自從Keyser&Roeper(1984)以來，對中動構式的研究一直是語言學的一個熱點問題，同時也是一個爭議較多的話題。對中動構式的研究一般集中在印歐語言中，而對漢語中動構式的研究是從Sung(1994)才開始的。儘管不少學者已經研究過英語和漢語中動構式，以往文獻中也有關於英漢中動構式的對比研究。然而，中動構式的本質特徵，特別是漢語中動構式的句法和語義等特點，仍然是一個有待解決的問題。因此，我們認為，有必要對英漢中動構式的句法、語義特徵進行全面的探討和對比研究
基本介紹
- 中文名：英漢中動構式的句法語義對比研究
- 類別：論文
- 摘要：對中動構式的研究
- 關鍵字：中動語義、英語中動構式
中文摘要,外文摘要,
中文摘要
自從Keyser&Roeper(1984)以來，對中動構式的研究一直是語言學的一個熱點問題，同時也是一個爭議較多的話題。對中動構式的研究一般集中在印歐語言中，而對漢語中動構式的研究是從Sung(1994)才開始的。儘管不少學者已經研究過英語和漢語中動構式，以往文獻中也有關於英漢中動構式的對比研究。然而，中動構式的本質特徵，特別是漢語中動構式的句法和語義等特點，仍然是一個有待解決的問題。因此，我們認為，有必要對英漢中動構式的句法、語義特徵進行全面的探討和對比研究。
本論文首先討論了中動語義的特徵，包括類指性、情態性、隱性施事的任指性、和...>> 詳細論文不只是討論了中動語義的這些特徵，還探討了它們之間的內部關係，指出中動語義的核心是類指性，具體來說，是屬性(dispositional)類指性。中動句是以非施事為主語的屬性類指句，或稱屬性歸因句。因此，被描述的對象，即事件非施事的參與者，被提升到主語的位置，同時，施事被降級為隱性論元。由於以非施事為主語的屬性類指句具有施事無關性，因此，中動句中的隱性施事在指稱上獲得了任指性解讀。另外，屬性類指句不匯報實際發生的特定事件，它表示的是主語的屬性促使事件發生的可能性，因此，它本身就具有情態性。可見，中動語義的屬性類指性導致了其它特徵的產生，所以，我們稱之為中動語義的核心。
本論文首先討論了中動語義的特徵，包括類指性、情態性、隱性施事的任指性、和...>> 詳細論文不只是討論了中動語義的這些特徵，還探討了它們之間的內部關係，指出中動語義的核心是類指性，具體來說，是屬性(dispositional)類指性。中動句是以非施事為主語的屬性類指句，或稱屬性歸因句。因此，被描述的對象，即事件非施事的參與者，被提升到主語的位置，同時，施事被降級為隱性論元。由於以非施事為主語的屬性類指句具有施事無關性，因此，中動句中的隱性施事在指稱上獲得了任指性解讀。另外，屬性類指句不匯報實際發生的特定事件，它表示的是主語的屬性促使事件發生的可能性，因此，它本身就具有情態性。可見，中動語義的屬性類指性導致了其它特徵的產生，所以，我們稱之為中動語義的核心。
中動語義具有跨語言的一致性，它在不同的語言中以不同的句法形式得以實現。換言之，中動構式在不同的語言中會有不同的表現形式。這也是對比研究的意義所在。英漢兩種語言屬於不同的語系，有著不同的語法特徵，因此，英漢中動構式必然會在形式上有諸多差異。
英語中動構式的存在是一個不爭的事實，而漢語中是否存在中動構式則是一個頗有爭議的問題。因此，論文在討論漢語中動構式之前首先證明了它的存在。論文把前人所研究的“NP+V起來+AP”結構假定為漢語中動構式，從構式語法有關構式的定義出發，證明了“NP+V起來+AP”結構獨立的構式地位，另外，論文把這個結構和類似結構，如主動結構、被動結構、受事主語句、話題句等，進行了詳細的對比，指出，“NP+V起來+AP”結構不等同於其它結構，因此，應當把它看成一個獨立的構式。在此之後，本文把以非施事為主語的“NP+V起來+AP”結構的意義和中動語義的特徵進行了對比，發現，這個結構所表達的意義正是中動語義，而且它在多方面和英語中動構式有諸多相似之處，因此，我們認為“NP+V起來+AP"結構有可能就是漢語中動構式。
