{ "paper_id": "Y02-1034", "header": { "generated_with": "S2ORC 1.0.0", "date_generated": "2023-01-19T13:43:37.990798Z" }, "title": "Double Subject, Double Nominative Object and Double Accusative Object Constructions in Japanese and Korean", "authors": [ { "first": "Hiroaki", "middle": [], "last": "Nakamura", "suffix": "", "affiliation": {}, "email": "nakamura@jcga.ac.jp" } ], "year": "", "venue": null, "identifiers": {}, "abstract": "This paper presents a unified account of three kinds of constructions in which more than one NP can show up with the same case in simple sentences in Japanese and Korean: double subject, double nominative object and double accusative constructions. Noting that the second NPs in these constructions are functional or relational, this paper proposes to assign them the category and type different from the first NPs. We show the derivations of these three constructions in a parallel manner, and explain the asymmetries in extractability between possessor and possessed NPs in relativization.", "pdf_parse": { "paper_id": "Y02-1034", "_pdf_hash": "", "abstract": [ { "text": "This paper presents a unified account of three kinds of constructions in which more than one NP can show up with the same case in simple sentences in Japanese and Korean: double subject, double nominative object and double accusative constructions. Noting that the second NPs in these constructions are functional or relational, this paper proposes to assign them the category and type different from the first NPs. We show the derivations of these three constructions in a parallel manner, and explain the asymmetries in extractability between possessor and possessed NPs in relativization.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Abstract", "sec_num": null } ], "body_text": [ { "text": "Much attention has centered on the so-called double (multiple) subject construction (hereafter, DSC) in Japanese and Korean,' as illustrated in (1) and (2), in the literature of empirical-and theoretical linguistics:", "cite_spans": [ { "start": 84, "end": 100, "text": "(hereafter, DSC)", "ref_id": null } ], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1" }, { "text": "(1) a. Syusyoo-no byooki-ga saikin omo-i. Prime Minister-GEN illness-NOM recently serious-PRES 'The Prime Minister is seriously ill.' b. Syusyoo-ga saikin byooki-ga omo-i. Prime Minister-NOM recently illness-NOM serious-PRES (2) a. Swunsang-uy pyeng-i choykun simha-ta. The Prime Minister-Gen illness-NOM recently serious-PRES b. Swunsang-i choykun peyng-i simha-ta. The Prime Minister-Nom recently illness-NOM serious-PRES In the (b) examples of (1) and (2), we insert the adverb saikin and choykun 'recently' between the two nominative NPs to show that these two nominative NPs do not form a constituent (if the 'possessor' NP shows up with genitive case, adverbs may not intervene between these NPs, as in *syusyoo-no saikin byoojoo 'lit. *Prime Minister's recently condition'). In the generative grammar framework, it has been argued that the sentence-initial nominative NP (hereafter, we refer to it as 'major subject') raises to a higher position (the targets vary depending on the movement theories the authors adopt) from the possessor position of the following subject by NP-movement in DSC. For example, Hasegawa (1999) proposes the derivation like (3):", "cite_spans": [ { "start": 1114, "end": 1129, "text": "Hasegawa (1999)", "ref_id": "BIBREF22" } ], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1" }, { "text": "( DSC has an IP-adjunction structure generated by NP-movement in (3) 2 . Notice here that the movement in (3) violates the limitation of 'one case per chain' since it involves two case (i.e., genitive and nominative case) positions in the chain formed by the movement of the possessor. Besides, this kind of explanation has ignored a lot of important characteristics of this construction, concentrating only on the multiple assignment/licensing of nominative case to the major and regular subjects. On the other hand, some Japanese traditional grammarians have proposed a kind of complex structure for DSC. (4) is cited from Sugimoto (1986) : Yamada-san-ga [si okusan-ga kirei-da]] Yamada-NOM wife-NOM beautiful-PRES 'Mr. Yamada's wife is beautiful.'", "cite_spans": [ { "start": 625, "end": 640, "text": "Sugimoto (1986)", "ref_id": null }, { "start": 643, "end": 656, "text": "Yamada-san-ga", "ref_id": null } ], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1" }, { "text": "(4) [s2", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1" }, { "text": "Sugimoto assumes that DSC has a complement sentence structure, in which the S1 is embedded under the matrix sentence S2. This argumentation immediately conflicts with the standard notion of a 'clausal complement', which is assumed to be a clause subcategorized for by a higher predicate. In (4), there is no matrix predicate which would take the clausal complement S1 as argument. He also dubs S1 BUN-JUTUGO 'Sentence-Predicate', but he does not give any explanation for the reason WHY this sentence, in which the argument structure of the predicate kirei-da 'is beautiful' is fully satisfied, can be predicated of the major subject. I believe that the availability of layers of predication in DSC like (4) is the most important feature of this construction which requires an explanation.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1" }, { "text": "Japanese and Korean also have interesting constructions related to DSC. Consider sentence (5b), where the object and its possessor shows up with nominative case:3 (5) a. Hanako-ni(wa) kono hon-no naiyoo-ga yoku wakar-u. Hanako-DAT(TOP) this book-GEN content-NOM well understand-PRES 'Hanako understands the content of this book well.' b. Hanako-ni(wa) kono hon-ga yoku naiyoo-ga wakar-u. Hanako-DAT(TOP) this book-NOM well content-NOM understand-PRES.", "cite_spans": [ { "start": 170, "end": 183, "text": "Hanako-ni(wa)", "ref_id": null }, { "start": 338, "end": 351, "text": "Hanako-ni(wa)", "ref_id": null } ], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1" }, { "text": "It is well-known that both the possessor and possessed NP's can be marked with accusative Case in 2 hl fact, some generative grammarians do not admit the IP-adjunction structure like (3) and suggest that a major subject move to a specifier position of some functional category. Mihara (1990) and Takezawa (2000) , while assuming that DSC have the IP-adjunction structure, argue that a major subject is not raised to the IP adjoined position but base-generated there. 