Specifically Chomsky b: 8 makes explicit the relation between language variation and the faculty of language as a biological system. On the basis of these conceptual premises we hardly expect variation to take the form of the broad generalizations entertained by the typological tradition, which takes as its starting point not coincidentally functionalist hence fundamentally behaviorist conceptions of language.
Nor do we expect that the same types of generalizations admit of a restatement within the mentalist model, say, in the form of macroparameters. It seems to us important that studies such as this provide what in our mind is compelling evidence in favor of the view that language variation results from the free interplay of elementary differences connected to the mental lexicon. If this conclusion is correct, it contributes a strong argument in favor of the biolinguistic perspective, which is the only model capable of predicting such variation. For external reasons, i.
Thus Italian dialects provide a rich and articulated picture of language variation that contrasts to some extent with that of other intensively studied linguistic varieties, say, French or English.
We should emphasize however that in our view the Italian situation reflects closely the kind of variation we expect to find in conditions of normal language growth and that the present theory of Universal Grammar and parameters predicts, in keeping 12 Introduction with the biolinguistic program. This means that it is the linguistic situation of, say, England or France that represents a somehow misleading picture of variation, reflecting not only the action of the internal shaping forces of language growth but also external mechanisms of social and political standardization.old.videovolunteers.org/map.php
Generative Grammar: Theory and its History (Routledge Leading Linguists)
The variation presented by Albanian dialects is compatible with that of Italian dialects. In this case our basic sample is more limited, including essentially a couple of varieties in Albania itself covering the major Geg vs. The latter witness a subtle and systematic variation in the morphosyntactic organization of the sentence for instance, in the verb inflection and in the case assignment system that has the same general characters as the microvariation observed in Romance dialects and may be explained in the same terms, essentially of lexical variation.
On several occasions in the course of this work we will also have the opportunity to treat variation characterizing not just two or more different dialects linguistic communities but showing up within the same dialect, or even within the productions of a single speaker. In accordance with the minimalist model, which predicts the absence of free alternations within any given grammar, we shall treat the relevant cases as revealing the simultaneous presence of slightly different lexicons hence grammars within the same speaker s.
This amounts to saying that, strictly speaking, there are no monolingual individuals, given that as we mentioned above each speaker will alternate at least between so-called stylistic choices according to situations of use. For instance, any cultivated Italian speaker of northern Italy will typically alternate a grammar inclusive of a simple perfective past used only in writing or comparable registers with a grammar where the present perfect covers the meaning of the simple perfective past as well.
Categorie funzionali del nome e del verbo — Introduction 13 Special thanks go to all our informants, both on Romance and on Albanian, though reasons of space prevent us from mentioning all of them here. Our debt to the teachers, friends and colleagues whose work inspired us should be obvious from the references.
However, we should take this opportunity to thank at least Carlos Otero, as the series editor, for helping our project along. Thanks also to Alan Pona for putting together the analytical index. The debt that this book owes to previously published work of the authors is acknowledged at the beginning of this Introduction.
Download Limit Exceeded
Rizzi ed. Earlier versions of chapters 1 and 3 appear respectively as Manzini, M. Manzini and L. Savoia, and as Manzini, M. An earlier version of chapter 8 appears in Italian as Manzini, M. Cinque secoli di cultura albanese in Sicilia.
Mirror, pp. This proposal is supported by several sets of data presented in section 2. The model we propose requires a different theory of agreement and chains, also discussed in section 3.
V corresponds to the predicative content of the event. It is again conventional to assume that the V position projects a set of arguments, including at least the object and the subject. We take it that the crucial property of the subject is a denotational property, which we provisionally notate D. In other words we identify the EPP property, which defines the subject, with the D property, an intuition which we share with Chomsky The so-called object essentially corresponds to the point of saturation of the obligatory internal argument of the predicate.
We take it that this property may be characterized by the category N, a label to which we return below. Therefore, connecting the V predicate with its N object and D subject yields a tree structure of the type in 1. Thus consider a finite sentence. Its structure minimally includes 2b , where the verb is in I.
- Towards an Elegant Syntax.
- Freely available.
- 100 Things Every Man Should Know.
This is also the case in languages with a considerable amount of verbal inflections, such as northern Italian dialects, where in declarative sentences the subject is lexicalized at least by a socalled subject clitic, as in 3. This holds for I, V and also for C, giving rise to a sentential schema of the type in 4.
Consider a finite verb, whose inflection alone represents the subject in a language such as Italian, as in 2a , yielding a case of so-called null subject. It is widely accepted in the literature that the inflection in a null subject language is pronominal Rizzi and that it alone suffices to satisfy the EPP requirement Pollock Our proposal represents a development of this line of thought, with some important differences. In the first place we adopt the conclusions of the introduction that morphological structures are identical to syntactic structures.
