Elsevier

Lingua

Volume 147, August 2014, Pages 9-24
Lingua

The dynamics of the PF interface: Negation and clitic clusters

https://doi.org/10.1016/j.lingua.2013.11.002Get rights and content

Highlights

  • Form alternations of the preverbal negative marker in Italo-Romance.

  • Evidence that phonology is sensitive to syntactic structure.

  • Connection between Jespersen's cycle and the internal structure of negation.

Abstract

In this article we take into account the different development of Italian and French with respect to the Jespersen cycle of negation: both languages started with a similar system, but are nowadays rather different. We argue that several different factors are involved in the activation of the cycle, which is the result of a general economy strategy. We claim that Jespersen's cycle can be blocked if speakers have access to any kind of evidence that the negative marker is complex. Here we provide evidence that the cycle is blocked when the preverbal negative marker is morphologically complex (i.e. at least bi-morphemic). We investigate several Italo-Romance varieties (both old and modern) and show that the alternation between two forms of the preverbal negative marker which depends on the presence of object clitics can either be a syntactic process or a phonological one, (although still sensitive to syntactic information). We argue that these morpho-phonological phenomena block the activation of the cycle as they make the bi-morphemic nature of the negative marker recognizable by the speakers. In addition, the data we present can shed light on the more general principles that map the PF interface.

Introduction

Our article is devoted to the analysis of some peculiar processes at the boundary between morpho-syntax and phonology that involve clusters of negation and object clitics in some old and modern Italian varieties. In the dialects we take into consideration the preverbal negative marker, whose longer form is identical or similar to the standard Italian [non], also displays a reduced form, [no] or [n-]. This shorter form requires specific morphological and phonological conditions, which vary across varieties, but it can be shown that in all dialects except the less conservative variant of modern Florentine, it is sensitive to syntactic information and, more precisely, to structural phrasing, and not simply to linear adjacency. This alternation also has interesting consequences for the explanation of the dynamics of the Jespersen cycle: since it requires that the preverbal negative marker is bi-morphemic, i.e. morphologically complex, it blocks the typical doubling of the negative marker through so called negative reinforcers (like adverbs or minimizers) which generally starts the cycle.1 In other words, the alternation between a complex and a simple form of the preverbal negative marker is the piece of evidence that induces speakers to analyze the negative marker as a bi-morphemic item, which does not need to be reinforced by any additional morpheme, given that it is already morphologically articulated.

As we will claim below, the activation of the Jespersen cycle is due, as originally proposed by Jespersen himself, to the weakness of the negative marker, whereby we surmise that this weakness must concern all the levels of grammar, not only phonetics, and that it is to be reinterpreted as (morphological, phonological or syntactic) complexity. This means that for the cycle to activate, there can be no component of the grammar that provides the speaker with evidence that the negative marker is complex. The phenomenon we analyze is only one of the possible clues that a speaker can have concerning the complexity of the negative marker. This means that we do not predict that, as soon as the morphological phenomenon investigated here disappears, the Jespersen cycle activates, because also all other levels of the grammar must provide the speaker with no clue that negation is complex (for further cases of evidence for complexity see Postma, 2002, Meisner, 2013).

The article is organized in the following way: in section 2 we discuss some general problems about the Jespersen's cycle and the differences between the languages that have undergone it, like French, and those where it has not fully developed, like Italian posing the question of the reason why Italian has a frozen syntax while French has undergone a change in the type of negative marker.

In section 3 we observe that some Italian varieties have two alternative forms of the preverbal negative marker, and describe the conditions of the phenomenon in Sicilian and Tuscan dialects comparing it with other similar phenomena involving negation–clitic clusters. In section 4 we present some general conclusions with respect to the interaction between phonology and syntax and argue in favor of the hypothesis that phonology does not see categorial labels but only syntactic (and in some cases possibly prosodic) units.

Section snippets

The Jespersen cycle and its trigger

Since Jespersen formalized for the first time the empirical generalization which is now commonly known as the Jespersen cycle, several linguists have tried to pin down what the actual trigger of the evolution of the negative marker is. According to Jespersen's original intuition, the cycle proceeds in three steps, which we briefly mention here: (a) in the first one the original negative marker is still stable and expresses negation alone, (b) in the second stage negation is represented by a

Reduced negation in clitic clusters

In this section we examine three types of systems where the negative marker interacts with clitics and argue that these phenomena drive the speaker to assume that the negative marker is composed of more than one morpheme, thus rendering the language more stable with respect to the Jespersen cycle.

Negative and clitic interaction: from syntax to phonology

In this work we have analyzed the patterns of interaction between clitics and the preverbal negative marker and have shown that the Old Italian Florentine and Sicilian varieties have a clear alternation between two forms of the negative marker non/nun and no/nu depending on the presence of an object clitic, which requires the shorter form of the negative marker. Given that neither variety displays any general phonological rule that deletes a nasal coda in the presence of another initial

References (31)

  • P. Benincà

    Punti di sintassi comparata dei dialetti italiani settentrionali

  • P. Benincà et al.

    On clausal architecture: evidence from complement clitic placement in Romance

  • A. Breitbarth et al.

    The distribution of preverbal en in (West) Flemish: syntactic and interpretive properties

    Lingua

    (2013)
  • A. Cardinaletti

    On pronoun movement: the Italian dative “loro”, PROBUS 3.2

    (1991)
  • A. Cattaneo

    It Is All About Clitics: The Case of a Northern Italian Dialect Like Bellinzonese

    (2009)
  • L. Cornips et al.

    Standardisation of syntactic elicitation techniques

    Lingua

    (2004)
  • V. Egerland

    Diachronic change and pronoun status: Italian dative “loro”

    Linguistics

    (2005)
  • J. Garzonio et al.

    Quantifiers as negative markers in Italian dialects

    Linguistic Variation Yearbook

    (2009)
  • J. Garzonio et al.

    Non and its companions: on the big NegP hypothesis. Ms.

    (2013)
  • K. Jaberg et al.

    Sprach- und Sachatlas Italiens und der Südschweiz

    (1928–40)
  • R. Kayne

    Parameters and Universals

    (2000)
  • P. Kiparsky et al.

    Tracking Jespersen's cycle

  • M. Loporcaro

    L’origine del raddoppiamento fonosintattico

    (1997)
  • M.R. Manzini et al.

    I dialetti italiani e romanci, Morfosintassi generativa

    (2005)
  • F. Martineau et al.

    Microvariation in French negation markers: an historical perspective

  • We thank the audiences of the Italian Dialects in Diachrony Conference (Leiden, May 19–21, 2011) and of the 43rd Linguistic Symposium on Romance Languages (New York, April 17–19, 2013), Paola Benincà, Ingo Feldhausen, Imme Kuchenbrandt, Diego Pescarini, Mair Parry and Silvia Rossi for comments and discussion. For the concerns of the Italian academy, Jacopo Garzonio takes responsibility over section 3 and Cecilia Poletto over sections 1 Introduction, 2 The Jespersen cycle and its trigger, 4 Negative and clitic interaction: from syntax to phonology.

    View full text