3 Note here that, in the sentence in (5), the nominative NP kono hon-ga 'this book' is NOT a subject as can be seen from the pair in (i):", "cite_spans": [ { "start": 278, "end": 291, "text": "Mihara (1990)", "ref_id": "BIBREF13" }, { "start": 296, "end": 311, "text": "Takezawa (2000)", "ref_id": null } ], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1" }, { "text": "(i) a. Taroo-ni jibuni-no koto-ga wakar-u. Taroo-DAT SELF-GEN thing-NOM understand-PRES Taro() understands himself.' b. * Jibun-ni", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1" }, { "text": "Taroo-no koto-ga wakar-u. SELF-DAT Taroo-GEN thing-NOM understand-PRES In (ia), the nominative jibun 'self can be corefer with the dative NP Taroo, whereas in (ib), Taroo cannot be the antecedent of the reflexive zibun (Notice that the word order is irrelevant, as can be seen in Jibun no koto-ga Tarooni wakaru vs *Taroo-no koto-ga jibun-ni wakar-u) . Also, the dative NP, but not the nominative objsect, induces subject-honorification, a kind of subject-agreement in Japanese, as illustrated in (ii):", "cite_spans": [ { "start": 297, "end": 350, "text": "Tarooni wakaru vs *Taroo-no koto-ga jibun-ni wakar-u)", "ref_id": null } ], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1" }, { "text": "(ii) Senseirni(wa)", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1" }, { "text": "imadoki-no seitoi-ga o-wakari-ninaraipi-nai. Teacher-DAT(-TOP) these-days-GEN students-NOM HON-understand-HON-NEG The teacher doesn't understand students of today.' All the tests for subjecthood clearly show that dative (experiencer) NPs are subjects, whereas nominative objects have much in common with normal accusative objects in transitive sentences.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1" }, { "text": "Korean transitive sentences (which are called the Inalienable Possessor Construction or Multiple Accusative Constructions) and these accusative NP's must stand in an inalienable possession relation. Observe the pair in (6) (taken from Cho 1993):", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1" }, { "text": "(6) a. Mary-ka John-uy tali-lul Mary-NOM John-GEN leg-ACC 'Mary kicked John's leg.' b. Mary-ka John-ul tali-lul Mary-NOM John-ACC leg-ACC cha-ss-ta. kick-PAST-DEC cha-ess-ta. kick-PAST-DEC These three kinds of constructions, i.e., DSC, the double (multiple) nominative object construction (hereafter, DNOC) and double (multiple) accusative object construction (hereafter, DAC), surely seem to share some interesting properties: in a simple sentence with only one predicate, more than one NP show up with the identical case. They must stand in a close relationship and the preceding (possessor) NP is NOT an argument of a predicate, but still assigned structural case.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1" }, { "text": "We will address to another similarity these three constructions have in common: possesee NPs resist extraction in relative clauses while possessor NPs can be the heads of relative clauses:", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1" }, { "text": "(7) a. byoojoo-ga omoi syusyoo illness-NOM serious prime minister 'the prime minister who is seriously ill' b. * syusyoo-ga omoi byoojoo the prime minister-NOM serious illness (8) a. Hanako-ni naiyoo-ga wakar-u hon Hanako-DAT content-NOM understand-PRES book 'the book the content of which Hanako understands' b. * Hanako-ni kono hon-ga wakar-u naiyoo Hanako-DAT this book-NOM understand content (9) a. Thalo-ka meli-lul simhakey ttayli-n Hanakho Thalo-Nom head-Acc hard hit-PAST Hanakho 'Hanakho whose head Thalo hit hard' b. * Thalo-ka Hanakho-lul simhakey ttayli-n meli Thalo-Nom Hanakho-Acc hard hit-PAST head", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1" }, { "text": "In this paper, we will argue that these three kinds of constructions (DSC, DNOC and DAC) in which more than one NPs show up with identical cases must be dealt with in a unified manner. The remainder of this article breaks down as follows. Section 2 outlines the framework we will adopt in this paper, the Combinatory Categorial Grammar proposed by Steedman (1996 Steedman ( , 2000 , and then shows the derivation of DSC in Japanese and Korean. Section 3 deals with DNOC and DAC, showing that these three kinds of constructions can and must be derived in a unified manner. Section 4 discusses consequences of the derivations of these constructions and the constraint on relative clause formation of DSC, DNOC and DAC. Section 5 presents conclusions.", "cite_spans": [ { "start": 348, "end": 362, "text": "Steedman (1996", "ref_id": "BIBREF16" }, { "start": 363, "end": 380, "text": "Steedman ( , 2000", "ref_id": "BIBREF17" } ], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1" }, { "text": "As a descriptive framework, we adopt the Combinatory Categorial Grammar proposed by Steedman (1996 Steedman ( , 2000 . CCG analyzes only surface strings of natural language, avoiding descriptive devices such as movement or deletion rules, abstract levels of representation, and empty categories. Combinatory rules may only apply to entities which are linguistically realized and adjacent, building up expressions from words to larger expressions. The modes of combination of expressions are entirely determined by lexical syntactic types, which specify semantic valency and canonical constituent order, and nothing else. For example, the verb eat in English has the category assignment in (1):", "cite_spans": [ { "start": 84, "end": 98, "text": "Steedman (1996", "ref_id": "BIBREF16" }, { "start": 99, "end": 116, "text": "Steedman ( , 2000", "ref_id": "BIBREF17" } ], "ref_spans": [], "eq_spans": [], "section": "Combinatory Categorial Grammar and Derivations of Double Subject Construction", "sec_num": "2" }, { "text": "(10) eats := (S/LNP)/NP A forward slash / means that an argument of the appropriate type must appear to the right of the functor; a backward slash, notated as '/L' in this paper, means that the argument must appear to the left. The category of eats indicates that it is a function that seeks an NP argument (i.e., an object) on its right to form a verb phrase of category S/ LNP, and then seeks an NP argument (i.e., a subject) on its left to form an S. I introduce only three combinatory rules in (11) relevant for the present discussions.