No full-text available
To read the full-text of this research,
you can request a copy directly from the author.
Last Resort with Move and Agree in Derivations and Representations
Abstract
Minimalism assumes that language consists of a lexicon and a computational system, with the latter embedded in two performance systems: articulatory-perceptual and conceptual-intentional. Two linguistic levels, phonetic form (PF) and logical form (LF), interface with the performance systems. A computation converges at the interface levels if it contains only legitimate PF and LF objects. However, we cannot define linguistic expressions simply as PF/LF pairs formed by a convergent derivation and satisfying interface conditions. The operations of the computational system which produce linguistic expressions must be optimal, in the sense that they must satisfy some general considerations of simplicity, often referred to as Economy Principles. One of them, the Last Resort Condition (LR), prohibits superfluous steps in a derivation. It requires that every operation apply for a reason. It has often been argued that a similar condition constrains representations, prohibiting superfluous symbols. These conditions require that derivations and representations in some sense be minimal. This article discusses the working of LR, as it applies to both derivations and representations. It starts with the former, examining how LR applies to both Move and Agree. © editorial matter and organization Cedric Boeckx 2011.
... who and what is bought 59 This in itself is quite interesting. We may be dealing here with an economy of representation effect (see [104] and references therein): when both a smaller and a larger structure are in principle available for X, if there is no evidence for the larger structure X is analyzed in terms of the smaller structure. (We would not necessarily expect to find this effect in all languages of this sort since the effect would hold only in an all-else-being-equal scenario, which is not always the case; e.g., lexical properties of elements that are elided on the larger structure option could block the effect-see [92].) ...
... Such examples have important consequences for the debated and unsettled issue of how subject-oriented anaphors should be analyzed. What is important here is that what is located in SpecvP are the individual conjuncts but what is located in SpecTP and what agrees with T is the whole coordination, as the simplified derivation in(104) shows. ...
... (104) [ TP [Pas i i kokoška j ] k su [ XP t k [ ConjP [ vP t i lajali u svom i dvorištu] i ...
The paper examines distributed extraction coordinations, in which different elements move out of conjuncts of a single coordination, as in Which book and which magazine did Mary buy and Amy read respectively, from a crosslinguistic perspective. A number of properties of such coordinations are discussed, which includes showing that they are also subject to the ATB requirement, which will shed light on the nature of the ATB phenomenon itself. It is also shown that there is a rather strong restriction on distributed extractions which confines such extractions to one context and completely excludes one type of movement, in particular head-movement, from participating in them. The higher coordination is shown to be formed during the derivation and to be semantically expletive. Distributed extraction constructions are also shown to have consequences for the proper analysis of a number of phenomena, including subject-oriented anaphors, right node raising, tough-constructions, agreement, and clausal structure. Regarding subject-oriented anaphors, the paper teases apart different approaches to subject-oriented anaphors based on constructions where different elements fill SpecvP and SpecTP (the latter undergoes agreement with T and the former binds subject-oriented anaphors).
... In pursuing this idea, I adopt Bošković's (2007Bošković's ( , 2008aBošković's ( , 2008bBošković's ( , 2011 moving-element driven movement approach, rather than Chomsky's (2000) target-driven approach, which appeals to the following assumptions: (i) The trigger for movement is placed on the moving element, not on the target. 12 (ii) ...
... Suppose that the feature compositions of a wh-C and a wh-phrase in English are like those in (33). In Bošković's (2007Bošković's ( , 2008aBošković's ( , 2008bBošković's ( , 2011 system, the wh-phrase with [uOp] will have to move to SpecCP, which is the closest position c-commanding its licenser, given the Shortest Move requirement. At the SpecCP position, all relevant features are checked off, which in turn inactivates the moving element (i.e., the wh-phrase) at the position, given Chomsky's (2000Chomsky's ( , 2001 Activation Condition. ...
