Since Chomsky rules out checking from adjoined positions, XP-adjunction can never satisfy Last Resort, and should therefore never take place. It is within this core component of the language faculty that we find the striking properties highlighted by minimalist guidelines. It seems increasingly reasonable to distinguish this component of the language faculty.
We return to the interpretative issues directly. A step in the direction of structuring the PF-component would be to assume that the PF-rules that reorder parts of the strings generated by the computational component, employ mechanisms similar to those in the overt component. That is, at least some PF-displacement rules too can be argued to be driven by the requirements of feature-checking, although presumably without the formation of checking domains, since the only option for Move in this component is adjunc- tion.
In this approach, the only function of the overt syntax is to ensure that these morphosyntactic features are licensed so that they are visible at the PF-interface. We suggest that this function is not exclusively the property of the overt syntax, and that feature-checking in the PF-component also shares the same function — to render morphological features visible to Morphology — a guiding intuition of the minimalist program. The question now is, what is the nature and content of features that can only be checked in the PF-component, since the standard notions of feature-checking cannot accommodate checking in this component, and it is to this question that we now turn.
Can it be argued to be the morphological property of a lexical item in a way analogous to Case, categorial features, etc. In all likelihood not, for, as Culicover 5 puts it: Focus is orthogonal to most other aspects of interpretation, in that it can vary independently of the literal sense of individual phrases and the compositional properties of the interpretation. So the book literally means the same thing whether or not it is in focus and read the book literally means the same thing whether or not the book is in focus.
Focusing is largely a pragmatic strategy, and especially the interpretation of non-wh focus is context-dependent and constrained by a number of discourse factors. For example, in Hindi-Urdu and Malayalam, positional focus construc- tions are not very good discourse initiators, since they presuppose discourse. It is nevertheless part of the grammar as it is the domain that evaluates the felicity of what otherwise seem to be optional rules of a particular grammar. Indeed, this seems plausible when one recalls fact that languages usually have more than one way of marking this feature.
And since we have claimed that PF-rules themselves are driven by the need to make linguistic items visible to Morphology, feature-checking in wh-focus in these languages actually has to be satisfied by the time the derivation reaches Mor- phology.
Chomsky — proposes that the lexical entry of a linguistic item consists of its intrinsic semantic features, such as categorial features, Case- assigning features, gender, etc. The operation that forms the numeration adds on the optional features that come into play in the derivation — features of number, Case, tense, etc. Additionally, this requirement may originate from the role that wh-phrases play in Domain D, functioning more or less as place-holders for the assertion.
It appears that focusing pertains more to the realm of the verbal paradigm than to the nominal one, and is in fact more predicational, and rather like mood, or choice of tense and aspect, i. Adjacency is sensitive, then, only to those elements that are relevant to the mapping process. Headedness is relevant linearization , while traces and empty projections are irrelevant PF-deletion.
UG Requirement on Morphology Features must be licensed. The Adjacency Condition Features can be licensed only under adjacency. We shall assume that the adjacency requirement in Morphology follows from the general poverty of the component itself — as it lacks the operation Move, it can only operate on strings that it can see as linear.
Of the possible configurations in 15 , only 15b provides an appropriate environment for adjacency, as traces are phonetically empty material. PF-feature checking movement of the type we are considering here is therefore motivated solely to satisfy this adjacency requirement, which may vary further in its specific formulation across particular languages. This feature-licensing is Morphology-driven. Adjacency between it and the focused category is required by the fact that Morphology can only operate on strings that it can see as linear, since it lacks the operation Move.
The feature is an intrinsic one of wh-foci, and thus it can be accessed for interpretation both at LF as well as Domain D.
Since Hindi-Urdu has overt verb agreement with the direct object, it is reasonable to assume that v licenses multiple specifiers at Spell-Out. This requirement of adjacency is, however, not fulfilled in the structure at Spell-Out, since the shifted DO inter- venes. Consequently, DO scrambling, expressed as left-adjunction to TP, takes place to yield adjacency between the verb and the subject. Hence, the feature can also be licensed by Morphology in 13a and Phonology in 13b and by all three strategies in 10 above.
Indeed for some speakers, preverbal focusing is indeed possible with rightward scram- bling, though not preferred. For example, in an example like 6a , is it is unambiguous to interlocutors that the scrambled DP is presupposed, specific and topical, but in 6d , none of these facts seem to be quite so clear. We suggest that this preference for leftward scrambling, i.
Годишњи број навода
This PVP measure classifies all those movements as optional and free which proceed in the canonical direction of the language. Now, Hindi-Urdu has a head-final setting of the Head Parameter, with the result that leftward movement is free. Rightward movement can, in this system, only take place, if it is needed for conver- gence, and hence the preference for leftward scrambling amongst native speakers.
