人物专栏 | Masatoshi Honda教授访谈录

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33期，总第167期）“人物专栏”与大家分享本站采编人员马晓彤对Masatoshi Honda教授进行采访的访谈录。Masatoshi Honda，日本宫崎大学教育学院副教授，日本筑波大学人文社会科学研究生院语言学博士。其研究领域为生成语法中小句域和名词域的句法结构，尤其是句法制图框架等相关研究。

本期访谈中，Masatoshi Honda教授首先就英语的言据性进行解释，其次介绍了生成语法中小句域和名词域的句法表现，随后对焦点的语言学研究以及汉英焦点表达的异同给出了自己的见解，最后给出了自己对于跨语言研究的建议，带给我们许多启发。

人物简介

Masatoshi Honda教授

Masatoshi Honda（本多正敏），宫崎大学教育学院副教授，筑波大学人文社会科学研究生院语言学博士。曾任横滨商学院英语教育中心助理教授、神田国际研究大学儿童英语教学中心研究员。其研究领域为生成语法中小句域和名词域的句法结构，尤其是句法制图框架等相关研究。目前他的研究兴趣是与语篇相关的词序变化，包括对比/非对比焦点前置、分句类型(如wh-感叹句)、感叹句/祈使句、言据性和意外的语言表现以及述谓。

Brief Introduction

Masatoshi Honda is an associate professor of the Faculty of Education, University of Miyazaki. He obtained a Ph.D. in Linguistics at the Graduate School of Humanities and Social Sciences at the University of Tsukuba.He worked as an assistant professor at the English Education Center at Yokohama College of Commerce and as a researcher at the Center for Teaching English to Children at Kanda University of International Studies. His primary research focus is on the syntactic articulation of the clausal and nominal domain within generative grammar, more specifically the cartographic framework. His current areas of interests are discourse-related word order alternations such as contrastive/non-contrastive focus fronting, clause typing (e.g. wh-exclamatives), exclamatory/optative sentences, linguistic realization of evidenitiality and mirativity, and predication. 

访谈内容

01.

马晓彤：与日语不同，英语是一种通常不带有言据性的语言；英语句子的言据性解释依赖于语境（Honda et al., 2015）。除了语境，您认为在英语中是否存在一些标志着言据性的词汇及表达方式？如果有的话，这些词汇和表达方式之间有何共性呢？

Masatoshi Honda教授：在语言类型学领域，言据性的定义为“一种主要用来表达信息来源的语言范畴”（Aikhenvald（2004：3））。一些语言（如日语）中存在蕴含言据性的语法成分（如：句末语气词），然而其他语言中并没有此类专门表达言据性的语法成分，英语就是其中之一。尽管英语缺少语法言据性成分，语言学家们认为英语可以通过某些类别的副词和动词短语来表达言据性（Anderson（1986），Chafe（1986），Cinque（1999），Rooryck（2001），Fox（2001），Aikhenvald（2004））。

在制图框架内，Cinque（1998）有关副词层级性的研究可以看作是对于言据性影响颇深的理论方法之一。Cinque的普遍层级性假定，通过Spec-Head一致性关系（[frankly Moodspeech act [fortunately Moodevaluative [allegedly Moodevidentiality [probably Modepistemic [once T（Past）… ]]]]]），副词被允准出现在相应的指示语位置。普遍层级性还假定言据性应该与情态（Mod（ality））这一句法范畴区别开来，相反，言据性应该属于语气（Mood）这一句法范畴。

Cinque（1998）的普遍层级性是很多学者研究跨语言言据性表达的基础。例如，Rooryck（2001）对插入语（即：主语和一个内嵌子句的动词结合，比如I hear的听说证据标记）的研究可被视为将Cinque的普遍层级性应用到英语言据性表达的尝试。Rooryck将插入语的句法与语义地位识别为言据性成分，并提议插入语的生成包含下述两步（i）将CP补足语前置到[Spec, MoodEvidP]位置；（ii）将主句动词显性或隐形移位到MoodEvid中心位置。

（1）

a. [MoodEvidP [CP John is back] [MoodEv° think] [TP I [T° think] [VP I think CP ]]]

b. [MoodEvidP [CP John is back] [MoodEv° said] [TP she [T° said] [VP she said CP ]]]

