Review: Ling Theories; Morphology; Syntax: Meklenborg Salvesen & Helland (eds.) (2013)

EDITOR: Christine  Meklenborg Salvesen
EDITOR: Hans Petter  Helland
TITLE: Challenging Clitics
SERIES TITLE: Linguistik Aktuell/Linguistics Today 206
PUBLISHER: John Benjamins
YEAR: 2013

REVIEWER: Hannah B. Washington, Ohio State University


This volume consists of a selection of 11 articles from the Challenging
Clitics workshop, which took place at the University of Oslo in October of
2011. The articles are comprised of a wide range of methodological approaches
in the treatment of clitic-related questions in a variety of languages. By
including a diverse body of articles, the editors aim to present a volume that
sheds light on what clitics are and how they function across syntax,
morphology, and phonology, as well as the interfaces that connect them.

By way of introduction, Christine Meklenborg Salvesen and Hans Petter Helland
(editors) begin the collection with their introductory chapter entitled “Why
challenging clitics?”, in which they offer a brief history of influential work
on clitics. This background begins with seminal descriptive works from the
19th century, which resulted in the well-known Tobler-Mussafia law for Romance
clitics and Wackernagel’s law of second position clitics in Indo-European.
More recent work on clitics starting in the 1970s with Kayne (1975), Zwicky
(1977), and Klavans (1985) provided the first theoretical frameworks through
which to understand clitics and clitic-related processes cross-linguistically.
For this volume, the editors adopt a model in which clitics are categorized
following Zwicky (1977) as either simple or special, sharing the basic
properties of being non-accentuated independent elements that require a host
and which may or may not have a lexical alternant. Modern work on clitics,
they argue, is most interested in the distribution of clitics in relation to
syntactic principles, although purely syntactic accounts have often had
difficulty accounting for clitic patterning. Some of the primary questions in
research on clitics from a syntactic point of view include those of whether
clitics are thematic arguments or functional heads. The editors further
provide a basic review of recent work that connects clitic movement to phases,
in contrast to prior analyses that suggest clitics move to a host within the
Inflectional Phrase (IP) field. With this background in mind, this volume
provides a new body of work from varied perspectives to contribute to the
understanding of clitics cross-linguistically.

Marios Mavrogiorgos, in his chapter entitled “Enclisis at the syntax-PF
interface”, explores the interaction between syntactic and PF (phonetic form)
operations using data from languages whose clitics are sensitive to finiteness
and languages that adhere to the Tobler-Mussafia law. In the
finiteness-sensitive languages of Western and Southern Europe, the alternation
between proclisis and enclisis is related to properties of the verbal host,
with enclisis predominant in the presence of a nonfinite verbal form and
proclisis found in the presence of tensed verbs. In contrast, Tobler-Mussafia
languages disallow clitics in utterance-initial position but allow them in any
other verb-adjacent position. Clitic placement in these languages typically
corresponds to the presence of certain left-periphery elements. Prior analyses
have suggested that enclisis is either the result of a PF operation on
proclitic inputs (i.e. clitic movement) or the result of the syntactic
movement of the verb to a higher head (i.e. verb movement), with either
syntactic or phonological motivations for the movement of a constituent; in
analyses that account for the PF-syntax interface, it is the verb that moves
with a phonological or prosodic motivation in Tobler-Mussafia languages.
Providing evidence that shows the theoretical problems with such analyses,
Mavrogiorgos argues that the alternation in clitic placement in both
finiteness-sensitive and Tobler-Mussafia languages is the result of both
syntactic and PF operations, and that these operations are not uniform across
all languages. In so doing, he delineates three distinct cases of proclisis
derivation and four distinct cases of enclisis derivation. In contrast with
other proposals relating to the nature of the alternation between proclisis
and enclisis in these two kinds of languages, his proposal suggests that
enclisis always depends on PF operations rather than on syntactic movement.