In the discussion on 8 and 9 , we demonstrated that focused XPs in Hindi-Urdu are more amenable to coreference. Following Erteschik-Shir , we suggest that this coreference is favored in scrambled constructions because of the fact that scrambled XPs tend to be interpreted as topics. Kidwai shows that even the putative reflexive binding in 8b does not follow from the Binding Theory, but rather from the fact that in Hindi-Urdu monomorphemic reflexives may, in certain discourse contexts, be used referentially. Since these coreference possibilities are heavily discourse-conditioned, we propose that coreference is determined in Domain D.
Since it must survive to the interface even when licensed, the fact that discourse-initial questions in Hindi-Urdu may fail Morphology does not yield non-convergence at the interface. It is therefore expected that in all instances of DO-focus, the structure at Spell- Out should by default meet the licensing requirements on focused phrases, thereby precluding the necessity of PF-movement in these cases.
PF-movement in the form of verb-raising to C0 would then take place to establish the relevant licensing domain for the wh-phrase.
We suggest a fairly literal interpretation is in order, by which anything which is a focus cannot be a subject. The question is that why such PF verb-raising is required at all, since in the order at Spell-Out in 22b no material relevant to the blocking of adjacency intervenes between the subject and the verb substituted into v. We suggest that languages are parametrized with respect to what is considered the correct direction of adjacency by the Morphology of that language, and that this parametrization is sensitive to settings of the Head Parameter, as first suggested by Horvath Tangale represents a case of further parametrization of the definition of adjacency.
As mentioned earlier, Tangale does not allow focused constituents to intervene between the verb and the direct object, as shown in 23 : 23 a. If Tangale were exactly like Western Bade, we would expect the PF verb-raising mechanisms to T0 to license subject focus to necessarily yield the order VSO, but since the language has overt object shift, the observed surface order is VOS. While in the last object-shift language we looked at the shifted object was found to be a barrier for adjacency, clearly this is not the case in Tangale. A consequence of these proposals is that we expect DOs, IOs and adjuncts in Tangale to be focused in situ by verb-raising to T0.
The Parametrization of Universal Grammar
In these two languages, directionality is relevant, whereas in Hindi-Urdu and Malayalam, those requirements are met by default, since they are head-final languages. The choice of the type of adjunction XP or X0 is a language-specific choice, since at least Hindi-Urdu appears to prefer XP-movement for focusing rather than X0-raising. In the analysis of wh-movement in Chomsky , the structure that exits Spell-Out of a query like 24 could be either 25a or 25b , but never 25c , since the [strong] Q-feature need only be satisfied once for convergence: 24 [CP Q [IP you will read what]] 25 a.
Of the three structures in 26 , the structure that cannot exit PF, 25c , is the only legitimate PF-output. The proposals we make regarding the licensing of wh-phrases in languages without wh-movement to [Spec,CP] carry over quite nicely to explain why this happens. A yes-no question, while it satisfies the Q-feature of a clause, does not have the same type of consequences for the discourse as the wh-question, since only the latter creates a presuppo- sition-assertion configuration. PF- movement of the wh-phrase to adjoin to CP substitution no longer being an option is ruled out by the PVP-measure, leftward movement being too expensive in a head-initial language.
Verb-raising at PF, an option in English, is therefore necessary to license the wh-phrase, for it to be interpreted as a question. Note that directionality of adjunction does not appear to be an issue in English, at least as far as auxiliaries are concerned. Although their claims about VP-topicalization seem to be on the right track, the proposals regarding verb-raising do not seem to be correct. Nor can it explain why main verb-raising in the presence of an auxiliary is good, or at least okay, in these contexts 27a—b , but terrible in all other cases of V-to-I, viz.
On analogy with our interpretation of the Extended Projection Principle for Tangale, let us assume that something which is a focus cannot simply raise to [Spec,TP] in the overt syntax. The auxiliary, in all the examples in 27 , we shall assume to be in a projection higher than the subject, and hence it undergoes no movement. If these claims are true, and main V-raising in English is a PF-movement rule for focus-licensing, then the impossibility of main verb raising in questions follows — main verb-raising will never satisfy the adjacency requirement.
As we have seen in this paper, current minimalist approaches open up avenues of inquiry into focus in natural language, the results of which lead to a re-appraisal of the very hypothesized shape of UG itself. Needless to say, the views held in this paper are not necessarily shared by any of them, and the mistakes and oversights are quite definitely my own. Josef Bayer personal communication points out the necessity of spelling out the fact that the notion of presupposition relevant to truth is SPEAKER presupposition rather than logical presupposition.
This, as Zubizarreta 3—4 observes, explains why the logically presupposed complements of factive predicates can be focused. Context-induced contrastive and presentational interpretations are not typical to foci alone, as the interpretations a topic may receive are identically constrained by context — if a contrast set of topics is available, the topic will be interpreted contrastively: A: Tell me about Ram and Sita.
Do you like them? These observations hold for morphological topics as well. Miyagawa and Kidwai trace the contrastive interpretation accorded to -wa topics in Japanese and -to topics in Hindi- Urdu respectively, to the role of the context-set provided by the discourse. Focus is indicated by upper case. Hindi-Urdu also exhibits the phenomenon of long-scrambling of arguments. While long- scrambling out of finite complement clauses is accepted by only a minority of speakers e. Mahajan , speakers are unanimous about the acceptability of long-scrambling out of non- finite complement clauses.
