Publishing Partner: Cambridge University Press CUP Extra Publisher Login
amazon logo
More Info

New from Oxford University Press!


Style, Mediation, and Change

Edited by Janus Mortensen, Nikolas Coupland, and Jacob Thogersen

Style, Mediation, and Change "Offers a coherent view of style as a unifying concept for the sociolinguistics of talking media."

New from Cambridge University Press!


Intonation and Prosodic Structure

By Caroline Féry

Intonation and Prosodic Structure "provides a state-of-the-art survey of intonation and prosodic structure."

Review of  Clitics between Syntax and Lexicon

Reviewer: Eva Monrós
Book Title: Clitics between Syntax and Lexicon
Book Author: Birgit Gerlach
Publisher: John Benjamins
Linguistic Field(s): Syntax
Issue Number: 14.2699

Discuss this Review
Help on Posting
Date: Sat, 4 Oct 2003 19:22:03 +0200
From: Eva Monrós Marín
Subject: Clitics between Syntax and Lexicon

Gerlach, Birgit (2002) Clitics between Syntax and Lexicon, John
Benjamins Publishing Co., Linguistik Aktuell/Linguistics Today 51.

Eva Monrós, Departament de Lingüística General, Universitat de Barcelona.


In this volume Birgit Gerlach offers a morphological analysis of
Romance clitics with frequent references to both the phonological and
syntactic levels. This kind of analysis allows for the integration of
constraints and rules from all grammatical modules in order to deal
properly with such a complex phenomenon. On the other hand, the book
covers the main problems of cliticisation discussed in the linguistic

CHAPTER 1. Introduction

A clitic is generally understood to be a word that cannot stand on
its own and leans on a host word in order to become integrated into a
prosodic constituent. Clitics lack stress and exhibit a specific
phonological behaviour, as well as syntactical and morphological
particularities. They are usually functional elements (i.e. pronouns,
determiners, auxiliaries, negation particles and question markers).

This book is devoted to the study of verbal clitics in several
Romance languages, mainly standard Italian, Iberian and Rio de la Plata
Spanish, standard French, Iberian and Brazilian Portuguese and standard
Romanian. In this chapter, three problematic issues are introduced: the
crosslinguistic differences in clitic paradigms (i.e. inventory, lexical
entries and syncretism), clitic sequencing (i.e. order, restrictions on
combination and separability), clitic placement (i.e. generation, surface
position) and doubling. These three issues are dealt in detail in
chapters 3-5. The main theoretical assumptions used in the analysis are
taken from the framework of Minimalist Morphology (MM), Lexical
Decomposition Grammar (LDG), Optimality Theory (OT), and Correspondence
Theory (CT) (see section 1.3).

CHAPTER 2. The status of Romance clitics between words and affixes

In chapter 2 the author argues in favour of considering clitics an
independent morphological category. The reasons go beyond the traditional
descriptive properties. Clitics have a different status from inflectional
and even phrasal affixes. Let us see the main prosodic and morphological

The phonological properties of clitics have been largely described
throughout linguistic research. Section 2.2 reviews the main
contributions of prosodic structure theories. On one hand, some proposals
regard clitics as constituting an autonomous prosodic unit, as they
generally do not affect stress distribution. On the other, they are also
assumed to integrate a prosodic word together with a stressed element.
Gerlach seems to prefer the second option, stress assignment and
readjustment being crucial in this choice.

As regards morphology, what are the differences between clitics and
inflectional affixes? The so-called 'Zwicky-criteria' (Zwicky, 1985) are
used to determine the morphological status of an element and its degree
of autonomy. Combining Zwicky-criteria and Nübling's (1992) bundle of
scales, Gerlach concludes that Romance pronominal clitics are neither
words nor affixes, but constitute an independent category.

Clitic morphology differs from affixal morphology in several
respects. All of them are explored and thoroughly exemplified in section
2.2. The following phenomena can only be explained by specific
restrictions on clitics, as affixal elements do not allow as much
variation: arbitrary gaps in clitic-verb and clitic-clitic combinations,
morphophonological idiosyncrasies (e.g. vowel deletion, consonant
neutralisations, opaque sequences), particular internal ordering, host
selectivity and partial changeability with full pronouns.

