| { |
| "paper_id": "Y96-1033", |
| "header": { |
| "generated_with": "S2ORC 1.0.0", |
| "date_generated": "2023-01-19T13:38:40.454348Z" |
| }, |
| "title": "Legitimate termination of nonlocal features in HPSG Hywel Evans Tsuru University takigmfLor.jp", |
| "authors": [], |
| "year": "", |
| "venue": null, |
| "identifiers": {}, |
| "abstract": "This paper reviews the treatment of wh-question facts offered by Lappin and Johnson 1996, and suggests that their account of certain island phenomena should be adapted by assuming that certain phrase structures license binding of inherited features. In Japanese, Lappin and Johnson's INHERILQUE feature appears to be dependent on INHERIQUE in order to terminate with a functional C head's TO-BINDIQUE. For certain languages, C's TO-BINDILQUE feature must be null if TO-BINDIQUE is null. In the spirit of Sag 1996 and Pollard and Yoo 1996, the facts can be handled by saying that TO-BINDILQUE is licensed on a wh-clause (wh-cl). As a wh-cl requires TO-BINDIQUE, the dependence of the less robust INHERILQUE on INHERIQUE is thus explained.", |
| "pdf_parse": { |
| "paper_id": "Y96-1033", |
| "_pdf_hash": "", |
| "abstract": [ |
| { |
| "text": "This paper reviews the treatment of wh-question facts offered by Lappin and Johnson 1996, and suggests that their account of certain island phenomena should be adapted by assuming that certain phrase structures license binding of inherited features. In Japanese, Lappin and Johnson's INHERILQUE feature appears to be dependent on INHERIQUE in order to terminate with a functional C head's TO-BINDIQUE. For certain languages, C's TO-BINDILQUE feature must be null if TO-BINDIQUE is null. In the spirit of Sag 1996 and Pollard and Yoo 1996, the facts can be handled by saying that TO-BINDILQUE is licensed on a wh-clause (wh-cl). As a wh-cl requires TO-BINDIQUE, the dependence of the less robust INHERILQUE on INHERIQUE is thus explained.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Abstract", |
| "sec_num": null |
| } |
| ], |
| "body_text": [ |
| { |
| "text": "In earlier versions of the Head-Driven Phrase Structure (HPSG) framework, inheritance of nonlocal features is terminated in accordance with the Nonlocal Feature Principle (NFP).", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Introduction", |
| "sec_num": "1." |
| }, |
| { |
| "text": "1. NONLOCAL FEATURE PRINCIPLE (Pollard and Sag 1994, page 164) For each nonlocal feature, the INHERITED value on the mother is the union of the inherited values on the daughters minus the TO-BIND value on the head daughter.", |
| "cite_spans": [ |
| { |
| "start": 30, |
| "end": 62, |
| "text": "(Pollard and Sag 1994, page 164)", |
| "ref_id": null |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Introduction", |
| "sec_num": "1." |
| }, |
| { |
| "text": "The Head-Filler Rule stipulates that a TO-BINDISLASH value appears on the head daughter, thereby terminating INHERISLASH. Adjectives like easy, which occur in tough-constructions, carry a TO-BINDISLASH feature which allows the INHERISLASH feature on its complement VP[inf] to be terminated at the AP level. Similarly, INHERIREL carried on to a relativizer phrase (RP) will be terminated legitimately as the RP modifies a nominal head which bears a TO-BINDIREL feature structure-shared with the INHERIREL feature on the RP.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Introduction", |
| "sec_num": "1." |
| }, |
| { |
| "text": "In line with this, Lappin and Johnson 1996a&b proposes that INHERIQUE is terminated by a wh-complementizer which carries TO-BINDIQUE. Languages like Chinese and Japanese --in which wh-phrases normally appear in situ -have a wh-complementizer which subcategorizes for S[fin] with a non-empty INHERIQUE structure-shared with the TO-BINDIQUE on the C head.", |
| "cite_spans": [ |
| { |
| "start": 19, |
| "end": 45, |
| "text": "Lappin and Johnson 1996a&b", |
| "ref_id": null |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Introduction", |
| "sec_num": "1." |
| }, |
| { |
| "text": "la. SUBCAT <S[fin,INHERIQUE: X]> b. NONLOCALITO-BINDIQUE: Y c. Condition: Y is a subset of X In 1, we have structure-sharing between the INHERIQUE value inherited onto the complement S and the TO-BINDIQUE value on the C head. The Japanese Q-particle ka is a candidate as a phonologically noninvisible complementizer. In the Government & Binding framework (GB), for example, ka was assumed to head a CP (in line with the 'extended' projection principle), with raising to spec of this structure.", |