... See Section 4 for a further discussion of this issue. (38) is subsumed under their analysis, it is not obvious how they would generalize a wide range of illicit outputs which have been analyzed in terms of syntactic freezing effects (for related data and additional discussion, see Rizzi (1997Rizzi ( , 2006Rizzi ( , 2014; Epstein (1992); Müller and Sternefeld (1996);Bošković (2007Bošković ( , 2008aBošković ( , 2008bBošković ( , 2011; and references therein). Even within English, it is not immediately clear how examples like (37) (repeated in (41)) and (42) could be accounted for under Epstein et al.'s (2014b) morpho-phonological, CI analysis. ...
Epstein et al. (2014b) suggest, based on the labeling analysis of Chomsky (2013), that there is no NS-specific halting constraint barring syntactic wh-movement from a criterial position, contra Rizzi (2014) and Epstein (1992); that wh-movement is allowed freely in narrow syntax but the ensuing ill-formedness is the result of violating a language-specific morpho-phonological, CI requirements. In this paper, I will argue that NS-specific criterial freezing is real, not a syntactic illusion as claimed in Epstein et al. (2014b), directly affecting both CI and SM interfaces; and show how a wh-in-situ language like Korean can serve as evidence for this claim. I will also show how the system that keeps both the criterial freezing constraint and Chomsky’s (2013) notion of labeling can treat scrambling, which would otherwise be problematic under Chomsky’s (2013) labeling system.
... One is that wh-movement takes place because the Q-feature or the wh-feature with an EPP property attracts a wh-phrase (Chomsky 1995(Chomsky , 2001(Chomsky , 2004(Chomsky , 2008(Chomsky , 2013. The other is that a wh-phrase has an uninterpretable Q-feature and it moves when it cannot delete its uQ-feature in situ (Bošković 2007(Bošković , 2011Zeijlstra 2012;Bjorkman and Zeijlstra 2019). ...
This article explores the possibility of providing a unified explanation for T-movement in English. The major claims of this article are three-fold. First, T can have an uninterpretable feature independently of C. In particular, T has an uninterpretable feature when the denotation of TP is not taken to be true. Second, the uninterpretable feature of T is deleted when T, or its projection, is merged with a constituent that has a matching feature. Third, T undergoes head movement as a last resort when its uninterpretable feature cannot be deleted in situ. This claim is shown to provide a unified account for subject-aux inversion in a variety of English constructions including interrogatives, negatives, optatives, and counterfactuals.
... To be more concrete, (36) shows that irrespective of the surface position of a wh-phrase, the wh-criterial position is always fixed at the agreeing C Wh in Korean. Under this analysis, English and Korean wh-questions are thus assumed to have the reverse feature matrix for a C Wh and a wh-phrase, as represented in (37) With this formulation, Jung (2015a) adopts the Moving-element Driven Movement Approach (Bošković's 2007(Bošković's , 2008a(Bošković's , 2008b(Bošković's , 2011, rather than the Target-driven Approach (Chomsky 2000), which is built upon the following assumptions: (i) The trigger for movement is placed on the moving element, not on the target. (ii) The element with the trigger (i.e., an uninterpretable Operator feature) moves to a position c-commanding its licenser to value the feature. ...
... Based on this and similar kinds of observation, Jung (2015) suggests that these interpretive consequences are a reflex of the syntactic wh-criterial freezing in the sense of Rizzi (1997Rizzi ( , 2006; that in Korean, the wh-criterial position is the position of a C Wh , while in English, it is "the Spec-head agreement position." 13 In implementing this idea, Jung (2015) adopts Bošković's (2007Bošković's ( , 2008aBošković's ( , 2008bBošković's ( , 2011 moving-element driven movement approach, rather than Chomsky's (2000) target-driven approach, which appeals to the following assumptions: (i) The trigger for movement is placed on the moving element, not on the target. (ii) The element with the trigger (i.e., an uninterpretable Operator feature) 13 The term "Criterial Freezing," as originally outlined in Rizzi (1997Rizzi ( , 2006, refers to a mechanism yielding island effects on XPs that moved into a Spec-head agreement position in the left periphery. ...