As for other clitic-like forms, most of them are also given the
morphological status of clitics, different from affixes and words. The
discussion focuses on Romanian auxiliaries and French negation, among

To close chapter 2, the problems of Anderson's model of phrasal
affixation are presented (Anderson, 1992). The empirical and theoretical
problems arise with respect to the domain of clisis, as it can depend
even on semantic conditions or on the internal morphological structure of
the verbal host (e.g. future forms in Portuguese).

CHAPTER 3. Clitics in the lexicon

Chapter 3 provides the first theoretical formulation of the book. As
mentioned above, the theoretical background combines MM, LDG, OT and CT.
According to MM, all linguistic elements are represented in the lexicon
with their phonological form, their morphosyntactic features and their
semantics; it is an early insertion model. The lexicon supplies the input
information for the utterance and a set of rules and constrains generate
a set of output candidates. The language-specific constraint ranking and
the Faithfulness constraints evaluate each of the output candidates. For
instance, a Faithfulness constraint such as the MAX(arg)M constraint will
require that an argument is realised as a clitic - or an affix - in the
output. OT and CT provide notions of ranking and violability of

Besides the theoretical background, this chapter analyses the clitic
inventories in each of the studied languages. The phonological
representation and morphological value of pronominal clitics are very
similar among languages, whereas the lexical entries differ a great deal
depending on the paradigmatic opposition inside the language-specific
clitic paradigm. As regards the realisation of arguments by clitics,
Gerlach analysis resembles Grimshaw's (1997) optimal approach. After
discussing Grimshaw's analysis of Italian, the author explains her own

Within the lexicon, clitic inventories reflect the contrasts and
syncretisms among elements. Syncretism is accounted for by
underspecification; for example, gender is underspecified in the lexical
entry of 1st and 2nd object pronouns in all Romance languages. The
relevant features consigned in the clitic paradigms are: person, gender,
number, case, reflexivity, category [+D], [-max,-min] (i.e. they are
neither words nor affixes) and [+V] (i.e. they are attached to a verb).
In general, 3rd person exhibits the most unambiguous forms.

As for the choice of clitics from the paradigm, the Faithfulness
constraints will assure that every feature of the input is represented in
the output sequence. On the other hand, the Markedness constraints
regulate the co-occurrence of certain features; this explains, for
example, the lack of gender marking in dative pronominal clitics (i.e.
case is ranked over gender).

The reader can see the details of the analysis for each language in
sections 3.2.2-3.2.7. Among them, Portuguese and Spanish have the
simplest paradigm, whereas French and Romanian have the most complex
ones; The French inventory incorporates the negation, a partitive, a
locative and the subject clitics, and Romanian includes also auxiliary

CHAPTER 4. Clitic sequences

Clitic combinations exhibit a particular behaviour as regards
linearisation, co-occurrence and surface forms. In chapter 4 the author
claims that clitic sequences form a morphological unit with its own

One of the most relevant properties of those units is
inseparability: all the clitics of the same verb are placed in the same
sentential position. Moreover, clitics are strictly ordered with respect
to each other within the sequence. The features responsible for the order
are case and person; for instance, IO clitics almost always precede DO
clitics, whereas 1st and 2nd person precede 3rd person. Another
well-known characteristic of the combinations of clitics is opacity, due
to deletion, substitution or changes in the surface form of the sequenced
elements (cf. Bonet, 1995). As usual in this book, a lot of examples are
offered for these properties.

Gerlach proposes an Optimality-theoretical explanation for clitic
sequences. As she points out, this kind of analysis allows combining
simultaneously constraints from different grammatical modules. The
relevant set of output candidates for the sequence is formed from
elements of the input; then, evaluation constraints operate. For the sake
of illustration, let us briefly see some examples.

At the morphological level, ALIGNMENT constraints - linked to
markedness in the person and argument hierarchy - are responsible for the
internal order of the elements. On the other hand, prohibited
combinations (i.e. 1/2OD+3OI) are prevented by means of a 'harmonic
alignment' restriction between person and argument hierarchies (cf.
section 4.3.4 for details). In the phonological level, clitic sequences
obey minimality conditions that, among other operations, trigger
phonological reduction or deletion. Besides, they also obey certain
dissimilation conditions (i.e. the elements of a clitic sequence must be
dissimilar) which are the reason for the opaque forms.