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Combinatory Categorial Grammar and Derivations of Double Subject Construction", "sec_num": "2" }, { "text": "(11) a. X/Y:f Y:", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Combinatory Categorial Grammar and Derivations of Double Subject Construction", "sec_num": "2" }, { "text": "a X:fa Y: a X/LY:f X:fa b. X/Y:g Y/Z:f X/Z:gf Y/LZ:f X/LY:g X/LZ:gf c. X:a T/L(T/X) or T/(T/LX): Alfa (11a)", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Combinatory Categorial Grammar and Derivations of Double Subject Construction", "sec_num": "2" }, { "text": "is the rule of function application. A function of category X/Y combines with an adjacent argument of category Y to yield a result of category X and interpretation fa, the result of applying f to a. This rule, for example, combines a transitive verb with an object to yield the verb phrase, and then, combines the verb phrase with a subject to produce the. sentence. The rule of function composition (11b) allows the main function of category X/Y to combine with the subordinate function of category Y/Z to yield a function of category X/Z. (11c) is the rule of type-raising, which is the process of reanalyzing an argument category as a new function which takes as its argument the function that would have applied to it before type raising. For example, this operation converts a subject NP, which would normally be an argument to a verb phrase of category NP/LS, into a function looking forward for a verb phrase on the right, S/(S/ LNP). We will see how these combinatory rules work to derive the surface constituents and their interpretations immediately below. Observing a close parallel traditionally maintained between syntax and semantics, we spell out semantic interpretations using italics following a colon, as in eat := (S/LNP)/NP: Ax.ily.eat'(x)(y). Assuming this simple apparatus, let us turn to the analysis of double subject sentences. As mentioned before, the most important property of these construction is that a simple sentence involves the layers of predication in that for each subject, the rest of the sentence denotes the property of its referent. To derive the structure involving the layers of predication in a sentence, we must address the syntactic and semantic properties of the second subject. For concreteness' sake, observe the contrast in grammaticality in 12 Even if Hanako in (12b) is Tanaka's wife, this sentence is completely unacceptable. Namely, in DSC; the second subject must be semantically (referentially) dependent upon the preceding (major) subject. The degraded status of (13b) seems to be due to the difficulty in getting a relational reading with kabu 'stock' (note that it is grammatical when the possessor Yamaguti-san is marked with genitive case as in Yamaguti-san no kabu, whereas (13b) is good because motikabu 'stock holdings' always requires to be construed relationally. In other words, the second subject in DSC does not denote an independent individual or individuals by itself, but denotes a part of the individual(s) denoted by the first subject, or some independent individual(s) standing in a certain close relations with its referent(s). In fact, some double subject sentences cannot appear without the major subjects, as illustrated in (14):(14) a. * Se-ga taka-i. height-NOM tall-PRES. b.?? Okusan-ga bijin-des-u.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Combinatory Categorial Grammar and Derivations of Double Subject Construction", "sec_num": "2" }, { "text": "wife-NOM beautiful woman-BE-PRES Though (14b) is not ungrammatical, it cannot denote truth values because it must be specified as to whose wife okusan-ga refers to in the context. The sentences in (14) should be taken to be open propositions denoting properties of some persons. The nouns like okusan 'wife' are argued to involve 'variables' in them as in 'wife-of-x' (Engdahl 1988 , Jacobsen 1999 , Baldridge 2000 , Kruijff and Vasishth 2001 . Many authors refer to this kind of nouns as 'relational nouns.' Let us assume that relational nouns have the category NP/LNP, which look for possessors on the left and returns a concatenated NP, and the type , i.e., functions from individuals to individuals (see Jacobson 1999 Jacobson , 2000 . We indicate the morphological case and semantic role on categories by subscripts for expository convenience, as in 15:", "cite_spans": [ { "start": 368, "end": 381, "text": "(Engdahl 1988", "ref_id": null }, { "start": 382, "end": 397, "text": ", Jacobsen 1999", "ref_id": null }, { "start": 398, "end": 414, "text": ", Baldridge 2000", "ref_id": "BIBREF0" }, { "start": 415, "end": 442, "text": ", Kruijff and Vasishth 2001", "ref_id": "BIBREF9" }, { "start": 713, "end": 726, "text": "Jacobson 1999", "ref_id": "BIBREF6" }, { "start": 727, "end": 742, "text": "Jacobson , 2000", "ref_id": "BIBREF7" } ], "ref_spans": [], "eq_spans": [], "section": "Combinatory Categorial Grammar and Derivations of Double Subject Construction", "sec_num": "2" }, { "text": "EQUATION", "cite_spans": [], "ref_spans": [], "eq_spans": [ { "start": 0, "end": 8, "text": "EQUATION", "ref_id": "EQREF", "raw_str": "(15) okusan := NPNomILNPuen.pos:Ax.wifei-oi(x)", "eq_num": "(15)" } ], "section": "Combinatory Categorial Grammar and Derivations of Double Subject Construction", "sec_num": "2" }, { "text": "states that okusan 'wife' is a functional expression looking leftward for a possessor marked with genitive case to return 'his wife' as a derived expression. Given this category and type for relational nouns, we can show the derivation of sentence (la) with the possessor marked with genitive case: The derivation in (16) requires no explanation. Syusyo-no 'prime minister' is concatenated to byojo-ga 'illness' by usual function application to yield the subject NP. Before proceeding to the derivation of the double subject sentences, let us assume the very simple definition of subject and object following Dowty (1997:7) . Since CCG allows flexible constituency, the relational NP byojo-ga 'illness-of-x' can combines with the adjectival predicate omoi 'serious' first via function composition, yielding the derived predicate byojo-ga omoi 'is seriously ill'. However, this complex predicate cannot be concatenated to the nominative NP syusyoo-ga 'prime minister' because it is an expression of category S/ NP L-Gen.Pos which look leftward for a possessor marked with GENITIVE case. When the major subject syushoo-ga combines with this derived predicate, case conflict immediately arises.", "cite_spans": [ { "start": 609, "end": 623, "text": "Dowty (1997:7)", "ref_id": null } ], "ref_spans": [], "eq_spans": [], "section": "Combinatory Categorial Grammar and Derivations of Double Subject Construction", "sec_num": "2" }, { "text": "We have to introduce