（1a） 的句法配置展示了传统上被成为重组（slifting）的生成过程，然而（1b）对应引用语倒装（quotative inversion）。Rooryck的分析让我们把重组（slifting）和 引用语倒装（quotative inversion）统一到言据性的句法实现。

尽管不可能提供一个包含英语中所有言据性标记成分的列表，我想在下文提及其中两种。其一是包含虚主语it的表达方式（如：X take it that …,  X have it that …, it be that …）（cf. Ikarashi（2015））。这种习语形式专门用来表达在主语位置出现的信息来源（例如，Rumor has it that John was fired.）。其二涉及传统上被认为有展示功能的倒装句。根据Shizawa（2015），如果说话者对发生在她/他面前的情况没有直接感知的证据，则不能使用英语中的地点倒装（如：Into the room came a cat.）。如果这一观察是正确的，地点倒装将被视为一种实现言据性（即直接感知）的句法结构。尽管对于把地点倒装看成一种言据性的句法实现方式仍需进一步的经验证据，这一方法对于言据性的研究提供了新视野。

02.

马晓彤：生成语法是20世纪下半叶最具影响力的语言学理论，您能介绍一些关于生成语法中小句域和名词域的句法表现吗？

Masatoshi Honda教授：自生成语法发展的早期开始，就一直在讨论小句域（CP）和名词域（NP）之间的某些对称性：语法功能，比如主语和宾语（如：The enemy destroyed the city./ the enemy’s destruction of the city）（Chomsky（1970））；转换规则，比如被动化（如：The city was destroyed by the enemy./the city’s destruction by the enemy）（Chomsky（1970））；曲折/一致关系 （如Szabolcsi（1984），Abney（1987），Takano（1989））；从属连词（如在介绍句子或名词论元时由句子中的矩阵谓词选择的C/D核心的作用（Szabolsci（1987，1994），Aboh（2004a））；谓语倒装（如：Den Dikken（1998, 2006））。这些研究表明，在这两个域的短语结构和它们所允许的句法操作方面，小句域和名词域中存在强烈的对称关系。

最近与小句域和名词域之间平行关系相关的主要问题之一是，这一关系是否能够拓展到信息结构的概念上。小句域和名词域构成了话语相关属性编码的结构区域，这一观点在制图框架中得到了理论上的实现（更多细节请查看Aboh, Corver, Dyakonova & van Koppen（2010））。就目前而言，关注信息结构方面小句域和名词域之间的对称性的研究至少有两个问题：（1）小句域和名词域之间的平行性在信息结构层面是否存在？（2）这种平行的观点是如何从跨语言的角度得到经验支持的？

03.

马晓彤：CP分裂假说认为CP域包含两种焦点功能投射:对比焦点和信息焦点。在您看来，哪种类型的焦点在最近的语言学研究中更重要？可以谈谈其中的原因吗？

Masatoshi Honda教授：Cruschina （2011）的CP分裂假说与Rizzi（1997）的CP分裂假说不同，前者认为CP域中的焦点可进一步分解为两个独立的句法范畴：对比焦点和信息焦点。Cruschina提出该假设正是基于他对罗曼语（包括他自己的方言，西西里语）的观察、分析。他发现这些语言中存在焦点前置现象并进行了对比分析。同时他还发现这些语言允许对比和非对比焦点前置。尽管对比焦点投射和非对比焦点投射的二分法允许我们从跨语言的角度探索它们之间潜在的形态句法差异，但对比焦点投射包括单个焦点投射还是两个焦点投射仍需进一步的讨论。

根据Cruschina（2011）的观点，非对比焦点前置包括QP前置（极性/过度焦点前置）和意外前置。后者是以“意外（Mirativity）”的概念命名的，指的是“传达说话者新得知的或意想不到的信息的语言标记”（Delancey（2001：369-370）。Cruschina的研究促使学者们从跨语言的角度来探索意外前置的存在（例如，Trotzke（2016）对德语中意外前置的研究；Authier和Haegeman（2019）对法语中意外前置的研究；以及Jiménez-Fernández（2015）和 Cruschina（2019）对西班牙语中mirative前置的研究。因此，非对比焦点前置研究是焦点语法标记实现研究的前沿课题。

04.