Francisco José Fernández-Rubiera, in his chapter entitled “Clisis revisited:
Root and embedded contexts in Western Iberian,” examines data from Asturian to
explore the difference in pronominal object clitic alternation between this
Western Iberian language on the one hand and Galician and European Portuguese
(EP) on the other. In contrast with EP and Galician, Fernández-Rubiera shows
that Asturian allows enclisis in finite embedded contexts due to differences
in the complementizer system between this language and the other two Western
Iberian languages mentioned. Specifically, Asturian displays variation in
clitic placement in embedded clauses in the presence of finite verb forms
depending on whether the preverbal subject bears focus. Meanwhile, if there is
no explicit subject in the left-periphery, both proclisis and enclisis are
available options. Fernández-Rubiera reviews prior syntactic and phonological
approaches, finding that both kinds of approaches are inadequate because they
require a single operationally-induced output. The author instead proposes a
new account using the concept of phases, such that Finiteness0 is a phase head
that carries an edge condition. Using this structure, the author argues that a
difference in interpretation is found between proclisis and enclisis in the
embedded clause context when there is not a focus-bearing preverbal subject
(or topic). Under this analysis, the pragmatic evidential feature
[+conviction] is found in the presence of enclisis, while proclisis
necessitates the presence of a truncated structure with the feature
[-conviction]. Fernández-Rubiera indicates that the primary benefit of this
approach is that it not only accounts for the alternation found in Asturian
but also can explain obligatory proclisis in Galician and EP.

In “Handling Wolof clitics in LFG”, Cheikh M. Bamba Dione offers a generative
approach to explore cliticization in Wolof within the Lexical Functional
Grammar (LFG) framework. Like special clitics in Romance, much of the focus in
the literature has depended on the idea of clitic movement (Kayne 1975)
despite the fact that clitic placement in Wolof relies on a combination of
phonological, morphological, and semantic properties. Unlike previous
treatments of Wolof clitics, Dione’s analysis avoids violation of the Lexical
Integrity Principle (Bresnan 2001) and thereby diverges from transformational
work that relies on phrase structure and constituent movement. In this paper,
Dione provides a sketch of LFG and its theoretical assumptions and offers a
clear presentation of how it can be employed to explain the factors that
affect object and locative clitic placement and clitic climbing within Wolof.
By including separate surface mappings between the diverse components in the
grammar, the author’s analysis accounts for the divisions and interfaces
between the phonology, morphology, and syntax without requiring an underlying
structure. Furthermore, this paper shows that LFG includes the necessary tools
and notation to develop each component of the grammar and argues for the
further use of LFG in the study of clitics cross-linguistically.

In “Clitic placement and grammaticalization in Portuguese,” Filomena Sandalo
and Charlotte Galves consider the question of diachronic change in object
clitic placement between Classical and Modern Portuguese, within Distributed
Morphology. These authors show that the alternation between proclisis and
enclisis of pronominal clitics in Portuguese cannot be explained by finiteness
restrictions as it is in other Romance languages. Distributed Morphology
allows the authors to employ derivational movement operations at different
levels within the grammatical architecture, as laid out by Embick & Noyer
(2001). Within this theoretical framework, both prosodic cliticization in
which a clitic leans on a host and post-syntactic affixation can be explained
through movement operations. The authors explore two kinds of Merger
operations in which units become more closely attached to other elements
through morphological incorporation: Lowering and Prosodic Inversion. Lowering
causes a head to lower to another head via a right dislocation operation and
can operate for long distance dislocation. Prosodic Inversion, meanwhile,
deals with elements in the periphery of the sentence, causing clitic elements
to lean on hosts rather than being subject to incorporation if the clitics
have only a prosodic dependency. Clitics in Portuguese surface as proclitics
only in the presence of a preverbal c-commanding functional head, in which
case the authors suggest that the clitic remains in situ. Enclisis, however,
is the result of a Merger operation (i.e. Lowering) by which clitics are not
c-commanded by another head and thus become more affix-like in nature.
Finally, the authors show that the situation is somewhat different in
Classical Portuguese, where enclisis is shown to be the result of a Prosodic
Inversion operation that incorporates the clitic in the prosodic domain.