| "cite_spans": [ |
| { |
| "start": 4, |
| "end": 32, |
| "text": "SUBCAT <S[fin,INHERIQUE: X]>", |
| "ref_id": null |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Introduction", |
| "sec_num": "1." |
| }, |
| { |
| "text": "2. Bill-wa John-ga nani-o katta ka shiritagatte-imasu yo Bill-top J-nom what-acc bought Q want-to-know emphatic dec particle \"Bill wants to know what John bought!\" *\"What does Bill want to know if John bought?\"", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Introduction", |
| "sec_num": "1." |
| }, |
| { |
| "text": "In 2 the wh-expression may only take narrow scope because of the presence of ka at that level and the forced declarative interpretation at the matrix clause.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Introduction", |
| "sec_num": "1." |
| }, |
| { |
| "text": "Therefore, the specifications given in 1 are unproblematic.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Introduction", |
| "sec_num": "1." |
| }, |
| { |
| "text": "3. Kimi-wa John-ga nani-o katta ka shiritagatte-imasu ka you-top J-nom what-acc bought Q want-to-know Q \"Do you want to know what John bought?\" \"What do you want to know if John bought?\"", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Introduction", |
| "sec_num": "1." |
| }, |
| { |
| "text": "In 3, the presence of the Q-particle ka forces an interpretation as a question for the clause at which it appears but does not require a wh-expression to take scope In 4, assuming optional TO-BINDIQUE values on the functional C head correctly predicts the interpretive possibilities for examples like 3. Thus, Lappin and Johnson's 1996 treatment --together with the assumption that TO-BINDIQUE features on functional heads may be optionally null --promises to explain the full range of facts relating to interpretation of wh-expressions.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Introduction", |
| "sec_num": "1." |
| }, |
| { |
| "text": "Certain constructions in Japanese provide puzzling evidence that there is some syntactic constraint on covert movement operations.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Subjacency in Japanese", |
| "sec_num": "2." |
| }, |
| { |
| "text": "5. ??Kimi-wa John-ga nani-o katta ka-dooka shiritagatteimasu ka? you-top J-nom what-acc bought Q-yes/no want-to-know Q \"What do you want to know whether or not John bought?\"", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Subjacency in Japanese", |
| "sec_num": "2." |
| }, |
| { |
| "text": "Assuming that ka-dooka has a necessarily null TO-BINDIQUE feature, the only possible interpretation for 5 gives the embedded wh-expression matrix scope. Most speakers of Western Japanese dialects find this construction odd. However, informants tend to agree that there is an improvement when there is a further wh-expression in the matrix clause. Furthermore, 9 is ungrammatical even though there is a wh-expression in the matrix clause. If no feature-checking is required by the embedded whexpression, there is nothing to explain these facts. It should be possible for the presence of the wh-expression in the matrix clause to allow checking of matrix C's Q-feature, thereby allowing a multiple wh-question. Lappin & Johnson account for these facts by assuming that S[fin] blocks the inheritance of QUE in Iraqi Arabic. This is formalized with reference to a parameterized defeasible version of the NFP.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Subjacency in Japanese", |
| "sec_num": "2." |
| }, |
| { |
| "text": "10. For each NONLOCAL feature F, the INHERITED value of F on a mother M is the union of the INHERITED values of F on the non-F-island daughters minus the value of TO-BIND on the head daughter.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Subjacency in Japanese", |
| "sec_num": "2." |
| }, |
| { |
| "text": "The facts in 7, 8, and 9 are successfully handled by assuming that, in Iraqi Arabic, S[fin] is a QUE island. However, this does not explain why the presence of a wh-QP sh in matrix clause initial position apparently allows INHERIQUE to be inherited from tensed clauses.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Subjacency in Japanese", |
| "sec_num": "2." |
| }, |
| { |