马晓彤：焦点现象几乎存在于所有语言中，研究者们对全世界范围内各种语言存在的焦点现象进行了大量的研究。然而，焦点的表达具有跨语言的差异性。例如，英语中，用“It is… that…”句型中的特殊句法位置来标记焦点；汉语中，则更多地用语用重音强调焦点。在您看来，汉英焦点表达的异同是什么？

Masatoshi Honda教授：虽然我不能直接回答这个问题，但简单比较一下语音重音和分裂结构就能给我们一些启发。在关于焦点的句法和语音实现的文献中，有人认为语音重音的语用效果应该与语法焦点的语用效果区分开来。例如，Horvath（1986）认为，在匈牙利语中，句子的焦点必须出现在动词之前。如果焦点成分（即一个音系重音成分）在该语言中占据了一个其他的句法位置而非动词前位置，这个句子就会不合语法；根据Horvath的观点，这样的句子只有在被解释为实现了元语言功能（例如，纠正另一个说话者对某个单词的发音错误）的情况下才能被接受。É. Kiss（1998: 245）认为匈牙利语中的动词前位置和英语中的分裂结构的焦点位置在语法上实现了可识别焦点，可识别焦点“代表谓词性短语可能拥有的上下文或情景给定元素集合的子集；被识别为谓词性短语实际所包含元素的集合的完备子集。”鉴于这一研究背景，我们可以说元语言纠正是通过韵律来实现的，必须与语法手段如分裂结构所表达的焦点分开；Cruschina（2011）还认为，元语言纠正不涉及意大利语的焦点前置。

因此，我很好奇汉语中的语用重音的性质和焦点的语法实现的本质；那么，一个相关的问题是，汉语中的语用重音和由语法实现的焦点之间是否存在关联。研究这些问题将对理解语言学理论中焦点的语音和语法实现做出经验和理论上的贡献。

05.

马晓彤：据我们所知，您不仅做了许多日语的研究，还做了很多德语的研究，关于不熟悉语言的句法研究，您有什么建议?

Masatoshi Honda教授：如果我们从生成语法发展的较早时期开始观察生成语法，会发现生成语法的研究原则之一是比较法。 生成语法框架下比较研究的主要目标是回答（至少回答）以下两个传统的核心问题：（1）什么是语言？（2）如何习得语言？为了解决这两个问题，许多基于原则与参数方法的研究一直致力定义什么是所有语言都有的一套基本原则，以及什么是决定语言间句法多样性的一套参数（如pro-drop参数，即决定一门语言是否允许句子省略主语）。

对我而言，从之前基于原则和参数的（传统的或经过改进的）研究中吸取经验，总是能为我从不同的角度研究目标语提供一种重要的方式（或者一种分析性的工具）。幸运的是我们有许多入门级的学术著作，概述了原则与参数理论采取的研究方法及其应用。我个人认为学习生成语法的发展过程和它对语言知识的见解，是我们知道如何比较第一语言、目标语言和不熟悉语言的一种好方法。

English Version

01.

Xiaotong Ma: English, unlike Japanese, is a language that does not normally encode evidentiality; the evidential interpretation of a sentence relies on the context (Honda et al., 2015). Except the context, do you think that there are some words or expressions whose appearances stand for the evidential expressions in English? If there are, what are similarities of these words and expressions?

Prof. Masatoshi Honda: In the field of linguistic typology, evidentiality is defined as “a linguistic category whose primary meaning is source of information” (Aikhenvald (2004: 3)). Some languages (e.g. Japanese) have grammatical items encoding evidentiality (e.g., sentence final particles), whereas other languages have no such grammatical items specialized in expressing evidentiality. Among those languages is English. Although English lacks grammatical evidential elements, it has been argued that English express evidentiality by means of certain types of adverbs and verb phrases (Anderson (1986), Chafe (1986), Cinque (1999), Rooryck (2001), Fox (2001), Aikhenvald (2004)).

Within the cartographic framework, Cinque’s (1998) research on the universal hierarchy of adverbs will be regarded as one of the influential theoretical approaches to evidentiality. In Cinque’s universal hierarchy, it is assumed that adverbs are licensed in their respective specifier positions via a Spec-Head agreement relation: [frankly Moodspeech act [fortunately Moodevaluative [allegedly Moodevidentiality [probably Modepistemic [once T(Past) … ]]]]]. The universal hierarchy also assumes that evidentiality should be treated as separated from the syntactic categories of Mod(ality); instead, it is assumed to belong to the syntactic categories of Mood.