In his paper entitled “Diachronic source of two cliticization patterns in
Slavic”, Krzysztof Migdalski gives an analysis of the change from
verb-adjacent to second position cliticization in the Slavic languages. While
Bulgarian and Macedonian retain the verb-adjacent clitics, second position
clitics are found in other Slavic languages (i.e. Czech, Serbo-Croatian,
Slovak, Slovene, and Polish). A study of Old Slavic reveals that clitic
placement resembles that of the first group with required verb-adjacency, with
a change in the latter group of languages to second position cliticization
that took place simultaneously with the loss of morphological tense. The
author interprets these simultaneous changes as indicative of the relationship
between verb-adjacency of clitics and the existence of
morphologically-expressed tense. He posits that these two interrelated issues
can be captured through a single syntactic generalization. That is, in a
language lacking morphological tense, the TP (Tense Phrase) projection is
unavailable, leaving clitics without a host to which they can adjoin in the
T-head. This lack of a suitable host leaves the clitic in need of a ‘second
best’ option that is not needed in languages where T is available. Migdalski
proposes that the preference for clitic adjunction to verbal hosts over second
position clitics is due to a preference for a syntactic constituent to
correspond with a prosodic word. Prosodic word status is available for
verb-adjacent clitics, while it is unavailable for second position clitics
because the elements of the prosodic word end up in different specifiers. This
analysis predicts that, unlike universal projections, TP is available to
emerge or be lost in the course of language change.

Gréte Dalmi’s chapter, “The Freezing Principle in Hungarian polarity,
non-polarity and multiple wh-questions,” considers the syntactic role of the
weak erotetic ‘vajon’ (“whether, if…at all”) in the presence of the
interrogative clitic ‘-e’ in Hungarian, as well as the interaction between
‘vajon’ and Long wh-movement. The author demonstrates that Long wh-movement in
subordinate interrogative clauses provides evidence of ‘vajon’ as a marker of
an operator-variable chain rather than as a complementizer. The author’s
evidence suggests a difference in the availability of Long wh-movement in
veridical vs. non-veridical contexts, which provides evidence for the presence
or absence of a freezing effect that is blocked or neutralized depending on
the status of the operator chain. That is, in veridical contexts, the chain
formed by the interrogative with the associated clitic blocks Long wh-movement
out of subordinate clauses with ‘vajon’. Meanwhile, non-veridical doxastic
epistemic predicates allow both Long and Partial wh-movement with ‘vajon’.

Francine Alice Girard presents data on pronouns from Cajun French (spoken in
the US state of Louisiana) in her article “Pronominal markers in Cajun
French”. The author considers both subject and object pronouns in Cajun
French, and discusses whether the pronouns fit into Zwicky’s (1977) special
clitic category or are better described as affixal elements. She also
addresses the question of how these pronominal objects compare with objects in
other varieties of French. Using extensive examples, the author shows that
pronouns in Cajun French differ in a number of ways from pronouns in Standard
French and other varieties of colloquial French, including in the expanded use
of strong pronouns and use of ça ‘that’ in Cajun French. Girard concludes that
Cajun French subject pronouns are not completely affixal in nature, given that
this variety shows both subject doubling and optionality with respect to
cliticization. The author further argues that the reduction of the accusative
and dative systems is suggestive that the object clitic system in Cajun French
displays more affixal behavior, similar to what has been found in other
varieties of French. These divergent tendencies between subject and object
pronouns in Cajun French challenge prior analytical approaches, leading the
author to understand the clitic system in this variety as uniquely situated at
an advanced stage in the grammaticalization process in which clitics are in
the process of being reduced completely and replaced with tonic forms.

The question of whether clitics can be adopted through language contact is
addressed in Natalia Pavlou and Phoevos Panagiotidis’s chapter “The
morphosyntax of ‘-nde’ and post-verbal clitics in Cypriot Greek”. The authors
show that the presence of the verbal clitic ‘-nde’ in Cypriot Greek imposes
certain morphological and syntactic restrictions on the clause. Specifically,
it can only appear with first-person plural verb forms. Given that pronominal
enclitics cannot co-occur with ‘-nde’ because both sets satisfy certain
features, the authors argue that ‘-nde’ functions as a clitic rather than an
affix. Furthermore, the authors suggest that this marker has been borrowed
from Turkish -dA and carries truth value information for the proposition.
Although it is moribund in Cypriot Greek, this chapter illuminates the
possibilities of contact-induced morphosyntactic borrowings.