another lexical rule which inserts a gap into an expression of functional category in order to avoid case conflict in DSC. According to Jacobson (1999) , this rule changes an argument of a functional category into a gap and encodes the information of the unrealized argument on the result as a feature. Let us call this rule the 'category shift.' The expression which undergoes this rule becomes a gap-containing category. A.x.wifet-of(x) In (20), the category shift rule changes the expression of category TPNom/LTPpos.Gen to the expression containing a gap in its possessor position of category TP NomITPpos, in which the case feature of the gap is deleted because it will not combine with a phonologically realized argument. The information of the missing argument ITPpos should be taken as a kind of feature encoded on the result category (Jacobson 1999:150) .", "cite_spans": [ { "start": 157, "end": 172, "text": "Jacobson (1999)", "ref_id": "BIBREF6" }, { "start": 864, "end": 883, "text": "(Jacobson 1999:150)", "ref_id": null } ], "ref_spans": [], "eq_spans": [], "section": "Combinatory Categorial Grammar and Derivations of Double Subject Construction", "sec_num": "2" }, { "text": "Given the CS rule in (19) and the category and type of relational NPs, we can show the desired derivation of sentence (lb) as in 21 => NPNomINPpos : Ax.condition'-of(x) cs --SITPpos : Ax.is-serioust(condition'-of(x))", "cite_spans": [ { "start": 135, "end": 168, "text": "NPNomINPpos : Ax.condition'-of(x)", "ref_id": null } ], "ref_spans": [], "eq_spans": [], "section": "Combinatory Categorial Grammar and Derivations of Double Subject Construction", "sec_num": "2" }, { "text": "B< Predication I Predication II", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Combinatory Categorial Grammar and Derivations of Double Subject Construction", "sec_num": "2" }, { "text": "S:is-serious'(conditiont-of(prime minister'))", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Combinatory Categorial Grammar and Derivations of Double Subject Construction", "sec_num": "2" }, { "text": "In (21), byoojoo 'condition' cannot directly combine with the overt possessor because it has undergone the category shift and become a gap containing expression. Therefore, it composes with the predicate omoi 'serious', saturating its 0-grid and resulting in the 'sentence' with a gap in the possessor position. For the sake of simplicity, we will omit the steps of lambda abstraction from the derivations shown below. It should be noticed here that the category of propositions in (20) is not of S or IP, as has been assumed in the literature of generative or traditional grammars, but of SITP, which indicates that these derived (open) propositions are property denoting expressions of type , exactly as normal VPs.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Combinatory Categorial Grammar and Derivations of Double Subject Construction", "sec_num": "2" }, { "text": "In the framework assuming the tight syntax-semantics relation in the theory of grammar, derived propositions with gaps can never have the category of sentence which denotes truth-values, not properties.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Combinatory Categorial Grammar and Derivations of Double Subject Construction", "sec_num": "2" }, { "text": "Consequently, the category of major subjects must also be changed. The category of the derived predicate in (21) is SINPpos (an open sentence containing a possessor gap), but the category of the gap, INP, cannot be cancelled by the NP category of the major subject because INP only indicates that the possessor is not phonologically realized. Let us assume that languages like Japanese or Korean have another kind of type raising rule, which raises (major) subject NPs to the category taking an open proposition as an argument to return a (fully saturated) proposition in 23b Steedman (2000) proposes to associate the rules of type raising with case assignment. Kruijff and Vasishth (2001:18) also uses the type raising to generate a noun in the subject position. That is, the raised category S/(S/LNP) specifies the noun as nominative case. If this speculation is correct, we may have the type-raised category (23b) for major subjects in addition to the standard raised category (23a) for normal subjects. (23b) assigns the interpretation as a generalized quantifier to major subjects in the same way as (23a). Normal subjects must be (external) arguments taken by predicates, whereas major subjects are not arguments of predicates at all. The category in (23b) intuitively states that a major subject can be licensed by an open sentence with a gap, which can occur anywhere in the open sentence in principle. In the derivation (21), the category of the major subject syusyoo-ga 'prime minister' should be type-raised over the open proposition byoojoo-ga omoi 'x's condition is serious' and assigned the category of S/(SITP).4 Then it can combine with the derived predicate (roughly, Xx.x's condition is serious) without causing case conflict.", "cite_spans": [ { "start": 576, "end": 591, "text": "Steedman (2000)", "ref_id": "BIBREF17" }, { "start": 662, "end": 692, "text": "Kruijff and Vasishth (2001:18)", "ref_id": null } ], "ref_spans": [], "eq_spans": [], "section": "Combinatory Categorial Grammar and Derivations of Double Subject Construction", "sec_num": "2" }, { "text": "Given the category and type for relational nouns and the category shift in (19), our analysis of the DSC proposed in Section 2 can be straightforwardly extended to the constructions with more than one nominative object or accusative object in Japanese and Korean. DNOC and DAC share the same syntactic and semantic properties with DSC (in fact, it seems that the semantic relationship between the possessors and possesses in these constructions is much more restricted and they must stand in a so-called inalienable possession relation or the whole-part relation s). It is clear that the preceding 4 Chierchia (1985:437) suggests the following rule of nominative case assignment: (i) Propositional functions assign case. In this sense, it is not surprising that open propositions, as well as normal VPs, can assign nominative case to major subjects in DSC. 5 We should classify relational nouns into two subgroup: reflexive relational nouns and pronominal relational nouns. The former refers to parts of individuals, whereas the latter refers to independent individuals standing in some close relation, e.g., a kinship relation, with individuals denoted by possessors. Reflexive relational