Clitic metathesis in the Friulian dialect of Forni di Sotto

Abstract

In this article we entertain the hypothesis that cliticization involves a rule of m-merge, which brackets a functional head with another constituent under linear adjacency to build a structure legible at the PF interface.We therefore argue for a division of labour between syntax and morphology in the spirit of Halle and Marantz (1993), although we depart from their model in rejecting a single post-syntactic Morphological Component, and instead assume that syntactic derivation and morphological operations such as m-merge are cyclically interleaved.In the first part of the article, we focus on the behaviour of clitics in contexts of V-to-C movement. As object clitics and negation are pied-piped by the verb to C, crossing the position of subject clitics, we argue that subject clitics are m-merged after V-to-C movement.The second part of the article deals with some puzzling permutations affecting the order of clitic elements. In particular, we focus on the Friulian dialect of Forni di Sotto (Manzini & Savoia 2005, 2009 to show that such permutations are due to morphological rules of fission and metathesis operating after m-merge. We therefore claim that the Forni pattern provides further evidence for syntactically void operations taking place at the Syntax/PF interface.

Interleaving Syntax and Postsyntax: Spellout before Syntactic Movement

“…My proposal is similar to that of Calabrese & Pescarini (), who argue that morphology and syntax in Friulian are interleaved, so that morphological cliticization of the object and the subject clitic onto the verb is interrupted by syntactic movement of the verb. The purpose of this article is to give more morphological evidence for the existence of this ordering of operations, the one in the right column of table (postsyntax ≺ syntax), and to argue that this view of the syntax–postsyntax interaction is superior to proposals that invoke idiosyncratic syntactic operations in specific structures to derive the correct morpheme order.…”

“…In this subsection I showed how the claim that a phase head first triggers Spellout of its complement and then proceeds to check its own features, establishing relations with elements in the spelled‐out domain, can explain the puzzling behavior of the clitic‐like past‐tense morpheme oon , which is sometimes pied‐piped by the verb and other times stranded below the verb. Later, in section 3, I show similarities between this phenomenon and (i) object cliticization in Friulian, which is analyzed in a similar way by Calabrese & Pescarini (), (ii) Extraordinary Left‐Branch Extraction in Slavic, in which a preposition is cliticized onto the extracted element, accounting for nonconstituent movement, and (iii) person prefixation in Classical Hebrew. All these cases can be argued to involve postsyntactic affixation with subsequent movement from the Spellout domain—in other words, situations in which a postsyntactic process feeds a syntactic one.…”

“…The goal of this article is to offer more morphosyntactic evidence that at least some Spellout operations can precede movement out of the spelled‐out domain, by showing that a specific prediction of this hypothesis is correct. Namely, if a postsyntactic process can feed a later syntactic operation, as Calabrese & Pescarini () argue is the case for cliticization in Friulian, we also expect that a postsyntactic process can be bled if the structural conditions for its application are not met because of a syntactic operation inside the Spellout domain. The main part of this article examines precisely such a case, concerning feeding and bleeding relations between syntactic head movement and postsyntactic affixation/cliticization in the verbal inflectional domain of the Niger‐Congo language Wolof.…”

“…Section 2 explores the affixation of the past‐tense morpheme in Wolof, showing that postsyntactic affixation feeds syntactic movement in one structure and in another is bled by earlier syntactic processes, leading to opacity effects. In section 3 I first briefly discuss the two cases already mentioned in the literature for which this analysis best explains the data: Friulian cliticization, as analyzed by Calabrese & Pescarini (), and Extraordinary Left‐Branch Extraction in Slavic. I then show how this analysis can also be applied to the data from Classical Hebrew that, according to Harbour (), challenge the PersonP Hypothesis.…”

“…An alternative approach is to change the order of syntactic and postsyntactic operations, so that (at least some) morphological processes can precede syntactic ones at particular points in the derivation. This is proposed by Calabrese & Pescarini () to account for the ordering of subject and object clitics and the verb in a dialect of Friulian, a language of northern Italy. Much earlier in the literature, Bresnan () shows that there are cases in which main sentence stress is assigned before movement, so that moved elements can have a level of stress that was assigned to them in a different position.…”