In the studies of wh-in-situ in Korean/Japanese, grammaticality judgments over wh-island constructions have been notoriously unstable, intricate, and even chaotic. For this prevailing controversy over the Subjacency issue, Kitagawa (2005) claims that the alleged Subjacency effect reported for wh-island constructions in Japanese is a “pseudo-grammatical” phenomenon; that native speakers’ preference over the subordinate wh-scope interpretation in wh-island constructions is a reflection of native speakers’ performance-related biases imposed on the perception and production of wh-questions. In this paper I will argue that Subjacency does exist as a real (i.e., formal) syntactic constraint, serving as a pivotal condition on ruling out illegitimate wh-questions in the narrow syntax; that the confusion and chaos in the judgment of Korean/Japanese wh-island constructions stem from the sound-meaning mismatch in the interpretation of wh-questions at the SM interface, which is the locus of linguistic variation in detectable forms. In so doing, I will show how Jung’s (2015) alternative analysis of wh-constructions can capture the precise nature of wh-licensing and Subjacency controversy in wh-in-situ languages like Korean/Japanese.
... Building on the insight offered by Watanabe (2010), I suggest here that as predicative adjectives cannot have their features valued in situ, they move to a position where their probes can initiate a search for matching features. Although this goes against Chomsky (2000 et seq.), for whom Agree does not trigger syntactic movement, Agree has been claimed to drive movement in other approaches (cf., a.o., Bošković 2007Bošković , 2011Epstein and Seely 2006). I suggest that predicate-argument agreement in the presence of GQ in the subject in Polish provides some support for the latter view. ...
This paper offers a minimalist account of the distribution of the features of case, number and gender in Polish phrases with cardinal numerals and vague quantifiers (Q-NPs) in the presence and absence of genitive of quantification (GQ) based on the assumption that GQ-cardinal numerals and vague quantifiers are phrasal and are merged as adjuncts to the projection of a null functional head (F), a phase head that values its complement with GQ in contexts in which GQ is not overridden by inherent theta-related or lexical case. Whether the null head F inserted in the derivation has unvalued features of number and gender has consequences for the syntactic computation of agreement relations, in which features may be shared. Heterogeneous case distribution in Q-NPs in nominative-accusative case contexts provides evidence that F is a phase head and triggers transfer of its complement. Variable agreement patterns of both attributive adjectival modifiers and adjectival predicates in the presence of GQ are argued here to reflect attempted-but-failed syntactic agreement, which does not lead to ungrammaticality, as some features unvalued in narrow syntax can still be interpreted by morphological realization rules in PF. GQ is shown here to provide evidence that predicative adjectives cannot have their φ-features valued in situ and must be remerged in a position where their φ-probes can initiate Search for matching features. Agreement in the presence of GQ in Polish thus demonstrates that the need of features to be valued drives syntactic movement.
... This principle was proposed in the literature as a constraint on syntactic derivations, and recent discussion has considered the claim that this principle may be relevant in computing processing cost of syntactic structures (see Enzinna 2013;Enzinna and Thompson 2013). According to the Principle of Economy of Representation (see Law 1991), formalized as the Minimal Structure Principle (MSP) (see Bošković 2011;Heck and Muller 2011), if two representations have the same lexical structure and serve the same function, once provided lexical requirements of relevant elements are satisfied, then the representation that has fewer projections is to be chosen as the syntactic representation serving that function. Taking into consideration results of recent research (Kang and Thompson 2013), the present experiment has been constructed by utilizing a counterbalanced self-paced reading task. ...