nouns are similar to standard reflexives in that the order of an antecedent and a reflexive or relational noun cannot be reversed, as shown in (i): =>cs _ IgAccMp i os:ALhead'-of(x) VP/(VPINP)", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Double Nominative Objects and Double Accusative Objects", "sec_num": "3" }, { "text": "The raised category for objects states that the objects are the expressions that combines with open verb phrases to yield verb phrases. The type raising rule (28) might be disallowed under standard categorical grammars in the Montague tradition because the raised object NPs can not easily be associated with the interpretation as generalized quantifiers. But if we extend the proposal of relating type raising and positioning of nouns in Steedman (2000) and Kruijff and Vasishth (2001) , we can posit (28) for possessor objects in DNOC and DAC as an additional type raising in Japanese and Korean. Let us leave open the question of whether this kind of special type raising rule should be required in our theory of grammar or not.", "cite_spans": [ { "start": 439, "end": 454, "text": "Steedman (2000)", "ref_id": "BIBREF17" }, { "start": 459, "end": 486, "text": "Kruijff and Vasishth (2001)", "ref_id": "BIBREF9" } ], "ref_spans": [], "eq_spans": [], "section": "Double Nominative Objects and Double Accusative Objects", "sec_num": "3" }, { "text": "Under the analysis proposed here, we can explain the syntactic derivations and semantic interpretations of DSC, DNOC and DAC in a unified manner. In languages like English, NPs marked with structural cases strictly correspond to external or internal arguments of predicates. On the other hand, in languages like Japanese and Korean which does not exhibit morphological agreement, NPs which are not licensed by the theta-grids of predicates can show up with structural cases. We propose that relational nouns contain variables inherently and that they are different in category and type from ordinary nouns denoting individuals.6 We also argue that some subjects and objects are licensed by open expressions in Japanese and Korean. The simple framework of CCG with function composition allows flexible constituency in derivations of expressions and gives a unified account to the phenomena of multiple occurrence of subjects and objects in simple sentences in these two languages.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Double Nominative Objects and Double Accusative Objects", "sec_num": "3" }, { "text": "possessee NPs, ignoring syntactic and semantic characteristics shared by all the possesees in the three kinds of constructions, the contrasts between possessors and possesees in extractability in (7)-(9) (repeated here as (29)-(31)) must be treated separately. In sentences (29)-(31), it is only possessor NP's that can be extracted from relative clauses, regardless of surface case-marking.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Double Nominative Objects and Double Accusative Objects", "sec_num": "3" }, { "text": "(29) a. byoojoo-ga omoi syusyoo illness-NOM serious prime minister 'the prime minister who is seriously ill' b. * syusyoo-ga omoi byoojoo the prime minister-NOM serious illness (30) a. Hanako-ni naiyoo-ga wakar-u hon Hanako-DAT content-NOM understand-PRES book 'the book the content of which Hanako understands' b. * Hanako-ni kono hon-ga wakar-u naiyoo Hanako-DAT this book-NOM understand content (31) a. Thalo-ka meli-lul simhakey ttayli-n Hanakho Thalo-Nom head-Acc hard hit-PAST Hanakho 'Hanakho whose head Thalo hit hard' b. * Thalo-ka", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Double Nominative Objects and Double Accusative Objects", "sec_num": "3" }, { "text": "Hanakho-lul simhakey ttayli-n meli Thalo-Nom Hanakho-Acc hard hit-PAST head Here, let us limit our selvesto the so-called head-external (standard) relative clauses which contain gaps corresponding to the heads. In this kind of relative clause, an open proposition containing a gap (i.e., a relative clause) modifies a head noun. Let us analyze the noun-modifying form of predicates as the verb stem plus the noun modifying suffix of category (N/N)/(S/LNP). First, let us look at the asymmetries in extractability between possessor and possessee NPs in DSC. The derivations of (29a) and (29b) can be shown as in (32a) and (32b):", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Double Nominative Objects and Double Accusative Objects", "sec_num": "3" }, { "text": "(32) a.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Double Nominative Objects and Double Accusative Objects", "sec_num": "3" }, { "text": "byoojoo-ga omo-i syusyoo NPINPpos: x.illness'of(x) SINP: Ax.serious \"(X In (32a) where the major subject is extracted, though the argument structure of the predicate omoi 'serious' in the relative clause is saturated, a gap remains in the possessor position of the subject. The relative clause byoojoo-ga omoi 'whose condition is serious' is, therefore, an open sentence of category SITP and type , and denotes a property of an individual corresponding to the head noun syusyoo 'Prime Minister'. On the contrary, the relational noun is extracted in ungrammatical (32b) and the relative clause will be construed as a complete proposition, which cannot modify a head noun and denote its property.7 range of phenomena involving multiple subject and object constructions and the asymmetries in extraction with relative clause formation, reflecting native speakers' intuition about syntactic constituency and semantic interpretations of these constructions.", "cite_spans": [ { "start": 25, "end": 34, "text": "NPINPpos:", "ref_id": null } ], "ref_spans": [], "eq_spans": [], "section": "Double Nominative Objects and Double Accusative Objects", "sec_num": "3" }, { "text": "It is well known that more than two NPs marked with nominative or accusative case can appear in simple sentences in Japanese and Korean, but we limit ourselves to those containing just two nominative NPs or accusative NPs in this paper. Our analysis can easily be extended to deal with multiple subject/object constructions in general.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "", "sec_num": null }, { "text": "Double Subject and Double Object Constructions and RelativizationSo far, we have dealt with the derivations of DSC, DNOC, and DAC in a unified manner. If our discussions centers only around the multiple assignment of an identical case to the possessor and", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "", "sec_num": null }, { "text": "representations. In fact, our approach is variable-free in the sense ofJacobson (1999Jacobson ( , 2000. Note that we use variables in the derivations throughout this paper only for ease of exposition.