在此基礎上，論文對“NP+V起來+AP”結構進行了詳細研究，並按照AP的語義指向將其分成A、B、C三個類型，認為，A型的動詞多為意義虛化的言說或者感官動詞，而B、C兩個類型中的動詞為動作動詞。另外，A型結構中的“V起來”可以移至主語之前，甚至可以從句中刪除，因此，它不是句子的謂語，而是屬於插入語的一種，是“獨立語”成分。A型句子真正的謂語是AP，也就是說，這個結構的主語NP可以直接和AP構成主謂短語。可見，AP本來就是NP的屬性，V所表達的動作只是使這個屬性凸顯的方式。“NP+V起來+AP”結構的B型和C型不同於A型，它們是以“V起來”為謂語中心的，AP只是作為狀語修飾“V起來”。可見，能表達中動語義的不是A型，而是B型和C型。所以，我們認為“NP+V起來+AP”結構的B型和C型才是漢語中動構式。其實，從本質上說，B型和C型應該歸為同一種類型，我們只是為了分析的方便把它們劃分開來。
也有研究者認為，漢語中動構式還有其它的表現形式，如“VP+V上去+AP”、 “VP+V著+AP”、“VP+V來+AP”、以及“NP+AP+v”結構。我們一一分析了這些結構，發現它們並不符合中動語義的所有限制，因此，它們都不是漢語中動構式的類型，至少不是典型的漢語中動構式。所以，我們將漢語中動構式界定為“NP+V起來+AP”結構的B型和C型。
在界定了漢語中動構式之後，我們詳細研究了英漢中動構式的各個組成部分，包括其論元、動詞、和修飾語。關於英漢中動構式的論元，論文討論了位於主語位置上的論元角色和沒有在句法層得到投射的隱性施事。通過對比英漢中動構式位於主語位置上的論元，發現英漢中動構式可以做主語的論元角色是有所不同的，例如英語中動構式可以以感事和來源為主語，而漢語中動構式不可以；另一方面，漢語中動構式的主語可以是成事、當事、目標、時間成分等，而這些成分在英語中動構式中不能做主語。總起來說，漢語中動構式以非內論元做主語的情況多於英語中動構式。
英漢中動構式對動詞的限制既有相同之處，也有相異的地方。一方面，典型英漢中動構式中的動詞在進入中動構式之前，一般具有[+自主]、[+及物]、[+完成]的語義特徵。另一方面，英語中動構式的動詞有時可以帶賓語或者結果補語等其它成分，而漢語中動構式的動詞一般不帶賓語、嚴格排斥結果補語。然而，一般來說，漢語中動構式的能產性比英語高，很多不能進入英語中動構式中的動詞可以構成合格的漢語中動句。
另外，我們對英漢中動構式動詞的及物性分兩個方面進行了探討。首先，我們討論了這些動詞在進入中動構式之前的及物性，認為，典型中動構式的動詞在其它語境中一般是及物動詞；然後，我們分析了這些動詞在中動構式語境裡的及物性，指出，中動詞受到中動構式義的壓制，由及物動詞變成了派生的不及物動詞。
我們知道，英漢中動構式一般需要某種形式的修飾語。英語中動構式的修飾語表現為副詞性修飾，而且在某些情況下，可以通過情態、否定、強調、焦點等手段來實現副詞性修飾的功能。漢語中動構式的修飾語是形容詞性修飾，並且不能用其它手段來代替。也就是說，漢語中動構式里的形容詞性修飾是不可缺少的。因而，所有的漢語中動句都有一定的評價意義，表達的是說話者對主語屬性的主觀評價。儘管英漢中動構式的修飾語表現為不同的形式，然而，這一點不能構成英漢中動構式的區別，因為，這些修飾語在英漢中動構式中的功能是一致的，即，作為狀語修飾謂語動詞。至於以何種形式作狀語和不同語言的特徵相關。漢語中可以作狀語的形式是多樣的，而英語則一般以副詞性成分作狀語。
還需要探討的另外一個重要問題就是英漢中動構式為什麼需要修飾語。這個問題有三種可能的解決方案，即，中動構式里的修飾語是句法要求、語義要求、或是語用需要。我們認為英漢中動構式里的修飾語是語用需要。詳細來說，修飾語是每個話語都必須有一個信息焦點的需要。換言之，修飾語作為中動句的信息中心為話語提供新信息。
在討論了英漢中動構式的各組成成分之後，我們又從整體上討論英漢中動構式範疇。論文首先總結了英漢中動構式的典型性特徵，並指出符合所有這些特徵的句子屬於典型英漢中動構式，或稱英漢中動構式的原型；而在某方面違反這些特徵的句子則是英漢中動構式的非典型成員。我們按照典型和非典型把英漢中動構式分成了不同的意義類型，並進行了詳細的對比。發現，漢語中動構式的意義類型不如英語中動構式豐富，如漢語中動構式沒有“性質中動構式”和“設計特徵中動構式”。這主要是因為漢語中動構式一般不能客觀描述某實體的特徵，一般都有評價意義。可見，有時，漢語中動構式中所表現出來的主觀性要比英語中動構式強。