Cinque’s (1998) universal hierarchy constituted a solid basis on which many scholars analyze evidential expressions across languages. For example, Rooryck’s (2001) study on parentheticals (i.e., a combination of the subject and a clause-embedding verb such as I hear for a hear-say evidential marker) will be seen as an attempt to apply Chinque’s universal hierarchy to English evidential expressions. Identifying the syntactic and semantic status of parentheticals as that of evidentials, Rooryck proposes that parentheticals are derived by (i) fronting the complement CP to [Spec, MoodEvidP] and (ii) overt or covert movement of the matrix verb to the MoodEvid head.

(1) 

a. [MoodEvidP [CP John is back] [MoodEv° think] [TP I [T° think] [VP I think CP ]]]

b. [MoodEvidP [CP John is back] [MoodEv° said] [TP she [T° said] [VP she said CP ]]]

The syntactic configuration in (1a) shows the derivation of what is traditionally known as slifting (Ross (1973)), whereas the one in (1b) corresponds to quotative inversion. Rooryck’s analysis will allow us to unify slifting and quotative inversion into the syntactic realization of evidentiality.

Although it is impossible to provide an exhaustive list of evidential markings in English, I would like to mention two other groups of them below. One is concerned with a set of expressions which involve the expletive it. (e.g., X take it that …,  X have it that …, it be that …) (cf. Ikarashi (2015)). Take the form have it that for example. This idiomatic form is specialized in expressing the source of information which appears in subject position (e.g., Rumor has it that John was fired.). The other involves certain inversion sentences which are traditionally known as involving the presentational function, which introduces a new referent into the discourse. According to Shizawa (2015), locative inversion in English (e.g., Into the room came a cat.) cannot be used if the speaker has no direct-perceptive evidence for the situation occurred in front of her/him. If this observation is correct, locative inversion will be seen as a syntactic device which realizes evidentiality (i.e. direct perception). Although further empirical justification will be necessary for the treatment of locative inversion as a syntactically-realized form of evidentiality, such an approach will provide new insights into the research on evidentiality.

02.

Xiaotong Ma: Generative Grammar is the most influential linguistic theory of the second half of the twentieth century. Could you introduce something about the syntactic articulation of the clausal and nominal domain within generative grammar?

Prof. Masatoshi Honda: Since the earlier stages in the development of generative grammar, certain symmetrical properties between the clausal domain (CP) and the nominal domain (DP) have been explored in terms of the following viewpoints: Grammatical functions such as subjects and objects (e.g., The enemy destroyed the city./ the enemy’s destruction of the city) (Chomsky (1970)); transformational rules such as passivization (e.g., The city was destroyed by the enemy./the city’s destruction by the enemy)) (Chomsky (1970)); inflection/agreement (e.g., Szabolcsi (1984), Abney (1987), Takano (1989)); subordinator roles (e.g., the role of the C/D head in introducing a sentential/nominal argument selected by the matrix predicate in a sentence (Szabolsci (1987, 1994), Aboh (2004a)); predicate inversion (e.g., Den Dikken (1998, 2006)). These studies imply that there are strong symmetrical relations between the clausal and nominal domains in terms of the phrase structure of these two domains and the set of syntactic operations that they allow.

One of the recent issues concerning the parallelism between the clausal domain and the nominal domain is whether it extend to the notion of information structure. The idea that the clausal domain and the nominal domain constitute the structural zone in which discourse-related properties are encoded is theoretically implemented in the cartographic framework (See Aboh, Corver, Dyakonova, and van Koppen (2010) for details and examples of discourse-related word order alternations in the nominal domain). As of the present, there seem to be at least two ongoing issues concerning a search of symmetry between the clausal domain and the nominal domain in terms of information structure: (i) whether the parallelism between the clausal and nominal domain actually is maintained at the level of information structure and (ii) how such a parallel view is empirically supported from a cross-linguistic perspective.

03.   

Xiaotong Ma: According to the Split CP Hypothesis, CP domain includes two functional projections for focus: Contrastive Focus and Information Focus. In your opinion, which type of focus is more important in recent linguistic studies? And what is the reason?

Prof. Masatoshi Honda: Cruschina’s (2011) split CP hypothesis departs from Rizzi’s (1997) original one in that the former involves two independent syntactic categories for focus in the CP domain: Contrastive Focus and Information Focus. Cruschina’s hypothesis is based on his comparative analysis of focus fronting phenomena observed in Romance languages (including his own dialect, Sicilian) and its crucial finding that they allow contrastive and non-contrastive focus fronting. Although it is still an ongoing issue whether the CP domain includes a single focus projection or two focus projections, we can safely say that the dichotomy of contrastive focus fronting and non-contrastive focus fronting allows us to explore (potential) morpho-syntactic differences between them from a cross-linguistic perspective.