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "", "sec_num": null }, { "text": "It should be noted here that the ungrammatical status of (32b) cannot simply be attributed to some semantic or pragmatic reason. We can easily come up with a DSC examples, in which relational nouns can be omitted and still grammatical. Consider: (i) a. Kono hebi-ga doo-ga naga-i.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "", "sec_num": null } ], "back_matter": [ { "text": "nominative and accusative objects are NOT arguments subcategorized by predicates though they seem to be assigned structural cases. We will argue that these possessor objects are licensed by certain constituents derived in a way similar to open propositions in DSC.The examples involving multiple occurrence of nominative and accusative objects are repeated here for convenience as (24) and (25).(24) a. Hanako-ni kono hon-no naiyoo-ga yoku wakar-u. Hanako-DAT this book-GEN content-NOM well understand-PRES 'Hanako understands the content of this book well.' . Hanako-m kono hon-ga yoku naiyoo-ga wakar-u. Hanako-DAT this book-NOM well content-NOM understand-PRES. (25) a. Thalo-nun Hanakho-uy meli-lul simhakey ttayly-essta.Thalo-NOM Hanakho-GEN head-ACC hard hit-PAST 'Taroo hit Hanako's head hard.' b. Thalo-nun Hanakho-lul simhakey meli-lul ttayly-essta.Thalo-NOM Hanakho-ACC hard head-ACC hit-PAST (26) illustrates the derivation of double nominative object sentence (24b):KNPExp. Dat: 2y.understand (content-of(this-boo0)(Y) As mentioned in foot note 3, let us assume that the dative NP Hanako is the subject and the nominative NP naiyoo-ga 'content' is the object in this sentence. Also assume that this case array is lexically specified for a set of psych-predicates which are subcategorized for experiencer and theme arguments. In derivation (26), the relational noun naiyoo-ga must undergo the category shift and get a gap inserted into its possessor position. This shifted nominative object combine with the transitive verb wakaru 'understand' to yield an open predicate naiyoo-ga wakaru 'understand the content-of-x', which still has an > interpretation. Since, according to our definition of the object in (17b), an object is any NP that combines with (S/NP)/NP (or (S/LNP)/NP)) to form S/LNP, this open predicate can take another OBJECT to form a predicate phrase, and the additional nominative object kono hon-ga 'this book' is licensed, as desired.This derivation holds for DAC in Korean, as we can see the derivation of (25) in (27):In contrast, pronominal relational nouns denote certain (close) relations between independent individuals, and can be easily passivised. (iii) a. John-ga hahaoya-o korosi-ta. John-NOM mother-ACC kill-PAST 'John killed his mother.' b. Hahaoya-ga John-ni koros-are-ta.mother-NOM John-BY kill-PASS-PAST 'His mother was killed by John.' In (iiib), hahaoya-ga 'mother' can naturally be construed as John's mother.Descriptively, we can circumvent this kind of semantic restriction by stating that only possessors of reflxive relational nouns (that is, nouns denoting inalienable possession relations) can be assigned the same case as the possessee in DNOC and DAC. This line of argument immediately holds for the asymmetries in extractability in DNOC and DAC. Let us look at the contrast in extractability between the possessor and possesee in relativization of DAC:(33) a. In the grammatical (33a), in which the possessor is extracted, the gap in the possessor position of the relational obbject meli-lul 'head' is passed up to the relative clause, even though the theta-grid of the transitive verb is saturated. Thus, the relative clause Thalo-ka meli-lul simhakey ttayl-in 'Taroo hit x's head' is an open expression of category SINP and type , and can modify the head noun, denoting a property of Hanakho. In (33b) where the relational noun is extracted, the relative clause is a complete sentence with no gap and cannot become a property denoting expression. The category and type of relational nouns and relative clause formation using function composition can give a unified account to the asymmetries in extractability between possessors' and possesees in DSC, DNOC and DAC in Japanese and Korean.", "cite_spans": [ { "start": 1005, "end": 1030, "text": "(content-of(this-boo0)(Y)", "ref_id": null } ], "ref_spans": [], "eq_spans": [], "section": "annex", "sec_num": null }, { "text": "In the literature, double/multiple subject and object constructions have been treated as separate phenomena because the discussions have centered around the multiple assignment (licensing) of identical cases to more than one NP in simple sentences. This approach can not derive DSC, DNOC, and DAC and explain the asymmetries in extractability in relative clauses uniformly. On the other hand, we have shown that these constructions share many important characteristics in addition to multiple case assignment. We argue that the possessors and possessees are different in categories and types in these constructions, and that the possessees are syntactically and semantically functional, taking possessors as arguments. We also propose a rule inserting a gap into a possessor position, which is encoded on the functional category as a feature and passed up to derived constituents. It is also argued that the possessor subjects and objects are not licensed by the theta-grids of predicates but by constituents like open propositions or open predicates containing free variables in these constructions. The flexible constituency assumed in CCG can give an unified account to a wide this snake-NOM body-NOM long-PRES This snake's body is long.' b. Kono hebi-ga naga-i.This snake is long.' However, we have asymmetries between the possessor and possessee in extractability with relative clause formation. (ii) a. . doo-ga naga-i hebi body-NOM long-REL snake 'the snake whose body is long' b.* kono hebi-ga naga-i doo this snake-NOM long-REL body In the ungrammatical (iib), kono hebi-ga nagai is semantically and/or pragmatically fine, but this clause does not contain any gap and cannot modify the head noun doo 'body'.