英漢中動構式範疇的成員有典型和非典型之分，而且，其邊界也不是清晰可斷的。換言之，英漢中動構式範疇和其它類似結構範疇之間不是界限分明的。這也證明了認知語言學關於範疇邊界模糊性的說法。
關鍵字：中動語義；英語中動構式；漢語中動構式；論元；動詞；修飾語
英語中動構式的存在是一個不爭的事實，而漢語中是否存在中動構式則是一個頗有爭議的問題。因此，論文在討論漢語中動構式之前首先證明了它的存在。論文把前人所研究的“NP+V起來+AP”結構假定為漢語中動構式，從構式語法有關構式的定義出發，證明了“NP+V起來+AP”結構獨立的構式地位，另外，論文把這個結構和類似結構，如主動結構、被動結構、受事主語句、話題句等，進行了詳細的對比，指出，“NP+V起來+AP”結構不等同於其它結構，因此，應當把它看成一個獨立的構式。在此之後，本文把以非施事為主語的“NP+V起來+AP”結構的意義和中動語義的特徵進行了對比，發現，這個結構所表達的意義正是中動語義，而且它在多方面和英語中動構式有諸多相似之處，因此，我們認為“NP+V起來+AP"結構有可能就是漢語中動構式。
在此基礎上，論文對“NP+V起來+AP”結構進行了詳細研究，並按照AP的語義指向將其分成A、B、C三個類型，認為，A型的動詞多為意義虛化的言說或者感官動詞，而B、C兩個類型中的動詞為動作動詞。另外，A型結構中的“V起來”可以移至主語之前，甚至可以從句中刪除，因此，它不是句子的謂語，而是屬於插入語的一種，是“獨立語”成分。A型句子真正的謂語是AP，也就是說，這個結構的主語NP可以直接和AP構成主謂短語。可見，AP本來就是NP的屬性，V所表達的動作只是使這個屬性凸顯的方式。“NP+V起來+AP”結構的B型和C型不同於A型，它們是以“V起來”為謂語中心的，AP只是作為狀語修飾“V起來”。可見，能表達中動語義的不是A型，而是B型和C型。所以，我們認為“NP+V起來+AP”結構的B型和C型才是漢語中動構式。其實，從本質上說，B型和C型應該歸為同一種類型，我們只是為了分析的方便把它們劃分開來。
也有研究者認為，漢語中動構式還有其它的表現形式，如“VP+V上去+AP”、 “VP+V著+AP”、“VP+V來+AP”、以及“NP+AP+v”結構。我們一一分析了這些結構，發現它們並不符合中動語義的所有限制，因此，它們都不是漢語中動構式的類型，至少不是典型的漢語中動構式。所以，我們將漢語中動構式界定為“NP+V起來+AP”結構的B型和C型。
在界定了漢語中動構式之後，我們詳細研究了英漢中動構式的各個組成部分，包括其論元、動詞、和修飾語。關於英漢中動構式的論元，論文討論了位於主語位置上的論元角色和沒有在句法層得到投射的隱性施事。通過對比英漢中動構式位於主語位置上的論元，發現英漢中動構式可以做主語的論元角色是有所不同的，例如英語中動構式可以以感事和來源為主語，而漢語中動構式不可以；另一方面，漢語中動構式的主語可以是成事、當事、目標、時間成分等，而這些成分在英語中動構式中不能做主語。總起來說，漢語中動構式以非內論元做主語的情況多於英語中動構式。
英漢中動構式對動詞的限制既有相同之處，也有相異的地方。一方面，典型英漢中動構式中的動詞在進入中動構式之前，一般具有[+自主]、[+及物]、[+完成]的語義特徵。另一方面，英語中動構式的動詞有時可以帶賓語或者結果補語等其它成分，而漢語中動構式的動詞一般不帶賓語、嚴格排斥結果補語。然而，一般來說，漢語中動構式的能產性比英語高，很多不能進入英語中動構式中的動詞可以構成合格的漢語中動句。
另外，我們對英漢中動構式動詞的及物性分兩個方面進行了探討。首先，我們討論了這些動詞在進入中動構式之前的及物性，認為，典型中動構式的動詞在其它語境中一般是及物動詞；然後，我們分析了這些動詞在中動構式語境裡的及物性，指出，中動詞受到中動構式義的壓制，由及物動詞變成了派生的不及物動詞。
我們知道，英漢中動構式一般需要某種形式的修飾語。英語中動構式的修飾語表現為副詞性修飾，而且在某些情況下，可以通過情態、否定、強調、焦點等手段來實現副詞性修飾的功能。漢語中動構式的修飾語是形容詞性修飾，並且不能用其它手段來代替。也就是說，漢語中動構式里的形容詞性修飾是不可缺少的。因而，所有的漢語中動句都有一定的評價意義，表達的是說話者對主語屬性的主觀評價。儘管英漢中動構式的修飾語表現為不同的形式，然而，這一點不能構成英漢中動構式的區別，因為，這些修飾語在英漢中動構式中的功能是一致的，即，作為狀語修飾謂語動詞。至於以何種形式作狀語和不同語言的特徵相關。漢語中可以作狀語的形式是多樣的，而英語則一般以副詞性成分作狀語。