| "text": "wh-QP-thought Mona Ali met who \"Who did Mona think Ali met?\"", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "sh-tsawwarit Mona [Ali gabal meno]?", |
| "sec_num": "11." |
| }, |
| { |
| "text": "11 is explained by positing the availability of two complementizers.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "sh-tsawwarit Mona [Ali gabal meno]?", |
| "sec_num": "11." |
| }, |
| { |
| "text": "}] [C[TOBINDILQUE111]sh-] [S [fin,INHERI LQUE{1}] tsawwarit Mona[ CP[INHERILQUE{14C [INHER I LQUE{1}]Ci [ S[fin,INHER I QUE{1}]Ali gabal [NKINHERIQUE{1}] meno]]]]] 11", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "12.[CP[INHERILQUE{", |
| "sec_num": null |
| }, |
| { |
| "text": "In 12, from Lappin and Johnson 1996b, a phonetically null complementizer which has the feature LQUE, unifies with the QUE feature inherited by the C's complement S. A phonetically realized wh-complementizer shcarries a nonoptional TO-BINDILQUE:X feature and subcategorizes for S[INHERILQUE:X]. Lappin and Johnson 1996a acknowledges Sag's 1996 analysis of relative clauses which dispenses with empty complementizers and the TO-BIND feature. In Sag's treatment, in which linguistic types are required to satisfy highly articulated linguistic type constraints, features are inherited through the head phrase, the presence of his QUE and REL feature being required to satisfy typing constraints on wh-and relative clauses. In contrast to Lappin and Johnson's treatment, the termination of QUE does not determine the scope of whexpressions in Sag's treatment. Rather, the presence of QUE at a left peripheral daughter in a phrase structure in English presumably licenses the termination of wh-features (whatever we decide to call them and by whatever means they are inherited on to a higher level of structure). Although Sag does not deal with this problem, the scope of wh-expressions will presumably be determined along the lines of Pollard and Yoo's 1996 treatment of QSTOREs. I do not deal with the problems of Pollard and Yoo's approach to quantifier scope determination here, but suggest that Lappin and Johnson's treatment of inherited features as purely \"syntactic\" features (rather than as necessarily stored interpretations to be unpacked eventually in the CONT feature structure) is attractive in that it promises a simpler and unified treatment of inheritance.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "12.[CP[INHERILQUE{", |
| "sec_num": null |
| }, |
| { |
| "text": "One ", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "12.[CP[INHERILQUE{", |
| "sec_num": null |
| }, |
| { |
| "text": "X],", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "12.[CP[INHERILQUE{", |
| "sec_num": null |
| }, |
| { |
| "text": "where X is non-null, is retained.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "12.[CP[INHERILQUE{", |
| "sec_num": null |
| }, |
| { |
| "text": "It is possible, then, to dispense with complementizers as functional binders of non-local features by adopting the basic idea contained in Pollard. and Yoo 1996 that scope is determined at certain target phrases at which termination of inherited features is licensed. Constraints on inheritance of features may be handled in the spirit of Sag 1996 by imposing language specific (and possibly language-universal) constraints preventing certain features from being inherited on to certain phrases. A rule of \"vehicle change\" allows inheritance of npros to proceed as elements of a different feature value with its own licensed binder.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "12.[CP[INHERILQUE{", |
| "sec_num": null |
| }, |
| { |
| "text": "The basic intuition here is that null complementizers may be dispensed with if we allow termination of non-local features (retrieval of QSTOREs in Pollard and Yoo's terms) to take place at certain target phrases which license this operation. One way to handle this would be to adapt the NFP to dispense with functional heads.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "The solution to the Japanese subjacency problem", |
| "sec_num": "3." |
| }, |
| { |