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Conclusion", "sec_num": "5" } ], "bib_entries": { "BIBREF0": { "ref_id": "b0", "title": "A Categorial Account of Extraction Asymmetries in Tagalog and Toba Batak", "authors": [ { "first": "J", "middle": [], "last": "Baldridge", "suffix": "" } ], "year": 2000, "venue": "", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Baldridge, J. 2000. A Categorial Account of Extraction Asymmetries in Tagalog and Toba Batak, University of Edinburgh.", "links": null }, "BIBREF1": { "ref_id": "b1", "title": "Formal Semantics and the Grammar of Predication", "authors": [ { "first": "Gennaro", "middle": [], "last": "Chierchia", "suffix": "" } ], "year": 1985, "venue": "Linguistic Inquiry", "volume": "16", "issue": "", "pages": "417--443", "other_ids": {}, "num": null, "urls": [], "raw_text": "Chierchia, Gennaro (1985) Formal Semantics and the Grammar of Predication. Linguistic Inquiry 16, 417-443.", "links": null }, "BIBREF2": { "ref_id": "b2", "title": "Inalienable-Type Multiple Accusative Constructions in Korean and Japanese", "authors": [ { "first": "D", "middle": [], "last": "Choe", "suffix": "" } ], "year": 1993, "venue": "Japanese/Korean Linguistics", "volume": "2", "issue": "", "pages": "319--337", "other_ids": {}, "num": null, "urls": [], "raw_text": "Choe, D. 1993. Inalienable-Type Multiple Accusative Constructions in Korean and Japanese,' Japanese/Korean Linguistics 2, 319-337.", "links": null }, "BIBREF3": { "ref_id": "b3", "title": "Syntactic Adjunction, A-Chains and the ECP --Multiple Identical Case Construction in Korean", "authors": [ { "first": "H", "middle": [], "last": "Choe", "suffix": "" } ], "year": 1987, "venue": "Proceedings of North Eastern Linguistic Society Annual Meeting", "volume": "17", "issue": "", "pages": "100--120", "other_ids": {}, "num": null, "urls": [], "raw_text": "Choe, H. 1987. Syntactic Adjunction, A-Chains and the ECP --Multiple Identical Case Construction in Korean. Proceedings of North Eastern Linguistic Society Annual Meeting 17, 100-120.", "links": null }, "BIBREF4": { "ref_id": "b4", "title": "Toward a Minimalist Theory of Syntactic Structure, ms. Ohio State University. Engdahl, E. 1988. Relational Interpretation", "authors": [ { "first": "D", "middle": [], "last": "Dowty", "suffix": "" } ], "year": 1997, "venue": "Metal Representation: The Interface between Language and Reality", "volume": "", "issue": "", "pages": "63--82", "other_ids": {}, "num": null, "urls": [], "raw_text": "Dowty, D. 1997. Toward a Minimalist Theory of Syntactic Structure, ms. Ohio State University. Engdahl, E. 1988. Relational Interpretation. In R. Kempson (ed.), Metal Representation: The Interface between Language and Reality, 63-82. Cambridge University Press.", "links": null }, "BIBREF5": { "ref_id": "b5", "title": "Head-Movement and Case-Marking in Japanese", "authors": [ { "first": "N", "middle": [], "last": "Fukui", "suffix": "" }, { "first": "T", "middle": [], "last": "Nishigauchi", "suffix": "" } ], "year": 1992, "venue": "Journal of Japanese Linguistics", "volume": "14", "issue": "", "pages": "1--35", "other_ids": {}, "num": null, "urls": [], "raw_text": "Fukui, N. and T. Nishigauchi. 1992. Head-Movement and Case-Marking in Japanese, Journal of Japanese Linguistics 14, 1-35, 1992.", "links": null }, "BIBREF6": { "ref_id": "b6", "title": "Towards a Variable-Free Semantics", "authors": [ { "first": "P", "middle": [], "last": "Jacobson", "suffix": "" } ], "year": 1999, "venue": "Linguistics and Philosophy", "volume": "22", "issue": "", "pages": "117--184", "other_ids": {}, "num": null, "urls": [], "raw_text": "Jacobson, P. 1999. Towards a Variable-Free Semantics, Linguistics and Philosophy 22, 117-184.", "links": null }, "BIBREF7": { "ref_id": "b7", "title": "Paycheck Pronouns, Bach-Peters Sentences, and Variable-free Semantics", "authors": [ { "first": "P", "middle": [], "last": "Jacobson", "suffix": "" } ], "year": 2000, "venue": "Natural Language Semantics", "volume": "8", "issue": "", "pages": "77--155", "other_ids": {}, "num": null, "urls": [], "raw_text": "Jacobson, P. 2000. Paycheck Pronouns, Bach-Peters Sentences, and Variable-free Semantics. Natural Language Semantics 8: 77-155.", "links": null }, "BIBREF8": { "ref_id": "b8", "title": "A Constraint-Based Approach to Some Multiple Nominative Constructions in Korean", "authors": [ { "first": "J", "middle": [], "last": "Kim", "suffix": "" } ], "year": 2000, "venue": "Proceedings of PACLIC", "volume": "14", "issue": "", "pages": "165--176", "other_ids": {}, "num": null, "urls": [], "raw_text": "Kim, J. 2000. A Constraint-Based Approach to Some Multiple Nominative Constructions in Korean. Proceedings of PACLIC 14, 165-176. Logico-Linguistic Society of Japan.", "links": null }, "BIBREF9": { "ref_id": "b9", "title": "Compence and Performance Modelling of Free Word Order", "authors": [ { "first": "G", "middle": [], "last": "Kruijff", "suffix": "" }, { "first": "S", "middle": [], "last": "Vasishth", "suffix": "" } ], "year": 2001, "venue": "", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Kruijff, G. and S. Vasishth. 2001. Compence and Performance Modelling of Free Word Order, ESSLLI 2001 Course Reader, Helsinki University.", "links": null }, "BIBREF10": { "ref_id": "b10", "title": "The Structure of the Japanese Language", "authors": [ { "first": "S", "middle": [], "last": "Kuno", "suffix": "" } ], "year": 1973, "venue": "", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Kuno, S. 1973. The Structure of the Japanese Language. MIT Press.", "links": null }, "BIBREF11": { "ref_id": "b11", "title": "A CCG Fragment of Korean", "authors": [ { "first": "K", "middle": [], "last": "Lee", "suffix": "" } ], "year": 2000, "venue": "Proceedings of PACLIC", "volume": "14", "issue": "", "pages": "219--229", "other_ids": {}, "num": null, "urls": [], "raw_text": "Lee, K. 2000. A CCG Fragment of Korean. Proceedings of PACLIC 14, 219-229, Logico-Linguistic Society of Japan.", "links": null }, "BIBREF12": { "ref_id": "b12", "title": "Meidai-no Bunpoo: Nihongo Bunpo Josetsu", "authors": [ { "first": "T", "middle": [], "last": "Masuoka", "suffix": "" } ], "year": 1987, "venue": "", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Masuoka, T. 1987. Meidai-no Bunpoo: Nihongo Bunpo Josetsu. Kurosio Publishers.", "links": null }, "BIBREF13": { "ref_id": "b13", "title": "Tajuu-Syugo Koobun-o Megutte", "authors": [ { "first": "K", "middle": [], "last": "Mihara", "suffix": "" } ], "year": 1990, "venue": "Nihongo-gaku", "volume": "9", "issue": "8", "pages": "66--76", "other_ids": {}, "num": null, "urls": [], "raw_text": "Mihara, K. 1990. Tajuu-Syugo Koobun-o Megutte. Nihongo-gaku 9:8, 66-76.", "links": null }, "BIBREF14": { "ref_id": "b14", "title": "A Parametric Approach to Case Alternation Phenomena in Japanese", "authors": [ { "first": "M", "middle": [], "last": "Morikawa", "suffix": "" } ], "year": 