還需要探討的另外一個重要問題就是英漢中動構式為什麼需要修飾語。這個問題有三種可能的解決方案，即，中動構式里的修飾語是句法要求、語義要求、或是語用需要。我們認為英漢中動構式里的修飾語是語用需要。詳細來說，修飾語是每個話語都必須有一個信息焦點的需要。換言之，修飾語作為中動句的信息中心為話語提供新信息。
在討論了英漢中動構式的各組成成分之後，我們又從整體上討論英漢中動構式範疇。論文首先總結了英漢中動構式的典型性特徵，並指出符合所有這些特徵的句子屬於典型英漢中動構式，或稱英漢中動構式的原型；而在某方面違反這些特徵的句子則是英漢中動構式的非典型成員。我們按照典型和非典型把英漢中動構式分成了不同的意義類型，並進行了詳細的對比。發現，漢語中動構式的意義類型不如英語中動構式豐富，如漢語中動構式沒有“性質中動構式”和“設計特徵中動構式”。這主要是因為漢語中動構式一般不能客觀描述某實體的特徵，一般都有評價意義。可見，有時，漢語中動構式中所表現出來的主觀性要比英語中動構式強。
英漢中動構式範疇的成員有典型和非典型之分，而且，其邊界也不是清晰可斷的。換言之，英漢中動構式範疇和其它類似結構範疇之間不是界限分明的。這也證明了認知語言學關於範疇邊界模糊性的說法。
關鍵字：中動語義；英語中動構式；漢語中動構式；論元；動詞；修飾語
外文摘要
Beginning from Keyser& Roeper (1984), the middle construction has been a hot topic in linguistics. Studies of the middle construction are basically about that of Indo-European languages, while studies of Chinese middle constructions start from Sung (1994). Although many researches have been done on English and Chinese middle constructions, the nature of middle constructions, especially the syntactic and semantic features of Chinese middle constructions, remains a mystery. Therefore, it is nece...>> 詳細
Beginning from Keyser& Roeper (1984), the middle construction has been a hot topic in linguistics. Studies of the middle construction are basically about that of Indo-European languages, while studies of Chinese middle constructions start from Sung (1994). Although many researches have been done on English and Chinese middle constructions, the nature of middle constructions, especially the syntactic and semantic features of Chinese middle constructions, remains a mystery. Therefore, it is necessary to conduct a thorough and in-depth research on the syntax and semantics of English and Chinese middle constructions, and a contrastive study will be helpful.