According to Cruschina (2011), non-contrastive focus fronting involves QP-fronting (polarity/verum-focus fronting) and mirative fronting. The latter is named after the notion of “Mirativity”, which “refers to the linguistic marking of an utterance as conveying information which is new or unexpected to the speaker” (Delancey (2001: 369-370)). Cruschina’s work has motivated scholars to explore the presence of mirative fronting from a cross-linguistic perspective (e.g., Trotzke (2016) for mirative fronting in German; Authier and Haegeman (2019) for mirative fronting in French; Jiménez-Fernández (2015) and Cruschina (2019) for mirative fronting in Spanish). Thus, the research on non-contrastive focus-fronting is at the frontier of the research on grammatical marking realization of focus.

04.

Xiaotong Ma: Focus is a phenomenon that exists in almost all languages, and the research on focus is very prosperous in many languages around the world. However, the expressions of focus vary from language to language. For example, in English, the special syntactic position in “It is… that…” sentence pattern is used to mark focus. In Chinese, however, pragmatic stress is more used to emphasize the focus. In your opinion, what are similarities and differences between expressions of focus in Chinese and in English?

Prof. Masatoshi Honda: Although I can’t directly answer the question, a brief comparison between phonological stress and the cleft construction will partially answer the question. In the literature on the syntactic and phonological realization of focus, it has been argued that the pragmatic effect of phonological stress should be separated from that of grammatical focus. For example, Horvath (1986) argues that the focus of a sentence must occupy an immediately preverbal position in Hungarian. If a focus constituent (i.e. a phonologically stressed element) occupies a syntactic position other than the preverbal position in the language, the sentence results in ungrammaticality; according to Horvath, such a sentence can be acceptable only if it is interpreted as fulfilling a metalinguistic function (e.g., corrections of a mispronunciation of a certain word by another speaker). Then, É. Kiss (1998: 245) proposes that the preverbal position in Hungarian and the focus position of the cleft construction in English grammatically realize identificational focus, which “represents a subset of the set of contextually or situationally given elements for which the predicate phrase can potentially hold; it is identified as the exhaustive subset of this set for which the predicate phrase actually holds.” Given this research background, we can say that metalinguistic correction is realized by means of prosody and must be separated from the focus which is expressed by grammatical means such as the cleft construction; Cruschina (2011) also argues that metalinguistic correction does not involve focus fronting in Italian.

Therefore, I’m curious about the nature of pragmatic stress and grammatical realization patterns of focus in Chinese; then, a related question is whether there is any correlation between pragmatic stress and grammatically-realized focus in Chinese. The insights obtained from these questions will make an empirical and theoretical contribution to understanding the phonological and grammatical realization of focus in linguistic theory. 

05.

Xiaotong Ma: As we know, you have done many researches not only in Japanese and English but also in German. Could you give us some suggestions if we want to do some syntactic researches in a language that we are not familiar with?

Prof. Masatoshi Honda: If we look at how generative grammar has developed since its earlier stages, we will see that one of its research guidelines is comparative approaches. The main goal of the comparative approach in the generative framework is to provide answers for (at least) the following two traditional core questions: (i) What is knowledge of language? and (ii) how it is acquired? Tackling with these two questions, much research based on the principles and parameters approach has been devoted to identifying a finite set of fundamental principles that are common to all languages and a finite set of parameters that determine syntactic variability among languages (e.g., the pro-drop parameter, which determines whether or not a language allows the subject of a sentence to be omitted.).

In my case, learning from the previous studies based on the (traditional or revised) principles and parameters approach always provide an important way (or an analytical tool) for me to study the target language from a different viewpoint. Fortunately, we have many introductory academic books which recapitulate the research methodology adopted in the principles and parameters approach and its application. I personally think that learning from the development of generative grammar and its insights into knowledge of language is one of the good ways to know how to compare our first language, the target language and a language which we are not familiar with.

往期推荐

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：马晓彤 闫玉萌 陈金玉 王平

排版：马晓彤 闫玉萌 高洁

审校：王丽媛 李芳芳