This work concerns the on-line processing of the Conditional Inversion Construction (CI) in English. According to the traditional approach to the syntax of Conditional Inversion, this construction involves I-to-C Movement (Iatridou and Embick 1994), identical to the movement involved in Subject-Auxiliary Inversion. However, recent analyses of Conditional Inversion claim that this construction involves raising of I-to-C-to-X, the head of an XP projected above CP (Bjorkman 2011). Assuming that these steps of movement are costly processing-wise, the traditional approach predicts that Conditional Inversion and Subject Auxiliary Inversion are equally costly with regard to their non-inverted structures, while the contemporary analysis predicts that Conditional Inversion is more costly than Subject Auxiliary Inversion with regard to their non-inverted structures. In order to investigate these predictions, we conduct a Self-Paced Reading Task experiment with seven monolingual-English speakers. Subjects are presented with one hundred sentences: fifty experimental and fifty fillers; half of the experimental sentences include CI and the other half includes SAI Constructions. After half of the total sentences, subjects are presented with a comprehension question. Reading times for each sentence are recorded. The result of a paired samples t-test revealed a significant difference in reading times between sentences with CI (M=64.95, SD=22.27) and SAI (M=34.28, SD=5.83), t(7)=4.821, p<0.003. For reading times per character, the difference is .003, which is significant. Assuming that " reading times are reflective of processing difficulty " (Levelt 1989; Rayner and Sereno 1994), the shorter processing time involved in computing SAI as contrasted with CI provides support for the claim that CI involves a more complex operation, which is predicted by Bjorkman's I-to-C-to-X Analysis.
The paper examines wh&wh coordinations (wh&wh), arguing that what makes them possible in a language is not the availability of multiple wh-fronting (MWF), as previously assumed, but the availability of an indeterminate system. Such coordinations are argued to involve derivational coordination formation via sideward merger where the affected elements undergo movement before coordination formation, sideward merger being disallowed from interpreted positions. The derivationally formed coordination undergoes movement upon insertion into the structure. The account captures a clause-mate requirement and ordering constraints on elements involved in wh&wh, as well as a single-pair reading restriction. A typological prediction is that since both MWF and wh&wh depend on the availability of an indeterminate system, but in different ways, there should be a good deal of overlap between MWF languages and languages allowing wh&wh but there should still be languages that allow wh&wh but not MWF, which is borne out.
This paper discusses the concept of weak head in Chomsky’s (2015. Problems of projection: Extensions. In Elisa Di Domenico, Cornelia Hamann & Simona Matteini (eds.), Structures, strategies and beyond: Studies in honour of Adriana Belletti , 3–16. Amsterdam: John Benjamins) sense from a typological perspective. This paper first establishes a novel generalization that large-scale pied-piping is available in a language only if the language has indeterminate pronouns in Kuroda’s (1965. Generative grammatical studies in the Japanese language . Cambridge, MA: Massachusetts Institute of Technology dissertation) sense and the clause to be pied-piped is head-final. To deduce this generalization, this paper first offers a labeling theoretic account of large-scale pied-piping. It then proposes that weak heads are (i) heads that have unvalued features and (ii) morpho-phonologically weak. It is shown that the generalization is deduced from this conception of weak heads, given Inaba’s (2011. The morphosyntax of constituent ordering patterns. The Hiroshima University Studies, Graduate School of Letters 71. 43–72) generalization that head-final complementizers are generally affixal. This paper further argues that the proposed conception of weak heads also allows us to deduce Agree from Minimal Search, which is a third factor principle according to Chomsky (2013. Problems of projection. Lingua 130. 33–49), hence minimizes UG.
This paper is consecrated to analyzing the imperative sentences in French language. We demonstrated that pro-element is generated in the extern agreement position of the verb [spec, VP]. We affirm that the complementizer projection (CP) is a superior projection of the imperative sentences. We also demonstrated that the movement verb in imperative tense must be moved in the syntactic position [C, CP], if there is no barrier-head, to checking the uninterpretable feature [uImperatif]. We concluded that the distribution of clitics pronouns in the imperative structure is not identical to the declarative sentences.
