人物专栏 | Kwang-Sup Kim教授访谈（上）
点击上方蓝字关注我们
编者按
《理论语言学五道口站》（2021年第64期，总第198期）“人物专栏”与大家分享本站采编人员王平对Kwang-Sup Kim教授进行采访的访谈录。Kwang-Sup Kim（金光變)，美国马里兰大学帕克分校博士，现任韩国外国语大学理论语言学、英语语言学教授。主要研究兴趣为句法理论、比较句法、句法与语义接口以及句法与音系接口。
此次访谈中，Kwang-Sup Kim教授首先结合韩语实例，提出了对轻动词理论的看法，然后对不同语言的参数差异等问题进行了讨论。
本次访谈内容共分为上下两期，后续内容将在下一期人物专栏中继续与大家分享，敬请期待。
人物简介
Kwang-Sup Kim教授
Kwang-Sup Kim（金光變），马里兰大学帕克分校博士，现任韩国外国语大学理论语言学、英语语言学教授。主要研究兴趣为句法理论、比较句法、句法与语义接口以及句法与音系接口。著有Minimalism and Last Resort (2008)（《最简方案和“最后一招”》）一书，并于Linguistic Inquiry、Lingua、Studies in Generative Gramma等期刊发表过多篇文章。曾任韩国生成语法学会主席及韩国外国语大学高翻学院院长。
Brief Introduction
Kwang-sup Kim received his Ph.D. from the University of Maryland, College Park and is currently Professor of General and English Linguistics at Hankuk University of Foreign Studies, Korea. His main interests are in syntactic theory, comparative syntax, the syntax-semantics interface, and the syntax-phonology interface. He wrote a book on the last resort strategies in minimalism, Minimalism and Last Resort (2018), and published numerous articles in journals such as Linguistic Inquiry, Lingua, and Studies in Generative Grammar. He also published many book chapters, which were published by Oxford University Press, Routledge, and Hankook Publishing Co. He was president of the Korean Generative Grammar Circle and Dean of College of Interpretation and Translation, Hankuk University of Foreign Studies.
访谈内容
01.
王平：传统意义上来说，词类包括动词、形容词、名词、副词等。随着研究的日益深入，语言学家们引入了“轻动词”这一概念。在您看来，轻动词应该属于一个独立的词类还是动词的一个子类呢？
Kwang-Sup Kim教授：在传统语法中，人们认为像kick这样的动词是不能进一步分析的。然而在最简方案中，我们认为动词kick可以进一步分析，准确来说，它是由轻动词v和它的补足语构成。轻动词类似于动词转换器，将其自身的补足语转化成动词。如（i）所示，乔姆斯基（2015）指出，轻动词的补足语“kick”是一个词根，没有自己的语法范畴。
(i) [v [root kick]]
如果只存在隐性轻动词，那么轻动词的存在就不那么可信。然而，事实上轻动词既有显性的，也有隐性的。就韩语来说，既包括显性轻动词又包括隐性轻动词。如(ii)所示，当形容词coh'good'与显性轻动词ha'do'合并时就变成了动词，表示“喜欢”的含义。
(ii) coh ‘good’ + ha ‘light verb’ = coha ha ‘like’
当然，在韩语中也有隐性轻动词。如（iii）所示，kkamppakgeli ‘blink (眨眼)’ 既可以表示不及物动词，也可以表示及物动词。对于这种现象可以理解为词根kkamppakkeli不同情况下分别与及物性轻动词v*和不及物性轻动词v的合并。
(iii)a. kkamppakkeli + v* = transitive verb
b. kkamppakkeli + v = intransitive verb
综上所述，动词远比它们看起来要复杂，它们是由轻动词和“动词词根或非动词成分”一起构成的。简要概括来说，轻动词是动词的核心，它并不属于一个独立的词类。
02.
王平：轻动词的引入主要是为了解决“格的指派”与“论元选择”的一些问题，乔姆斯基没有对轻动词进行区分。而部分学者，如Kratzer（1996）对轻动词进行了详细的区分。所以我们是否有必要对轻动词进行详细的分类呢？您能谈谈对这个问题的看法吗？
Kwang-Sup Kim教授：如果我们认同轻动词是“动词转换器”，那么轻动词至少包括两种类型：一种是“及物性轻动词”，另一种是“不及物性轻动词”。当然，除了上述两种，轻动词还包括其他类型。前面我提到过，在韩语中ha是一个显性轻动词，它就有着不同的功能。首先，作为一个动词转换器，它后面可以跟不同类型的补语，如(i)中所示，ha可以把其后的形容词、从句、词根都转变为动词。
(i)
a. kippe ‘happy’ + ha ‘light verb’ = kippe ha ‘pleased’
b. [kippe + key] + ha ‘light verb’ = kippu-key ha ‘make someone pleased’
此外，韩语中的ha也可以作为“形容词转换器”。当ha附加到副词kule上时，kippe ha则变成形容词，如(ii)所示。
(ii) kule ‘so’ + ha = kule ha ‘so’
所以这样来看，显性轻动词的分类不止一种，那么隐性轻动词更是如此了。
03.