In the beginning, this dissertation talks about the features of middle semantics, including its genericity, modality, the arbitrary reference of the implicit agent, and the responsibility of the grammatical subject. In the meantime, the dissertation also explores the interrelations among those features, and finds that the core of middle semantics is genericity, dispositional genericity, to be specific. In other words, middle constructions are disposition ascriptions. As a result, the described passive participant is promoted to the subject position, while the agent is demoted. Since dispositionals are subject-oriented generic sentences, the implicit agent acquires an arbitrary interpretation. Moreover, instead of reporting real happenings, dispositional generics convey the conduciveness of the subject entity to the action designated by the predicate verb. In other words, as disposition ascriptions, middle constructions inherently embrace a modal meaning. Thus, the feature of dispositional genericity, as the core of middle semantics, leads to all the other features of middle semantics.
Middle semantics is universal across languages, but it is realized by different syntactic structures in different languages. That is to say, middle constructions are encoded in different ways in different languages. As English and Chinese belong to different language families, English and Chinese middle constructions are bound to have many differences in their syntactic representations.
The existence of English middle constructions does not arouse much opposition, but linguists disagree when it comes to whether there are middle constructions in Chinese. Due to that, we first of all try to prove the existence of Chinese middle constructions. Based on previous researches, we presuppose "NP+ Wqilai+AP" as the Chinese middle construction. Before all, we prove its status as an individual construction, and then we compare that construction with other similar constructions, including active constructions, passive constructions, patient-subject constructions, and topic constructions. We come to the conclusion that "NP+ Vqilai+AP" cannot be taken as a part of other constructions and its meaning is independent of other constructions. Then by comparing a subsection of "NP+ Vqilai+AP", i.e, the part with a non-agentive subject, with middle semantics, we find that the meaning of "NP+ Vqilai+AP" is consistent with middle semantics. Moreover, we also compare "NP+ Vqilai+AP" with the English middle construction, and find many similarities between the two constructions. Therefore, we conclude that there is a chance that "NP+ ''Vqilai+AP" is the Chinese middle construction.
Based on that hypothesis, we divide "NP+ Vqilai+AP" into three types, i.e, types A, B, and C, according to the different semantic orientations of AP. We find that type A is very different from types B and C in many ways, while types B and C are essentially the same. After comparing the three types with middle semantics, we come to the conclusion that types B and C of "NP+ Vqilai+AP" are the Chinese middle construction. Some researchers take structures like "VP+ Vshangqu+AP", "VP+ Vzhe+AP", "VP+ Vlai+AP", and "NP+AP+V" as other realizations of middle semantics in Chinese. After analyzing those structures one by one, we find that they do not conform to the restrictions of middle constructions. Therefore, we define Chinese middle constructions as types B and C of "NP+ Vqilai+AP".
After the delimitation of Chinese middle constructions, we go on to talk about the component parts of English and Chinese middle constructions, including their arguments, verbs, and adjuncts. As for their arguments, we discuss both the argument in the subject position and the implicit agent which is not syntactically present. By comparing the subjects of English and Chinese middle constructions, we find that English and Chinese middle constructions allow different thematic roles in the subject position. Generally speaking, Chinese middle constructions allow more cases of non-intemal-argument subject than their English counterparts.