| "text": "A phrase XP licensed as a binder for feature F inherits the INHERIF:Y features of its daughters minus its own TO-BINDIF:Z value. Condition: Z is a subset of Y Assume that a wh-clause in English is licensed as a binder for QUE. A constraint will determine that TO-BINDIQUE:X of a wh-filler clause will contain the npro contributed by the non-head daughter. In Iraqi Arabic, S[fin] is also licensed as a binder for QUE. As a constraint prevents inheritance of QUE onto S[fin], INHERIQUE:X on Whirs daughters must appear as TO-BINDIQUE:X on mother S[fin] unless QUE is modified to LQUE. Only CP with a sh-head, such as 11, is licensed as a binder for LQUE in Iraqi Arabic.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "New NFP.", |
| "sec_num": "13." |
| }, |
| { |
| "text": "The Japanese \"subjacency\" data can be handled by assuming that Japanese also has a rule allowing \"vehicle change\" of INHERIQUE to INHERILQUE. One might suggest --provisionally --the rule in 14 to handle these cases.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "New NFP.", |
| "sec_num": "13." |
| }, |
| { |
| "text": "14. QUE to LQUE \"vehicle change\" rule If XP[INHERIQUE:X], then either XP [INHERIQUE:X] or XP [INHERILQUE:X] The rule in 14 may be applied to salvage INHERIQUE in contexts in which there is illicit inheritance of the feature on to a particular structure. It allows INHERIQUE to modify to INHERILQUE, but not vice versa. An idiosyncratic constraint applying to CP with ka-dooka as its C head is that it must have an empty INHERIQUE value. This is entirely in line with Sag's suggestions for explaining island phenomena, and is in the spirit of his general program of using type constraints to dispense with null complementizers. The constraint preventing INHERIQUE on ka-dooka CP forces \"vehicle change\" of INHERIQUE to INHERILQUE in order for inheritance of the npros contained in INHERIQUE to proceed to successful termination at a licensed binder structure.", |
| "cite_spans": [ |
| { |
| "start": 73, |
| "end": 86, |
| "text": "[INHERIQUE:X]", |
| "ref_id": null |
| }, |
| { |
| "start": 93, |
| "end": 107, |
| "text": "[INHERILQUE:X]", |
| "ref_id": null |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "New NFP.", |
| "sec_num": "13." |
| }, |
| { |
| "text": "The \"subjacency\" data is then explained by assuming that only a whclause is a licensed binder for INHERILQUE. If a wh-clause in Japanese must satisfy the constraint that it has a non-empty TO-BINDIQUE feature (not TO-BINDILQUE), then the facts fall out perfectly naturally. In 6, by contrast, the \"vehicle change\" rule allows INHERIQUE to modify to INHERILQUE and be inherited successfully on to the ka-dooka CP. The presence of a wh-expression in the matrix clause means that there will be a whclause available to act as a licensed binder for INHERILQUE.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "New NFP.", |
| "sec_num": "13." |
| }, |
| { |
| "text": "The Japanese \"subjacency\" data repeated above has played an important role in the development of the MP. Watanabe's 1991 famous analysis assumed that there were two levels of movement, one an optional operator feature movement taking place at S-structure in order to trigger a CP [wh] , the other at LF --of the whole category --to determine the scope of the wh-expression. This second level of movement was assumed to be free from subjacency restrictions. The approach recommended here has the advantage over Watanabe's solution that there is no need to consider different levels of movement, and therefore no need to motivate the claim that different constraints apply at those levels. Instead, we assume that language-particular constraints prevent certain features being inherited on to certain phrases, certain strategies being available to effect extraction.", |
| "cite_spans": [ |
| { |
| "start": 280, |
| "end": 284, |
| "text": "[wh]", |
| "ref_id": null |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Advantages of this approach", |
| "sec_num": "3.1" |
| }, |
| { |
| "text": "A considerable body of data (Tancredi's 1990 evidence relating to only, for example) has seriously undermined the claim that whole categories moved at LF. In the development of the MP, the need to eradicate certain putatively unmotivated movement operations was widely accepted by linguists working in the Principles and Parameters tradition. Chomsky 1995 assumes that some other interpretive strategy like unselective binding is sufficient to determine the scope of wh-expressions. Therefore, something like Watanabe's original operator movement has been retained to handle checking of features of C[wh]; but further movement is assumed to be blocked as a violation of Economy. Thus although wh-in situ constructions in many of the world's languages exhibit Island effects typical of movement, the MP must resort to completely unrelated mechanisms in order to account for wh-movement and wh-in situ constructions. In this approach, whin situ and wh-movement are both subject to constraints on inheritance --although these may not be the same.", |