王平：尽管大多数学者都认为名词性结构的核心都是“名词”，但是也有人认为不然。对此您秉持的观点是名词性结构的核心不应该是“(N)名词”，而应该是“(D)限定词”也就是说，名词短语、形容词性短语都可以作为“(D)限定词”的补足语，从而组成名词性结构。此外，“限定词”后也可以存在没有补足语这种情况。当然，该理论在一些具有明显限定词的语言中比较有说服力。但就汉语而言，其限定词是隐性的，而且其限定词短语往往没有补足语。对此您有什么看法呢？
Kwang-Sup Kim教授：部分学者认为语言可以分为两种类型：NP语言与DP语言（Bošković 2008, 2010）。这两种语言其实都是有限定词的，但是在NP语言中，限定词并不能进行投射。韩语和汉语类似，限定词做名词短语的修饰语。如(i)所示，当韩语量化词motu（全，都）出现在名词前，其必须带上一个形容词性的后缀n。
(i) Motu-n haksayng
All-NP-modifying marker student
另外值得注意的是，量化词motu“全”也可以放在名词后面，如(ii)所示。
(ii) haksayng motu
Student all
在例(ii) 中，量化词motu“全”充当了该名词性结构的核心。如果这种分析可行的话，那么也可以说韩语是一种兼具NP/DP的语言。这也就意味着，语言的类型可以分为三种：NP语言，DP语言以及NP/DP语言。因此，这个问题可以理解为并非所有的语言都有DP（限定词短语）这个层级，但这又不意味着这些没有DP的语言中不存在限定词，也就是说它们的限定词并没有占据核心位置，而是变为了名词短语的修饰语。
04.
王平：生成语法认为语言具有普遍性。然而，英语中的一些句法操作，如“Do-insertion”和 “There-insertion”现象在其它的语言中并不存在。您认为这些区别仅仅是因为语言间的参数不同吗?
Kwang-Sup Kim教授：语言间是有参数差异的。插入规则确实不具有普遍性，很大程度上也并不容易被大家认同和接受。但是值得注意的是，“Do-insertion”现象并非英语独有，其实在韩语中也有类似的现象。比如，在英语中do一般出现在否定句和强调句中（如(i)所示），韩语中ha也可以出现在否定句和强调句中。
(i) a. John does not like Mary.
b. John DOES like Mary.
类似结构在韩语中也得到了证实。如在否定句和对比话题结构中ha的使用。
(ii)
a.John-i Mary-lul cohaha-ci ani-ha-n-ta.
John-NomMary-Acclike-CInot-HA-Pres-Dec
‘John does not like Mary’
b.John-i Mary-lul cohaha-ki-nun ha-n-ta.
John-NomMary-Acclike-Com-Contrastive TopicHA-Pres-Dec
‘It is true that John like Mary, but…’
这些规则在狭义句法中并不适用，但在PF层可以得到解读。使用这些规则是为了避免出现词缀无依附等问题。在韩语和英语中，时态可以通过词缀体现，当时态无依附的时候，do/ha就会出现。汉语因为没有显性时态词素，所以也就不会有插入操作。然而，德语、法语、西班牙语以及许多其它语言有显性的时态词素，它们的动词移位到T，T不会无依附，因此不需要插入规则。
English Version
01.
Ping Wang: Traditionally, word categories include verb, adjective, noun, adverb, etc. As linguistic research becomes increasingly mature, linguists introduced light verbs to solve the problems of case and argument. Do you think light verbs should be regarded as an independent category or a subcategory of verb？
(i) [v [root kick]] If there is only a covert light verb, its existence is less convincing. However, there are overt light verbs as well as covert. In Korean both types of light verbs are present. For instance, the adjective coh ‘good’ turns into a verb when it is merged with the overt light verb ha ‘do’. If coh is merged with ha, the grammatical category of the resulting word is verb and denotes ‘like’.
(ii) coh ‘good’ + ha ‘light verb’ = coha ha ‘like’
Of course, in Korean there is a covert light verb as well. For instance, kkamppakgeli ‘blink’ is ambiguous between an intransitive verb and a transitive verb. This can be analyzed as follows: the root kkamppakkeli may be merged with a transitive light verb v*, or it may be merged with an intransitive light verb v.