1993, "venue": "", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Morikawa, M. 1993. A Parametric Approach to Case Alternation Phenomena in Japanese. Hituzi Syobo.", "links": null }, "BIBREF15": { "ref_id": "b15", "title": "Nihon-go no Daisyugo to Syudai", "authors": [ { "first": "T", "middle": [], "last": "Sugimoto", "suffix": "" } ], "year": 1990, "venue": "Kyusyu Daigaku Kougakubu Kiyou", "volume": "3", "issue": "", "pages": "165--182", "other_ids": {}, "num": null, "urls": [], "raw_text": "Sugimoto, T. 1990. Nihon-go no Daisyugo to Syudai, Kyusyu Daigaku Kougakubu Kiyou 3, 165-182.", "links": null }, "BIBREF16": { "ref_id": "b16", "title": "Surface Structure and Interpretation", "authors": [ { "first": "M", "middle": [], "last": "Steedman", "suffix": "" } ], "year": 1996, "venue": "", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Steedman, M. 1996. Surface Structure and Interpretation. MIT Press.", "links": null }, "BIBREF17": { "ref_id": "b17", "title": "The Syntactic Process", "authors": [ { "first": "M", "middle": [], "last": "Steedman", "suffix": "" } ], "year": 2000, "venue": "", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Steedman, M. 2000. The Syntactic Process. MIT Press.", "links": null }, "BIBREF18": { "ref_id": "b18", "title": "Remarks on Double-Nominative Sentences", "authors": [ { "first": "M", "middle": [], "last": "Shibatani", "suffix": "" }, { "first": "C", "middle": [], "last": "Cotton", "suffix": "" } ], "year": null, "venue": "Papers in Japanese Linguistics", "volume": "5", "issue": "", "pages": "1976--77", "other_ids": {}, "num": null, "urls": [], "raw_text": "Shibatani, M. and C. Cotton. 1976-77. Remarks on Double-Nominative Sentences. Papers in Japanese Linguistics 5, 261-277, 1976-77.", "links": null }, "BIBREF19": { "ref_id": "b19", "title": "Topicalization and Subjectivization in Japanese: Characterization and Identificational Information", "authors": [ { "first": "K", "middle": [], "last": "Takami", "suffix": "" }, { "first": "A", "middle": [], "last": "Kamio", "suffix": "" } ], "year": 1996, "venue": "Lingua", "volume": "99", "issue": "", "pages": "207--235", "other_ids": {}, "num": null, "urls": [], "raw_text": "Takami, K. and A. Kamio. 1996. Topicalization and Subjectivization in Japanese: Characterization and Identificational Information. Lingua 99, 207-235.", "links": null }, "BIBREF20": { "ref_id": "b20", "title": "A Configurational Approach to Case-marking in Japanese", "authors": [ { "first": "K", "middle": [], "last": "Takezawa", "suffix": "" } ], "year": 1987, "venue": "", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Takezawa, K. 1987. A Configurational Approach to Case-marking in Japanese. Ph. D. dissertation, University of Washington.", "links": null }, "BIBREF21": { "ref_id": "b21", "title": "A CCG Account of Tzotzil Pied-Piping", "authors": [ { "first": "Frank", "middle": [ "R" ], "last": "Trechsel", "suffix": "" } ], "year": 2000, "venue": "Natural Language and Linguistic Theory", "volume": "18", "issue": "", "pages": "611--663", "other_ids": {}, "num": null, "urls": [], "raw_text": "Trechsel, Frank R. 2000. A CCG Account of Tzotzil Pied-Piping. Natural Language and Linguistic Theory 18: 611-663.", "links": null }, "BIBREF22": { "ref_id": "b22", "title": "Seisei-Nihongogaku Nyuumon. Taisyuukan", "authors": [ { "first": "N", "middle": [], "last": "Hasegawa", "suffix": "" } ], "year": 1999, "venue": "", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Hasegawa, N. 1999. Seisei-Nihongogaku Nyuumon. Taisyuukan.", "links": null } }, "ref_entries": { "FIGREF0": { "uris": null, "text": ": (12) a. Tanaka-san-ga okusan-ga bijin-des-u. Tanaka-Mr.-NOM wife-NOM beautiful-woman-BE-PRES 'Mr. Tanaka's wife is beautiful.' b. * Tanaka-san-ga Hanako-san-ga bijin-des-u. Tanaka-Mr.-NOM Hanako-Ms.-NOM beautiful-woman-BE-PRES (13) a. Yamaguti-san-ga motikabu-ga kyuuraku-s-ita. Yamaguti-Mr.-NOM shares-NOM sudden-drop-DO-PAST 'The price of Mr. Yamaguti's shares fell suddenly.' b.??Tanaka-san-ga kabu-ga kyuuraku-s-ita. Yamaguti-Mr.-NOM stock-NOM sudden-drop-DO-PAST Cf. Yamaguti-san-no kabu-ga kyuuraku-s-ta. Yamaguti-Mr.-GEN stock-NOM sudden-drop-DO-PAST", "type_str": "figure", "num": null }, "FIGREF1": { "uris": null, "text": "prime minister'NP I NP : Ax.illness'-of(x)< S/ NP : 2y.serious'(y) '(illness'-of(prime-minister))", "type_str": "figure", "num": null }, "FIGREF2": { "uris": null, "text": "17) a. A SUBJECT is any NP that combines with S/NP (or S/LNP) to form S. b. A DIRECT OBJECT (DO) is any NP that combines with (S/NP)/NP (or (S/LNP)/NP)) to form S/LNP. Let us turn to the derivation of double subject sentences like (lb): : 2x.serious'(illness'-oj(x)) Pos.Gen", "type_str": "figure", "num": null }, "FIGREF3": { "uris": null, "text": "19) the Category Shift (CS): Let us indicate an argument of a functional category by T. Any expression of category (A/B)/...0 can shift to (AIC)/B. (19) is a unary rule mapping an expression into a new expression. (19), for instance, change the category of relational expressions like okusan 'wife', as in (", "type_str": "figure", "num": null }, "FIGREF4": { "uris": null, "text": "prime_minister' NPNonANPp os.Gen : ibccondition '-of(x) S/ L TP Nom: 2y.is-serious'(y)", "type_str": "figure", "num": null }, "FIGREF5": { "uris": null, "text": "(i) a. John-ga jibun-o seme-ta. John-NOM self-ACC blame-PAST 'John blamed himself.' b. * Jibun-ga John-ni semer-are-ta. SELF-NOM John-BY blame-PASS-PAST (ii) a. John-ga asi-o ot-ta.John-NOM leg-ACC brake-PAST 'John broke his leg.' b. * Asi-ga John-ni or-are-ta.leg-NOM John-BY break-PASS-x.head-of(x) : Ax.2t,y.hit'(x)(y)", "type_str": "figure", "num": null }, "FIGREF6": { "uris": null, "text": ")B (N/N)/(SINP) N: prime-minister' SINP : Ax.seriouslillness'-of(x)) Pos N/N: A.P.Ax.serious'(illness\"-of(x)) & P(x) N: Ax.serious'(illness 1 (x)) & prime-minister'(x)", "type_str": "figure", "num": null }, "TABREF1": { "num": null, "type_str": "table", "text": "In fact, expressions which have undergone the category shift or inherited gaps from shifted expressions via function composition are simply open expressions containing free variables. The derived expression byojo-ga omoi 'seriously ill' in (21), for example, is an open proposition with an unbound variable in the possessor position of the second subject. This open proposition can be a predicate by the standard lambda abstraction over a free variable, defined as in (22).(22) If F is a formula and x is a variable, then Ax.Fx is a predicate.", "content": "", "html": null } } } }