In the traditional conception, reflected by generative theory and by the minimalist model Chomsky , , , syntactic categories and morphological features represent two different subsystems. In particular the traditional conception of feature distinguishes the feature itself from its value, which is binary Jakobson ; Chomsky and Halle To begin with, we can observe that the distinction between feature and value may be dispensed with. Thus instead of saying that there is a number feature whose value can be positive plural or negative singular , we can say that the number feature identifies with the plural.
As a consequence, the notion of default and markedness is excluded from our grammar; for example, if number coincides with the plural, it is not possible to treat the singular as a default value of the feature, but only as its absence. If so, a morphological feature comes to have the same formal nature as a syntactic category. For Romance dialects, in Chapters 3—4 we reach the conclusion that the case, person, number and gender features generally employed in the description of pronominal clitics are inadequate. Our proposal is instead that clitics can be adequately described in terms of syntactic categories, namely D, already introduced above for subject clitics, and several categories for object clitics.
These include R referentiality for properties of specific quantification; Q for properties of indefinite quantification; P for person, i. Again N has been introduced above for objects; the independent categorization of P elements makes it clear that the N category is in fact restricted to the third person. The categories mentioned above coincide with those independently postulated by many theories for the internal structure of the noun phrase.
In other words, they do not constitute a separate list of morphological categories, but they are the same as syntactic categories. Given this identification of morphological and syntactic categorization it is natural to assume that morphological structure, understood descriptively as the structure of words, is articulated into the same hierarchies as syntactic structure.
The unification of morphological and syntactic structures can be crucially Agreement inflections of the verb 17 applied to verb inflection.
Linguistics 14 by Routledge - Issuu
The traditional and generative approach to verb inflection and to its role in the structure of the sentence is well illustrated by the theory of Chomsky , , according to which the verb inflection consists of an unspecified set of features associated to the verb in I, whose value is set by the corresponding, valued features of the sentential subject. The unification of morphology and syntax that we adopt here leads however to the conclusion that the verb inflection has exactly the same status as a pronominal subject, in particular the clitic subject in 5.
Thus we take the verb inflection to be categorizable as D. What is more, we assume that it is inserted in a D position within a morphological structure which reproduces exactly the syntactic structure of the sentence. Concretely, for a verb like the one in 2 , we propose the analysis in 8a , where the verbal root corr-, corresponding to the predicative content of the verb, is generated in I while the inflection occupies the D position. Crucially the structure in 8a is to be understood as word-internal, while a structure like 7 , which most directly compares with it, is to be understood as syntactic in the conventional sense of the term.
The two combine as indicated in 8b. In essence, the verb in the sentential I position takes a D subject to its right; in turn the verb in the sentential I position has a structure that strictly parallels that of the sentence, with the verb root in I and its inflection in the D position to its right. This property of distributed morphology holds of the current system as well; but in other respects the two models differ. First, Halle and Marantz , distinguish two sets of categories for morphology and syntax, forcing the two components to be separate in turn.
On the contrary, in the current model there is a unique set of categories and structures and therefore a completely integrated morphosyntactic component. We shall devote this chapter to providing empirical evidence in favor of this conclusion, as it concerns the internal structure of the verb and of the sentence. At the same time it seems to us that it gives rise to a simpler grammar than the one envisaged by Halle and Marantz , Indeed to the extent that their morphological and syntactic components largely overlap, there is considerable redundancy between the two, which is absent from our theory.
Late Insertion is furthermore governed by the underspecification principle, according to which a lexical element may be inserted under a terminal node only if its features are a subset of the features of the terminal node itself. For cases in which this condition is apparently violated, Halle and Marantz , posit a rule specific to the morphological component, namely impoverishment, which deletes features from a terminal node. In this respect Distributed Morphology has the same power as do theories employing ranking of constraints, such as Optimality Theory.
Our model is based instead on the minimalist idea that linguistic structures are projected from lexical material, thus excluding the existence of a morphosyntactic structure independent of the lexical items that it will eventually host, and a fortiori the existence of readjustment rules such as impoverishment. It seems to us that such a model is once again more restrictive in that it does not have the power of extrinsic ordering provided by constraint ranking in Optimality Theory and by readjustment rules in Distributed Morphology.
In introducing the syntactic representation in 8 for a sentence such as 2 , we noted that the idea that the verb inflection has a pronominal status is adopted by much current literature for null subject languages, though without an explicit formalization. The treatment we propose for verb inflection however is meant to be universal and not bound to the null subject status of languages such as Italian.