(iii) a. kkamppakkeli + v* = transitive verbb. kkamppakkeli + v = intransitive verb
In short, verbs are more complex than they appear, and they consist of light verbs and roots or non-verbal constituents. To provide a short answer to the question, the light verb is the head of V, so that it is not an independent category.
02.Ping Wang: The introduction of light verbs is targeted at the problems of case and argument. Chomsky did not distinguish them in his series of works. However, Kratzer (1996) and other scholars have made a detailed distinction between light verbs. So is it necessary for us to make a subordinate distinction between light verbs? Could you talk about your views on this issue? Prof. Kwang-Sup Kim: If we adopt the view that light verbs are verbalizers, there are at least two types of light verbs: one is a transitive verbalizer, and the other is an intransitive verbalizer. In addition, there are many other types. I mentioned that in Korean ha is an overt light verb. There are various types of ha. Basically, it is a verbalizer, and it can take various types of complements. It can take an adjective, a clause, and a root as its complement, turning them into verbs. (i)a. kippe ‘happy’ + ha ‘light verb’ = kippe ha ‘pleased’b. [kippe + key] + ha ‘light verb’ = kippu-key ha ‘make someone pleased’ In fact, Korean ha can also function as an adjectivalizer. Kule is an adverb, and when it is attached by ha, an adjective is generated. (ii) kule ‘so’ + ha = kule ha ‘so’ There are various types of overt light verbs. If so, it is not surprising that there are various types of covert light verbs.
03.Ping Wang: There are nominals with no N although there is no antecedent for the missing N. You propose that the absence of N is not surprising, since D is the head of a nominal, and NP is just one type of complements that D subcategorizes for: besides NP, D can take AP as its complement, and it may take no complement at all. This phenomenon may not be surprising in the languages with overt determiners, but it is somewhat inconceivable in Chinese, a language with covert determiners, especially D take no complement. How will you explain these cross-linguistic differences with your theory? Prof. Kwang-Sup Kim: We can classify languages into two types: NP-languages and DP-languages (Bošković 2008, 2010). NP-languages as well as DP-languages have determiners, but the determiner do not project their labels in NP-languages. In Korean, just like Chinese, determiners are NP-modifiers. For instance, motu ‘all’ must be marked by an adjectival ending n, when it precedes a noun. (i) Motu-n haksayngAll-NP-modifying marker student In fact, it is possible to place the quantifier motu right after the noun. (ii) haksayng motu Student all In (ii) the quantifier motu ‘all’ can be taken to be the head of the nominal. If this analysis is correct, Korean is a NP/DP language. This means that there are three types of languages: DP-languages, NP-languages, and NP/DP languages. The point is that not every language has the category DP. This does not mean that there is no determiner in those languages, but the determiners are not the heads taking NP-complements, but they are NP-modifiers.
04.Ping Wang:Generative grammar holds that language is universal. However, some syntactic manipulations commonly found in English, such as Do-insertion and There-Insertion, do not seem to exist in other languages. Do you think these differences can only be classified as differences in parameters? Prof. Kwang-Sup Kim: Yes, there are parametric differences. Insertion rules are not likely to be universally attested. They are last resort operations that must be avoided if possible. But it is noteworthy that English is not alone in making use of Do-insertion. Korean has a similar operation. Ha-insertion takes place in negative and emphatic sentences. It is well-known that do is inserted in negative and emphatic sentences in English. (i) a. John does not like Mary.b. John DOES like Mary. Quite an analogous pattern is attested in Korean. In negative sentences and contrastive topic constructions, dummy ha shows up. (ii) a. John-i Mary-lul cohaha-ci ani-ha-n-ta.John-Nom Mary-Acc like-CI not-HA-Pres-Dec‘John does not like Mary’ b. John-i Mary-lul cohaha-ki-nun ha-n-ta.John-Nom Mary-Acc like-Com-Contrastive Topic HA-Pres-Dec ‘It is true that John like Mary, but …’ These insertion rules do not apply in the narrow syntax, but apply at PF. They are required in order to avoid stray affixes. In Korean and English, tense is an affix, and when tense is stranded, do/ha-insertion takes place. Chinese is known to have no overt tense morpheme.So it is not likely to have that kind of insertion operations. German, French, Spanish, and many other languages have overt tense morphemes, but they have no corresponding insertion rules, because verb must move to T in those languages, and so T is not stranded.
往期推荐
Noam Chomsky | Mind Your Language
本文版权归“理论语言学五道口站”所有，转载请联系本平台。
编辑：闫玉萌 赵欣宇 雷晨
排版：闫玉萌 赵欣宇 雷晨
审校：李芳芳 田英慧