CHAPTER 5. Clitics at the interfaces: clitic placement and clitic

Chapter 5 is concerned with the placement of clitics in the sentence
which is different from that of full DPs. To begin with, Gerlach
summarises two well-known purely syntactic approaches: movement (Kayne,
1975) and base generation (Jaeggli, 1982). Next, she offers an analysis
in terms of OT, combining syntactic, morphological and prosodic ALIGN
constraints. These constraints regulate the syntactic domain of
cliticisation (i.e. V), the position of the clitics within this domain
(i.e. enclisis or proclisis), and the prosodic attachment. Among the most
problematic issues we find mesoclisis in European Portuguese.

As regards clitic doubling, it is a prohibited phenomenon in many
Romance languages. When allowed, clitic doubling must obey certain
properties, such as animacy, specificity, and the argument grammatical
relation and discourse status.


One of the most outstanding qualities of this book is probably the
important empirical background provided. Each property, condition or
restriction is exemplified in the specific language containing it,
rending the exposition comprehensible. The data come from a wide range of
several sources, such as Perlmutter (1970), Monachesi (1995), Grimshaw
(1997), Crysmann (2000), among others.

The examples almost always reach deep detail. For instance, when
dealing with arbitrary gaps, Gerlach presents the personal restriction
*1+2 and does not forget to remark that this combination is possible in
Spanish if one of the pronouns is an ethical dative. However, on other
occasions - certainly rare -, the lack of detail can be misleading: in
pp. 222-223 the author states that clitic doubling is obligatory in
Spanish with indirect objects, which would require further explanations.

On the other hand, the analysis is well framed within the reference
theoretical models. The postulate presentation of the before
mentioned models given in chapter 1 allows the reader to get familiar
with the theoretical framework and comfortably follow the author
explanations. The exposition is explicit throughout the analysis despite
its complexity. I would like to highlight that this observation does not
stem from a critical position, it only pretends to make evident that the
book is written for a specialized reader.

The study of the particularities of the Romance pronominal clitics
leads to several theoretical conclusions. From the point of view of
clitics themselves, we have already pointed out the fact that there is
enough evidence to consider them an independent morphological category.
Therefore, we can claim that cliticisation is an interface phenomenon.
However, Gerlach's analysis has much deeper implications such as the need
to postulate an autonomous morphological component within the grammatical
system (cf. section 3.1.)

To conclude, this volume offers us an exhaustive revision of the main
aspects concerning Romance pronominal clitics. To put it in nutshell, we
are dealing with a book which contains both a solid empirical basis and
a consistent theoretical formulation.


BONET, E. (1995). "Feature structure of Romance clitics". Natural
Language and Linguistic Theory 13, pp. 607-647.

CRYSMANN, B. (2000). "Clitics and coordination in linear structure".
Dins: Gerlach, B. I J. Grijzenhout (eds.) Clitics in Phonology,
Morphology and Syntax, Amsterdam/Philadelphia, John Benjamins, pp.

GRIMSHAW, J. (1997). "The best clitic: constraint conflict in
Morphosyntax". Dins: Haegemann, L. (ed.) Elements of Grammar, Dordrecht,
Cluwer, pp. 169-196.

JAEGGLI, O. A. (1982). Topics in Romance Syntax. Dordrecht: Foris.

KAYNE, R. S. (1975). French Syntax: the Transformational Cycle.
Cambridge MA: MIT Press.

MONACHESI, P. (1995). A Grammar of Italian Clitics. PhD Diss.
Tilburg University.

NÜBLING, D. (1992). Klitika im Deutschen. Schriftsprache,
Umgangssprache, Alemanische Dialekte. Tübingen: Gunter Narr.

PERLMUTTER, D. M. (1970). "Surface structure constraints in syntax".
Linguistic Inquiry 1, pp. 187-255.

ZWICKY, A. M. (1985). "Clitics and particles". Language 61.2, pp.
ABOUT THE REVIEWER Eva Monrós holds a degree in Catalan Language and Literature (1997) and in General Linguistics (2000). She is a researcher at the Universitat de Barcelona (Spain) under a grant for postgraduate students from the Catalan Government and is currently working with a team of researchers from Brazil. Her research interests are syntactic typology and minimalist framework. Her on-going PhD project deals with ergativity in Amazonian languages.