In the article “Acquisition of Italian object clitics by a trilingual child”,
Elizaveta Khachaturyan examines a child who acquires Italian alongside Russian
and Norwegian. Building on the work of Cipriani et al. (1993) for the
acquisition of pronominal object clitics, the author demonstrates that the
process that the trilingual child ‘Esther’ undergoes in her acquisition of
Italian clitics differs from that of monolingual children. While monolingual
children display errors in gender, number, and morphophonological processes as
they acquire clitics, Esther’s primary errors relate to the use of tonic forms
instead of clitics. Additionally, the child acquires third-person clitics
differently from first- and second-person clitics in Italian, although their
initial attestations in Esther’s language begin similarly as enclitic around
the age of 2;04. While third-person forms are acquired and then consistently
produced in a target-like fashion, first- and second-person clitics are
produced correctly in the first stage of acquisition but are later replaced by
non-clitic strong forms. Khachaturyan takes this as evidence of two distinct
phases of acquisition. In the first premorphological stage, the clitics used
by the child are among the very few morphological operators in her language,
often appearing as enclitic with imperative verb forms. The second stage
corresponds to a protomorphological stage in which Esther begins to use
morphology creatively, with a greater importance of pragmatic and syntactic
features for clitic use. In this later stage, the child begins using tonic
first- and second-person pronouns instead of clitics following verbs. The
errors produced in the second phase of acquisition correspond to changes in
word order that suggest influence from Norwegian and Russian, despite the
child’s continued correct, though less frequent, usage of third-person

Italian clitics are further analyzed in Diego Pescarini’s chapter entitled
“Clitic clusters in early Italo-Romance and the syntax/phonology interface”.
This paper examines morphophonological properties of Italo-Romance clitics,
leading the author to support the claim that clitic clusters should be
understood as a phonological Foot rather than as a series of Prosodic Words.
While the analysis of clitics as Prosodic Words suggests that clitics carry
the same prosodic status as their verbal host, the Foot analysis is indicative
of the prosodic deficiency of clitics. Using the Foot analysis, the author
describes apocope in third-person clitics in early Italo-Romance, suggesting
that the distribution derives from constraints on syllable structure and
alignment. Whereas apocope in modern Italian arises within the bounds of
certain morphosyntactic constraints, early Italian and Italian vernaculars
have a more productive employment of apocope, which favors the prosodic

Christine Meklenborg Salvesen analyzes the internal organization of clitic
clusters between Old French and Modern French in her paper on “Reflexive verbs
and the restructuring of clitic clusters”. She argues that the abrupt change
from ACC-DAT to DAT-ACC ordering of pronominal object clitics occurring in the
15th and 16th centuries is a result of an increase in the use of reflexive
forms. Much of the prior research on the topic of clitic ordering within
clusters focuses on phonological or prosodic changes, though some researchers
have suggested that a pragmatically-conditioned placement changed to a
syntactically-conditioned one. In this view, old information such as the
accusative clitic would come first; however, because the direct object holds a
closer link to the verb than the indirect object, it is argued that the
accusative form came to require a closer placement to the verb than the dative
form. Using corpus data to show frequencies of different object clitics, the
author executes a series of Fisher’s exact tests to show that clusters
involving first- and second-person singular clitics show a change in ordering
prior to  clusters containing their plural counterparts. Furthermore, she
demonstrates that the functional distinctions made by the placement of clitics
in Old French led to a reanalysis by which all forms that carried functional
differences came to require placement closer to the subject. Finally, the data
seem to suggest that the French verbal system experienced a significant
increase in the use of reflexive pronouns with transitive verbs from the 12th
to 15th centuries. This change in linearization is accounted for by
distinguishing between split clusters and true clusters, reflective of the
syntactic placement of the two clitics.


The assortment of languages and methodologies in this edited volume is
indicative of just how much the study of clitics has advanced since Kayne
(1975), Zwicky (1977), and Klavans (1985). Although a number of the collected
papers focus on clitics in Romance, which have dominated much of the prior
literature on the subject, the inclusion of clitic-related questions in
Slavic, Hungarian, Wolof, Cypriot Greek, and lesser-studied Romance varieties
such as Asturian and Cajun French offers the reader a wide range of languages
displaying varied phenomena.