| "cite_spans": [ |
| { |
| "start": 28, |
| "end": 44, |
| "text": "(Tancredi's 1990", |
| "ref_id": "BIBREF5" |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Advantages of this approach", |
| "sec_num": "3.1" |
| }, |
| { |
| "text": "While Lappin and Johnson assume that ka-dooka CPs are F-islands for INHERIQUE, they do not resort to LQUE in order to explain the Japanese \"subjacency\" data. Instead, Johnson (pc) proposes an update function on feature values to the effect that the QUE value of the phrase dare-ga in 6 is the union of the value of its own QUE feature and the value of the QUE feature of the F-island ka-dooka CP. While this approach offers a simple solution to the problem, it suffers from a number of problems. The biggest difficulty is that it means that apparently parallel island phenomena are handled in radically different ways for different languages. In Iraqi Arabic, QUE escapes islands by resort to a resumption strategy which replaces QUE with LQUE. However, this solution is rejected for the Japanese data and appeal is made to a completely different mechanism. It would clearly be advantageous to deal with the problem with recourse to the same strategy. In doing so, however, the requirement that INHERIT features are terminated by a functional category carrying TO-BIND features means that considerable optionality has to be factored into their analysis.", |
| "cite_spans": [ |
| { |
| "start": 167, |
| "end": 179, |
| "text": "Johnson (pc)", |
| "ref_id": null |
| }, |
| { |
| "start": 393, |
| "end": 396, |
| "text": "CP.", |
| "ref_id": null |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Advantages of this approach", |
| "sec_num": "3.1" |
| }, |
| { |
| "text": "As already mentioned, treating ka as a functional TO-BIND head requires that we let it carry this feature only optionally. In addition to this, it is necessary to specify that ka allows TO-BINDILQUE, but only if there is also a non-null TO-BINDIQUE value. Therefore, 4 has to be modified to 16. There is considerable optionality here which is factored into functional heads in a more or less arbitrary manner in order to account for the facts. In the account suggested here, by contrast, the new NFP allows npros to be bound at certain licensed levels of structure, with no need to treat ka as a functional head. INHERILQUE arises by the same mechanism in Japanese and Iraqi Arabic, with no need for empty complementizers. These last two points are of considerable importance given the fact that Q complementizers are actually optional in matrix clauses in Japanese. Another problem with Johnson's update function is that nothing prevents adjunct wh-expressions contributing INHERIQUE and these being inherited onto ka-dooka CPs and updated to extract them from the F-island. This in fact makes very wrong predictions as adjunct wh-expressions do not easily extract from interrogative clauses in Japanese at all. One may speculate that this problem can be dealt with very simply by assuming that interrogative clauses are subject to the constraint that any INHERIQUE they carry must have the head feature noun. This requires that npros would contain head feature information, but such a solution would not seem to be beyond the theory, and would be in line with Sag's 1996 suggestions regarding weak islands for SLASH. This would, in fact, solve the problem for Johnson's update function. Interestingly, however, accepting that constraints apply to block certain kinds of INHERIT features on certain kinds of structures weakens the case for F-islands as stated in Lappin and Johnson's version of the NFP.", |
| "cite_spans": [ |
| { |
| "start": 186, |
| "end": 203, |
| "text": "TO-BINDILQUE, but", |
| "ref_id": null |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Advantages of this approach", |
| "sec_num": "3.1" |
| }, |
| { |