U ovom se radu analizira struktura glagolske sintagme u bosanskom jeziku te se nastoji odrediti bazni položaj upravnih članova sintagme, njihovih argumenata, ali i adjunkta. Ponašanje strukturnih elemenata koji čine glagolsku sintagmu u bosanskom jeziku i mogućnost njihovog pomjeranja prati se u procesu deriviranja površinske strukture rečenice koja sadrži elidirano mjesto. Različita ostvarenja elipse objašnjavaju se jedinstvenim položajem pojedinih strukturnih elemenata i njihovom mogućnošću da projiciraju zasebne maksimalne projekcije koje, u procesu deriviranja površinske strukture rečenice, mogu biti ili obuhvaćene elipsom ili izbjeći elidiranje. U vezi s tim, u radu se ukazuje na različito ponašanje pojedinih vrsta glagola u procesu VP elipse u bosanskom jeziku, te se uspostavlja paralela s istim strukturnim elementima u engleskom jeziku. U radu se analiza navedenih pojava zasniva na principima i parametrima generativnegramatike.
The chapter is concerned with grammar and lexis as core components of human language. Its main aim is to elaborate on the interrelation between the two components as discussed in a number of specific formal and functional linguistic theories. From the former group, the issue is traced in such models as Government & Binding (G&B), the Minimalist Program (MP), Distributed Morphology (DM), Head-Driven Phrase Structure Grammar (HPSG) and Lexical-Functional Grammar (LFG) from the latter - in Dik's and Halliday's functional models and the usage-based models by Langacker and Goldberg.
This paper investigates the acquisition of a new uninterpretable feature (uF) under the
circumstance that support from an interpretable feature (iF) counterpart is not available, through the study of L2 acquisition of Quantifier Raising (QR) by Japanese learners of English. QR is triggered by the uQUANT feature, which does not have the interpretable counterpart (i.e., iQUANT) (cf. Bruening, 2001). While this feature is active in English, it is missing in Japanese. Thus, Japanese learners of English need to acquire the new uF in order to yield QR. The results of the acceptability judgment task suggest that intermediate learners’ grammar lacks the uQUANT feature in accord with the Interpretability Hypothesis (IH) of Tsimpli (2003) and Hawkins and Hattori (2006).
The aim of the present research is to contrast the degree of grammaticalization of factitive operators in Romanian, French, Spanish and Italian. While Fr. faire, Sp. hacer and It. fare function as free morphemes, i.e., grammatical markers, Rom. a face and its contextual synonyms (a da, a cǎuta, a pune, a lǎsa) do not display morphological traces of grammaticalization, but they lack semantic and, partly, syntactic autonomy.
This paper considers the genesis of Case markers, which constitute a new functional category (K(case)) hitherto unanalysed in formal (Minimalist) analyses of grammaticalization.
In Xhosa VSO clauses, subject agreement exhibits default features, objects cannot be pronominalized, a subject focus reading is obligatory, and experiencer verbs with two DP arguments are precluded. We argue that impoverished versions of T and v* in VSO clauses lack the probe features involved in subject agreement, EPP, object shift, and nominative/accusative valuation within Xhosa SVO sentences. Only an unusual focus-linked strategy can Case-license full DPs in VSO clauses, but this is incompatible with inherent Cases borne by arguments of experiencer verbs. We show that CPs and augmentless NPs appear in positions where DPs cannot surface because uCase is a feature of D. Given the striking evidence for abstract Case in Xhosa, we propose Case-friendly analyses for Bantu Case-theoretic anomalies that Xhosa shares.