Advanced scholars of clitics can learn a great deal from this volume, not only
about clitics in a variety of languages but also about different theoretical
approaches employed to describe their patterns. This is the primary strength
of this volume: it is not limited to a single subfield or linguistic
methodology. Rather than dedicating itself to syntax or morphology or
phonology, the editors have brought together a wide range of approaches to and
perspectives on the study of clitics. The methodologically diverse chapters
herein include generative approaches in the Minimalist Program, historical
accounts that work with the syntax-morphology-phonology interfaces, corpus
work that takes into account quantitative and qualitative evidence,
morphological studies that use the framework of Distributed Morphology to show
changes in constraints over time, the study of clitic acquisition by a child
whose input languages have differing clitic patterns, and clitic borrowing
between languages in close contact, among other approaches. Subsequently, this
volume builds on prior work that treats clitics cross-linguistically,
including Halpern & Zwicky’s (1996) edited volume on second position clitics
and Spencer & Luís’s (2012) more recent publication that looks at clitic
patterns in a wide range of languages.

While the variety of languages and methodological diversity of this
compilation are impressive, it is most appropriate for audiences interested in
categorical phenomena studied through a generative lens. Since the coherence
of this volume largely hinges on the generative, variation-free approaches
used by the authors, it may not be appropriate for those whose primary
interest lies in patterns of usage rather than purely theoretical models that
may not reach descriptive adequacy. Discussions of pragmatically-conditioned
patterns in this volume are superficial at best, and there is little to no
substantive discussion of variation in clitic placement, internal ordering, or
expression that goes beyond the generative understanding of competing
grammars. With that in mind, “Challenging Clitics” could benefit from the
inclusion of the discussions found in Schwenter & Silva (2003), Schwenter
(2006), and de Andrade (2010) on the role of pragmatics in patterns of
expression and variation in placement patterns. Furthermore, greater
quantitative analysis with up-to-date usage of statistical measures could
greatly improve both the descriptive and the explanatory power of the analyses
presented in this book.

Overall, the volume’s breadth of issues and methodologies is notable. The
interdisciplinary, cutting-edge formal approaches to the interfaces at which
clitics reside make it well worth the read for those interested in clitic
phenomena. Despite its limitations from a functionalist point of view, this
volume is successful in its goal of offering diverse perspectives on clitics
cross-linguistically, and it is recommended for linguists who work on the
theoretical side of the syntax-morphology interface.


Bresnan, Joan. 2001. Lexical-Functional Syntax. Oxford: Blackwell.

Cipriani, Paola, Anna Maria Chilosi, Piero Bottari & Lucia Pfanner. 1993.
L’acquisizione della morfosintassi in italiano. Fasi e processi. Padova:

de Andrade, Aroldo Leal. 2010. A subida de clíticos em português: Um estudo
sobre a variedade europeia dos séculos XVI a XX. Doctoral dissertation.
Campinas, SP.

Embick, David & Rolf Noyer. 2001. Movement operations after syntax. Linguistic
Inquiry 32(4): 555-596.

Halpern, Aaron L. & Arnold M. Zwicky (eds.). 1996. Approaching Second: Second
Position Clitics and Related Phenomena. Stanford: CSLI Publications.

Kayne, Richard. 1975. French Syntax: The Transformational Cycle. Cambridge,
MA: The MIT Press.

Klavans, Judith. 1985. The independence of syntax and phonology in
cliticization. Language 61(1):95-120.

Schwenter, Scott A. 2006. Null Objects across South America. Selected
Proceedings of the 8th Hispanic Linguistics Symposium, ed. by Timothy L. Face
and Carol A. Klee, 23-36. Somerville, MA: Cascadilla Proceedings Project.

Schwenter, Scott A. and Gláucia Silva. 2003. Overt vs. Null Direct Objects in
Spoken Brazilian Portuguese: A Semantic/Pragmatic Account. Hispania

Spencer, Andrew & Ana R. Luís. 2012. Clitics: An Introduction. Cambridge:
Cambridge University Press.

Zwicky, Arnold. 1977. On Clitics. Bloomington, IN: Indiana University
Linguistics Club.


Hannah Washington is a PhD student in Hispanic Linguistics at The Ohio State
University with a specialization in morphosyntactic and pragmatic variation.
Her doctoral dissertation (in progress) uses corpus data to examine the
variation in pronominal clitic object placement in Portuguese and Spanish,
including special reference to the role of frequency, grammaticalization, and

LINGUIST List: Vol-25-680