| "text": "10. For each NONLOCAL feature F, the INHERITED value of F on a mother M is the union of the INHERITED values of F on the non-F-island daughters minus the value of TO-BIND on the head daughter.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Advantages of this approach", |
| "sec_num": "3.1" |
| }, |
| { |
| "text": "If Lappin and Johnson need to resort to type constraints to block inheritance of certain features in any case, it would be desirable to reduce all inheritance constraints to the same mechanism. There is no obvious way that the block on extraction of adjunct npros out of interrogative clauses can be handled through recourse to F-islands. By contrast, type constraints can be made to straightforwardly handle all the island effects dealt with here. As the suggestions offered here make reference to type constraints rather than F-islands, there is good reason to prefer this as a general approach. Furthermore, dispensing with the unmotivated update function simplifies things considerably and is therefore to be preferred.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Advantages of this approach", |
| "sec_num": "3.1" |
| } |
| ], |
| "back_matter": [ |
| { |
| "text": "The treatment of Relative Clauses in Sag 1996 suggests that imposing highly articlulated constraints on linguistic types will allow us to dispense with invisible categories. Pollard and Yoo's 1996 approach to quantifiers suggests how \"nonlocal\" features can be terminated without recourse to (often invisible) functional heads. Lappin and Johnson's 1996 account of island phenomena in a variety of languages is extremely suggestive of how the presence of whexpressions in situ may give rise to ungrammaticality. The suggestions made above can be viewed as a synthesis of these three approaches. Islands may be handled in general as constraints preventing features from appearing at certain levels of structure, with no need for F-islands. Eradicating functional complementizers as terminators of nonlocal features solves the problem of optionality with respect to these functions.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Conclusion", |
| "sec_num": null |
| } |
| ], |
| "bib_entries": { |
| "BIBREF0": { |
| "ref_id": "b0", |
| "title": "The Minimalist Program", |
| "authors": [ |
| { |
| "first": "N", |
| "middle": [], |
| "last": "Chomsky", |
| "suffix": "" |
| } |
| ], |
| "year": 1995, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Chomsky, N. 1995. The Minimalist Program. The MIT Press.", |
| "links": null |
| }, |
| "BIBREF1": { |
| "ref_id": "b1", |
| "title": "A Critique of the Minimalist Program", |
| "authors": [ |
| { |
| "first": "D", |
| "middle": [], |
| "last": "Johnson", |
| "suffix": "" |
| }, |
| { |
| "first": "S", |
| "middle": [], |
| "last": "Lappin", |
| "suffix": "" |
| } |
| ], |
| "year": 1996, |
| "venue": "Linguistics and Philosophy", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Johnson, D. and S. Lappin. 1996a. A Critique of the Minimalist Program, To appear in Linguistics and Philosophy.", |
| "links": null |
| }, |
| "BIBREF2": { |
| "ref_id": "b2", |
| "title": "Wh-questions: Unification in a Typed Feature Structure Grammar vs Feature Checking in the Minimalist Program, presented at the Annual Conference of the Israel Association for Theoretical Linguistics", |
| "authors": [ |
| { |
| "first": "D", |
| "middle": [], |
| "last": "Johnson", |
| "suffix": "" |
| }, |
| { |
| "first": "S", |
| "middle": [], |
| "last": "Lappin", |
| "suffix": "" |
| } |
| ], |
| "year": 1996, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Johnson, D. and S. Lappin. 1996b. Wh-questions: Unification in a Typed Feature Structure Grammar vs Feature Checking in the Minimalist Program, presented at the Annual Conference of the Israel Association for Theoretical Linguistics, Hebrew University, Jerusalem.", |
| "links": null |
| }, |
| "BIBREF3": { |
| "ref_id": "b3", |
| "title": "Concepts of Logical Form in Linguistics and Philosophy", |
| "authors": [ |
| { |
| "first": "S", |
| "middle": [], |
| "last": "Lappin", |
| "suffix": "" |
| } |
| ], |
| "year": 1991, |
| "venue": "The Chomskyan Turn", |
| "volume": "", |
| "issue": "", |
| "pages": "300--333", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Lappin, S. 1991. \"Concepts of Logical Form in Linguistics and Philosophy\" in A. Kasher (ed), The Chomskyan Turn, Blackwell, Oxford, pp. 300-333.", |
| "links": null |
| }, |
| "BIBREF4": { |
| "ref_id": "b4", |
