive verb, like “给” (give), the unmarked word order will be:
(1). 他给了我一本书。
He give-PAR me one-CLA book. – in the form of S-V-IO-DO.
In this case, the word order of Standard Chinese is exactly the same as that of English, which is a typical SVO language: ‘he gave me a book’. The ditransitive verb construction is usually referred to as a VP-shell structure; for those who are interested in how Chomsky and his fellows solve the problem of the theta-role assignment of a ditransitive structure, just type this word into Google and you will get piles of literature.
Then we talk about the application of ba in constructions with a transitive (not ditransitive) verb. The nature of ba is rather complicated and even controversial; here I would like to follow the light verb assumption (gladly we have Julio’s post on light verbs; that saves my day), in which the nature of ba is a bit similar (but not equal) to the auxiliary verb in German. A light verb is a verb (of course), but its function is more like a particle which leads to the shift of structure within the sentence, e.g. focus, stress, and so on. For a typical transitive verb, like ‘吃’ (eat), both unmarked and ba-construction sentence express the same meaning:
(2a). 我吃了一个苹果。
I eat-PAR one-CLA apple. – in the form of SVO.
(2b). 我把一个苹果吃了。
I ba one-CLA apple eat-PAR. – in the form of S-ba-O-V
(2c). *我一个苹果吃了。
*I an-CLA apple eat-PAR. – in the form of SOV. An asterisk indicates that the sentence is ungrammatical to native speakers like me.
Both (2a) and (2b) are grammatical and acceptable by a native speaker; (2c), a pure SOV order, sounds horrible unless uttered with a particular intonation pattern. We can see that the SOV order is possible only when ba is present.
Then we put the two together and go to the example 1 in the question, which is a combination of the double-object and ba-constructions. For the ba-construction of (1), we can list a set for comparison:
(3a). 他给了我一本书。 (a replica of example (1).)
He give-PAR me one-CLA book. – in the form of S-V-IO-DO.
(3b). 他把一本书给了我。
He ba one-CLA book give-PAR me. – in the form of S-ba-DO-V-IO.
(3c). #他给了一本书我。
# He give-PAR one-CLA book me. – in the form of S-V-DO-IO. I use the hash here because in some Chinese dialects, e.g. Cantonese, the structure is perfectly acceptable, and I also see some of my Hong Kong friends using it in their variety of Standard Chinese; I would like to ignore it here because I’m talking about Standard Chinese, but the sentence is not that ‘standard’.
(3d). *他把我给了一本书。
* He ba me give-PAR one-CLA book. – in the form of S-ba-IO-V-DO.
For other possible structures, most of them are ungrammatical and unacceptable to native speakers; that includes all the structures in which V is at the very end of the sentence, including but not limited to S-IO-DO-V and S-ba-DO-IO-V. There should be at least one component after the main verb of the sentence for it to be grammatical; for Standard Mandarin Chinese, it can never be the case in a double-object construction that a sentence is a pure SOV sentence – which is rather convincing in showing that the structure of DO-V in both transitive and ditransitive sentences is a result of triggered movement, rather than a based-generated structure.
For a typical SOV language – unless certain extraposed structures are present (which is too complicated for non-syntacticians) – all objects should be in front of the main verb. At the same time, it is grammatical for an SOV language, like Japanese or German, to exchange the position between IO and DO – a process called ‘scrambling’. Scrambling, as a widely-present feature among OV languages, is impossible in Standard Chinese; this can be another piece of evidence, although the logic is not fully convincing. Sad story.
As for the remaining two examples, ‘从家里出去’ (from home-in go out) and ‘在家里睡觉’ (at home-in sleep), they belong to the problem of propositional phrases (PP) rather than that of objects. In Archaic Chinese, the position of a PP in relation to a main verb is rather flexible (or in Chomskyan terminology, a “free parameter”), but in Standard Mandarin Chinese, a PP attaching to a VP is more frequently a pre-verbal one. This is definitely interesting – but also more difficult; so if you think the following content may be beyond your grasp (which, actually, is the case for some of my classmates in our linguistics masters programme), just skip it and jump to the end.
Contemporary syntactic theories, especially the Chomskyan one, assume that any given sentence forms a tree structure, the large tree containing a series of smaller trees, corresponding to phrases. For every phrase, there is a head (the most important word), a complement providing the essential part of the rest of the phrase, and a specifier which is more or less like a modifier. For a VP including a transitive verb, the verb itself is the head, the object is its complement, while the adverbial (e.g. a PP) is its specifier.
In Standard Chinese, most of the time, the specifier is in front of the head, and that structure is less flexible. For instance, an NP (noun phrase) is in the form of AP-N, which is why we say ‘美丽的|姑娘’ (beautiful girl) rather than ‘姑娘|美丽的’ (girl beautiful); this is the same in the construction of the VP, so that we put the PP in front of the VP, and say ‘从家里出去’ rather than ‘出去从家里’, or ‘在晚上|看电视’ (in the evening watch TV) rather than ‘看电视|在晚上’ (watch TV in the evening).
Since we are talking about the specifier but not the complement when we discuss the PP-V construction in Chinese, that structure can never be used to argue for an SOV structure. The example ‘在晚上|看电视’ (in the evening watch TV, in the form of PP-V-O), on the contrary, obviously shows that Standard Chinese is a VO language.
I will not further extend my argument into an analysis of Greenberg’s Linguistic Universals, in which he suggests that a VO language always has head-complement order (while Chinese does not); that will be beyond the reach of 99% of my readers, and may even earn me a page in Linguistic Inquiry. I hope that it is clear enough for you to know why Standard Mandarin Chinese is an SVO language rather than an SOV one: basically, what you regard as an OV structure is not fully qualified. Next time try some more delicate examples and you are always welcome to ask me more questions.
Enjoy learning Chinese! All the best, Chris.
You do not need to learn Chinese in order to be a syntactician. Try reading these:
Greenberg, Joseph H. Language universals: With special reference to feature hierarchies. Walter de Gruyter, 2005.
Haegeman, Liliane. Introduction to government and binding theory. 1991.
Li, Charles N., and Sandra A. Thompson. “An Explanation of Word Order Change SVO→SOV.” Foundations of Language, vol. 12, no. 2, 1974, pp. 201–214. www.jstor.org/stable/25000832.
Peyraube, Alain. “On word order in Archaic Chinese.” Cahiers de linguistique-Asie orientale 26.1 (1997): 3-20.
But Chinese IS an SVO language. was originally published on CamLangSci