The facts of preverbal focusing in these configurations are quite complex, but appear to center around the basic generalization that long-scrambling of an XP licenses preverbal focusing only of a constituent of the clause of origin of the scrambled XP itself. We shall have little to say about these constructions in this paper. We will assume that the analysis of Hindi-Urdu scrambling as XP-adjunction that we provide below carries over to Malayalam.
This is not, however, to be taken to imply that FP analyses ignore the question altogether.
Brody suggests that positional focus languages have overt movement, triggered by the [strong] features of F0, to a universally available FP projection and that languages that apparently do not, actually exhibit covert raising to [Spec,FP] at LF. This enables him to unify what are, intuitively, very distinct types of focus — the focusing involved in association with focusing particles, which we refer to as BOUND FOCUS, appears to originate in the set of grammatical conditions imposed upon an operator like only — only must be associated with a focused element in its c-command — whereas positional focusing, which we refer to as FREE FOCUS, derives its existence from speaker intentions and the properties of wh-phrases, but not from the immediate grammatical environment.
In free focus, on the other hand, a contrastive focus interpretation of the focused element is not obligatory, and even when it does obtain it is not induced by a grammatical operator analogous to those like even and only. The use of word-stress for contrastive focus is however an option in the grammar.
It turns out that besides the adjacency facts and the fact that it is a feature that appears to be checked by a verbal projection, there is very little actual support for this characterization of focus — it does not seem to be the property of particular types of predicates, the morphological form of foci and verbal morphology is invariant across all GF-foci. Josef Bayer personal communication points out that our characterization of focus as pragmatic and hence evaluated in terms of speaker-presuppositions at Domain D attributes too much to the PF-interface to really distinguish it from the LF-interface, as we conceive of it currently.
While this is true, at least part of the research agenda of the approach suggested here would be to restrict interpretations at LF to core semantic properties of natural language by transferring the burden of discourse-dependent interpretations to Domain D. This is by no means a trivial task, and further raises the question as to whether there can be Domain D and LF interactions, and how these are to be captured. An obvious problem with this proposal is the problem of wh-in-situ in languages with overt wh-movement.
In this context, note that languages like German tend to locate wh-in-situ items left- adjacent to the verb Josef Bayer, personal communication : i a. The data in ii could be explained by considering the PP nach Hause to be part of the lexical meaning of nach Hause fahren, hence the adjacency condition is met by considering the complex predicate to be the relevant category licensing the in-situ wh- phrase. This conclusion, however, is hasty, as there are many languages, e. Hindi-Urdu, which also allow PP and NP constituents of what are clearly complex predicates to be extracted.
Furthermore, the fact that a constituent can be topicalized does not necessarily entail that it may be scrambled the operation that enables positional focusing. With this in mind, consider the contrast between i and ii. In ia what is clearly involved is PP-scrambling, i. The fact that the preverbal focusing requirement is not very strong in German witness the relative acceptability of ia , however, then puts the question back into the domain of language-particular enforcement of Domain D requirements.
Chomsky suggests that it may well be that focus interpretation takes place when various conditions like adjacency obtain, and Lecarme suggests the proper characterization of focus to be a post-Spell-Out feature adjoined in the PF- component.
Similar accounts of DO and adjunct focus can be given, assuming that IOs are generated as vP adjuncts in some languages without the dative alternation Kidwai and with the order IO-DO — the DO must be scrambled out of the way for adjacency between the licenser and licensed. However, there is a problem here, since Tuller, citing Kenstowicz , provides evidence to show that tone sandhi processes in the language argue for a distinction between the syntactic position occupied by a focused DO versus a nonfocused one. Significantly, 32 involves cases of paired contrastive focusing, with the topicalized VP bearing narrow focus as well.
Topicalization in English has often been noticed to involve narrow focus. Although the question why all topicalization in English does not require verb-raising at PF must remain unanswered. This volume. Bobaljik, Jonathan. In Heidi Harley and Colin Phillips eds. Brody, Michael. Chomsky, Noam. Lectures on Government and Binding. Dordrecht: Foris. You know that reading The Parametrization Of Universal Grammar Fanselow Gisbert is effective, because we are able to get information from the resources. We can read books on our mobile, tablets and Kindle, etc.
Eric Reuland - Наводи Google академика
Hence, there are lots of books being received by PDF format. Listed below are some websites for downloading free PDF books where you can acquire just as much knowledge as you wish. Nowadays everybody, young and old, should familiarize themselves with the growing eBook industry. Ebooks and eBook viewers provide substantial benefits above traditional reading. Ebooks cut down on the employ of paper, as advocated by environmental enthusiasts.
Presently there are no fixed timings for study. There will be no question of waiting-time for new editions. Presently there is no transportation to be able to the eBook shop. Typically the books in a eBook store can be downloaded quickly, sometimes for free, sometimes for any fee.