We investigate the nature of ergativity through Basque. In Basque, ergative case and agreement reflect structural rather than inherent Case: Agree/Move rather than thematic or other c/s-selectional dependencies. Evidence comes from the core distinctions between these dependency types, including ergative-absolutive alternations due to raising-to-ergative of internal arguments and absolutive Exceptional Case Marking of external arguments. Thus structural Case systems cannot be reduced to a nominative-accusative basis with an inherent ergative, as recent work proposes (Woolford 2006, Legate 2008). In turn, our investigation clarifies the nature of structural ergativity. First, ergativity, like nominativity, comes from the supra-vP T-system, whereas absolutivity and accusativity are lower. Second, ergative agreement can occur under unbounded c-command through Agree, like nominative and accusative case and agreement, but ergative case assignment requires movement to Spec,T, bearing out the ergative as a 'marked' structural Case. Third, structural Agree/Case systems show not only ergative and accusative alignments, but also arbitrary islands of each pattern within the other, calling for suitable parametric tools. We develop a theoretical account of these results in the current Principles-and-Parameters framework (Chomsky 2000, 2008), building on theories of structural ergativity (Laka 2000, Bobaljik 1993, Bittner and Hale 1996).
In Lubukusu and Lusaamia, the wh-expression ‘how’ agrees in φ-features with the subject of its clause. We show that agreement on ‘how’ is not always identical to subject agreement on the verb: the two diverge in certain locative inversion and subject extraction environments. On the basis of these facts, we argue that ‘how’ is a vP adjunct with downward-probing uφ independent of the uφ that underlies subject agreement. We also explore locality paradoxes that arise in connection with agreeing ‘how’ in locative inversion constructions. These present challenges to the traditional notion of equidistance from a probe as an explanation for inversion, show that operators may have ‘‘active’’ φ-features even while they are Ā-opaque, and offer insight into the mechanisms making locative inversion possible.
This paper closely examines two types of predicate contrastive topic constructions in Korean, i.e. Ha-construction and R-construction, and argues that they are derived from the same underlying structure via the same process, which involves a phrasal movement including a subject, leaving behind the copy of the moved constituent (cf. Chomsky, 1995), in contrast to the head movement analysis. That is, as far as syntactic process is concerned, Ha- and R-constructions are one and the same construction. Two constructions and their surface variations arise as a result of selective PF-deletion process in the lower copy of the moved constituent. The analysis proposed in this paper provides a uniform analysis of two types of predicate contrastive topic constructions in Korean and also provides a systematic account of both similarities and differences with regard to morpho-syntactic and semantic properties found in two construction types. To the extent the current proposal holds, it provides strong empirical evidence for copy theory of movement and late lexical insertion along the lines of Distributed Morphology (cf. Halle and Marantz, 1993). Furthermore, it constitutes strong empirical evidence for the assumption that verbal inflections as well as verbal roots are independently projected to syntactic structure in Korean.
Chapter 5 develops the last-resort mechanism ℜ, through a cross-linguistic study of the repairs of person constraints by the
emergence of otherwise unavailable ergatives, accusatives, PPs, and enriched DPs. They are unified as the minimal enrichments
of a non-convergent syntactic structure by an Agree/Case dependency, that is, by an uninterpretable phi-feature (probe). ℜ
thus extends the hypothesis that uninterpretable features create syntactic dependencies to meet Full Interpretation, from
features that are lexically fixed, to those that are dynamically inserted for this reason. The point of departure is Chomsky's
(1995 et seq.) proposal that syntactic content can be licensed by an 'effect on output', confined by a strong modular architecture
that restricts its scope as an interface mechanism. ℜ can detect illegibility at the interfaces of syntax with PF and LF,
but not problems that arise within these modules. In response, it can enrich the numeration interface between syntax and the
lexicon with an uninterpretable feature, but cannot modify syntactic computation, nor search the lexicon for interpretable
content. ℜ is extended to other Full Interpretation failures, notably the ergative and accusative 'dependent Case' of all
transitives, construed as a response to Case licensing.
ResearchGate has not been able to resolve any references for this publication.