| "title": "Quantifier, Wh-Phrases and a Theory of Argument Selection. Unpublished manuscript, Ohio State University. Sag, I. 1996. English Relative Clause Constructions. Unpublished manuscript", |
| "authors": [ |
| { |
| "first": "C", |
| "middle": [], |
| "last": "Pollard", |
| "suffix": "" |
| }, |
| { |
| "first": "E", |
| "middle": [ |
| "J" |
| ], |
| "last": "Yoo", |
| "suffix": "" |
| } |
| ], |
| "year": 1996, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Pollard, C. and E. J. Yoo. 1996. Quantifier, Wh-Phrases and a Theory of Argument Selection. Unpublished manuscript, Ohio State University. Sag, I. 1996. English Relative Clause Constructions. Unpublished manuscript, Stanford University.", |
| "links": null |
| }, |
| "BIBREF5": { |
| "ref_id": "b5", |
| "title": "Not only EVEN, but even ONLY", |
| "authors": [ |
| { |
| "first": "C", |
| "middle": [], |
| "last": "Tancredi", |
| "suffix": "" |
| } |
| ], |
| "year": 1990, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Tancredi, C. 1990. Not only EVEN, but even ONLY, Ms., MIT.", |
| "links": null |
| }, |
| "BIBREF6": { |
| "ref_id": "b6", |
| "title": "Wh-in-situ, Subjacency, and Chain-formation", |
| "authors": [ |
| { |
| "first": "A", |
| "middle": [], |
| "last": "Watanabe", |
| "suffix": "" |
| } |
| ], |
| "year": 1991, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Watanabe, A. 1991. Wh-in-situ, Subjacency, and Chain-formation. Ms., MIT.", |
| "links": null |
| } |
| }, |
| "ref_entries": { |
| "FIGREF0": { |
| "type_str": "figure", |
| "text": "there. The functional head ka, then, optionally carries TO-BINDIQUE:Y, where Y is a subset of X inherited by complement S[INHERIQUE:X]. 4a. SUBCAT <S[(INHERIQUE: X)]> b. (NONLOCALITO-BINDIQUE: Y) c. Condition: Y is a subset of X", |
| "uris": null, |
| "num": null |
| }, |
| "TABREF1": { |
| "type_str": "table", |
| "num": null, |
| "content": "<table/>", |
| "text": "obvious way to recast Lappin and Johnson's treatment in Sag's framework would be to have a type constraint preventing the INHERIQUE feature from appearing on S[fin] in Iraqi Arabic. This could be handled by saying that a finite phrase must have the VFORM fin, and have a null INHERIQUE value. If we then have a rule which allows INHERIQUE:X to be inherited in modified form as INHERILQUE:X, we can handle the Iraqi Arabic facts. The npros in INHERIQUE:X must be inherited as INHERILQUE:X in order to survive inheritance on to a finite phrase. Termination of INHERILQUE is licensed at wh-clauses which have the non-null wh-complementizer sh-, as in 12. The subcategorization requirement that shtakes an S[INHERILQUE:", |
| "html": null |
| }, |
| "TABREF5": { |
| "type_str": "table", |
| "num": null, |
| "content": "<table><tr><td>18. a. Kimi-wa John-ga nani-o tabeta ka shiritai?</td></tr><tr><td>you-top J-nom what-acc ate Q want-to-know</td></tr><tr><td>\"What do you want to know if John ate?\"</td></tr><tr><td>\"Do you want to know what John ate?\"</td></tr><tr><td>b. *Kimi-wa John-go nani-o tabeta shiritai?</td></tr><tr><td>you-top J-nom what-acc ate want-to-know</td></tr><tr><td>In 18a, both</td></tr><tr><td>17. Kimi-wa nani-o taberu?</td></tr><tr><td>you-top what-acc eat</td></tr><tr><td>\"What are you going to eat?\"</td></tr></table>", |
| "text": "In 17, a perfectly natural sentence, there is no Q-particle. It is not possible to both dispense with null complementizers and retain complementizers as functional heads required to terminate INHERIT features. In the present account, by contrast, S is simply licensed as a binder for QUE in Japanese. Therefore, we have a mechanism allowing INHERIQUE to terminate without the complementizer. The full range of facts can be explained by assuming that ka is required on a wh-clause in embedded contexts to satisfy the subcategorization requirements of selecting verbs, but not required in matrix positions because INHERIQUE can terminate perfectly well without the complementizer. Positing an invisible complementizer raises the question of why a verb like shiritai (want to know) may not take a clause without a complementizer. narrow and wide scope is possible for the embedded whexpression, as expected given that there is no constraint blocking INHERIQUE out of ka CP. By contrast, 18b is completely ruled out, as expected on the assumption that shiritai needs a ka CP.", |
| "html": null |
| } |
| } |
| } |
| } |