The restrictions English and Chinese middle constructions impose on their predicate verbs have something in common. For example, in other contexts rather than the middle construction, those verbs share semantic features like [+volitional], [+transitive], and [+perfective]. On the other hand, however, the restrictions on middle verbs between English and Chinese are also different from each other. For instance, English middle verbs sometimes allow the presence of objects or resultative complements. Nevertheless, Chinese middle verbs rarely come with objects and they are incompatible with resultative complements. Generally speaking, Chinese middle constructions have higher productivity than English ones, and many verbs that are not eligible in English middle constructions form perfectly acceptable Chinese middle constructions.
Apart from that, we talk about the transitivity of English and Chinese middle verbs from two perspectives. In the first place, we discuss the transitive nature of those verbs in other environments rather than the middle construction. Then, we analyze their transitivity in the context of middles, and conclude that middle verbs become derived intransitives as a result of their interaction with the construction meaning.
As we know, both English and Chinese middles need some kind of adjuncts. English middles require adverbial modification, which can also be replaced by negation, stress, focus and so on. Chinese middles require adjective modification, and there are no ways to substitute this kind of adjuncts. Therefore, all Chinese middle sentences have appraisal meaning, i.e. they express speakers'' subjective evaluation of the properties of the subject entities. Although the adjuncts in English and Chinese middle constructions are very different from each other, this does not constitute a distinction between English and Chinese middles, since the adjuncts share the same function in both constructions. This superficial difference is due to the different characteristics of English and Chinese. To be specific, Chinese verbs allow different forms of modifiers, while verbs in English are usually modified by adverbs.
Besides, we also have to understand why English and Chinese middles require adjuncts. There are three possible answers to that question: it can be structural, semantic, or pragmatic. We find that adjuncts in middles are due to pragmatic constraints. That is to say, adjuncts are used to satisfy the requirement that every utterance have a focus that serves to convey new information in the discourse.
After an in-depth study of the components of English and Chinese middles, we then talk about English and Chinese middles as a category. We first of all summarize the defining features of English and Chinese middles, and point out that sentences conforming to all those features are prototypes in this category, while sentences violating one or two of those features are peripheral members of the category. Afterwards, we divide English and Chinese middles into different types. Through a careful comparison, we find that English middle constructions convey more various meanings than Chinese middles and that all Chinese middles contain a shade of subjective meaning, which is not the case in English middles. Moreover, the boundary between middles and other similar constructions is not clear-cut. Therefore, to some extent, this study also confirms the fuzzy nature of boundaries between categories.
Key words: middle semantics; English middle constructions; Chinese middle constructions; arguments; middle verbs; adjuncts
In the beginning, this dissertation talks about the features of middle semantics, including its genericity, modality, the arbitrary reference of the implicit agent, and the responsibility of the grammatical subject. In the meantime, the dissertation also explores the interrelations among those features, and finds that the core of middle semantics is genericity, dispositional genericity, to be specific. In other words, middle constructions are disposition ascriptions. As a result, the described passive participant is promoted to the subject position, while the agent is demoted. Since dispositionals are subject-oriented generic sentences, the implicit agent acquires an arbitrary interpretation. Moreover, instead of reporting real happenings, dispositional generics convey the conduciveness of the subject entity to the action designated by the predicate verb. In other words, as disposition ascriptions, middle constructions inherently embrace a modal meaning. Thus, the feature of dispositional genericity, as the core of middle semantics, leads to all the other features of middle semantics.
Middle semantics is universal across languages, but it is realized by different syntactic structures in different languages. That is to say, middle constructions are encoded in different ways in different languages. As English and Chinese belong to different language families, English and Chinese middle constructions are bound to have many differences in their syntactic representations.
The existence of English middle constructions does not arouse much opposition, but linguists disagree when it comes to whether there are middle constructions in Chinese. Due to that, we first of all try to prove the existence of Chinese middle constructions. Based on previous researches, we presuppose "NP+ Wqilai+AP" as the Chinese middle construction. Before all, we prove its status as an individual construction, and then we compare that construction with other similar constructions, including active constructions, passive constructions, patient-subject constructions, and topic constructions. We come to the conclusion that "NP+ Vqilai+AP" cannot be taken as a part of other constructions and its meaning is independent of other constructions. Then by comparing a subsection of "NP+ Vqilai+AP", i.e, the part with a non-agentive subject, with middle semantics, we find that the meaning of "NP+ Vqilai+AP" is consistent with middle semantics. Moreover, we also compare "NP+ Vqilai+AP" with the English middle construction, and find many similarities between the two constructions. Therefore, we conclude that there is a chance that "NP+ ''Vqilai+AP" is the Chinese middle construction.
Based on that hypothesis, we divide "NP+ Vqilai+AP" into three types, i.e, types A, B, and C, according to the different semantic orientations of AP. We find that type A is very different from types B and C in many ways, while types B and C are essentially the same. After comparing the three types with middle semantics, we come to the conclusion that types B and C of "NP+ Vqilai+AP" are the Chinese middle construction. Some researchers take structures like "VP+ Vshangqu+AP", "VP+ Vzhe+AP", "VP+ Vlai+AP", and "NP+AP+V" as other realizations of middle semantics in Chinese. After analyzing those structures one by one, we find that they do not conform to the restrictions of middle constructions. Therefore, we define Chinese middle constructions as types B and C of "NP+ Vqilai+AP".
After the delimitation of Chinese middle constructions, we go on to talk about the component parts of English and Chinese middle constructions, including their arguments, verbs, and adjuncts. As for their arguments, we discuss both the argument in the subject position and the implicit agent which is not syntactically present. By comparing the subjects of English and Chinese middle constructions, we find that English and Chinese middle constructions allow different thematic roles in the subject position. Generally speaking, Chinese middle constructions allow more cases of non-intemal-argument subject than their English counterparts.
The restrictions English and Chinese middle constructions impose on their predicate verbs have something in common. For example, in other contexts rather than the middle construction, those verbs share semantic features like [+volitional], [+transitive], and [+perfective]. On the other hand, however, the restrictions on middle verbs between English and Chinese are also different from each other. For instance, English middle verbs sometimes allow the presence of objects or resultative complements. Nevertheless, Chinese middle verbs rarely come with objects and they are incompatible with resultative complements. Generally speaking, Chinese middle constructions have higher productivity than English ones, and many verbs that are not eligible in English middle constructions form perfectly acceptable Chinese middle constructions.
Apart from that, we talk about the transitivity of English and Chinese middle verbs from two perspectives. In the first place, we discuss the transitive nature of those verbs in other environments rather than the middle construction. Then, we analyze their transitivity in the context of middles, and conclude that middle verbs become derived intransitives as a result of their interaction with the construction meaning.
As we know, both English and Chinese middles need some kind of adjuncts. English middles require adverbial modification, which can also be replaced by negation, stress, focus and so on. Chinese middles require adjective modification, and there are no ways to substitute this kind of adjuncts. Therefore, all Chinese middle sentences have appraisal meaning, i.e. they express speakers'' subjective evaluation of the properties of the subject entities. Although the adjuncts in English and Chinese middle constructions are very different from each other, this does not constitute a distinction between English and Chinese middles, since the adjuncts share the same function in both constructions. This superficial difference is due to the different characteristics of English and Chinese. To be specific, Chinese verbs allow different forms of modifiers, while verbs in English are usually modified by adverbs.
Besides, we also have to understand why English and Chinese middles require adjuncts. There are three possible answers to that question: it can be structural, semantic, or pragmatic. We find that adjuncts in middles are due to pragmatic constraints. That is to say, adjuncts are used to satisfy the requirement that every utterance have a focus that serves to convey new information in the discourse.
After an in-depth study of the components of English and Chinese middles, we then talk about English and Chinese middles as a category. We first of all summarize the defining features of English and Chinese middles, and point out that sentences conforming to all those features are prototypes in this category, while sentences violating one or two of those features are peripheral members of the category. Afterwards, we divide English and Chinese middles into different types. Through a careful comparison, we find that English middle constructions convey more various meanings than Chinese middles and that all Chinese middles contain a shade of subjective meaning, which is not the case in English middles. Moreover, the boundary between middles and other similar constructions is not clear-cut. Therefore, to some extent, this study also confirms the fuzzy nature of boundaries between categories.
Key words: middle semantics; English middle constructions; Chinese middle constructions; arguments; middle verbs; adjuncts
