LINGUIST List 32.787

Wed Mar 03 2021

Review: Romance; Linguistic Theories; Syntax: Fernández-Sánchez (2020)

Editor for this issue: Jeremy Coburn <>

Date: 30-Sep-2020
From: Morgane Jourdain <>
Subject: Right Peripheral Fragments
E-mail this message to a friend

Discuss this message

Book announced at

AUTHOR: Javier Fernández-Sánchez
TITLE: Right Peripheral Fragments
SUBTITLE: Right dislocation and related phenomena in Romance
SERIES TITLE: Linguistik Aktuell/Linguistics Today 258
PUBLISHER: John Benjamins
YEAR: 2020

REVIEWER: Morgane Jourdain


In ‘Right peripheral fragments: Right dislocation and related phenomena in Romance’, Javier Fernández-Sánchez aims at identifying why right-dislocated constituents exhibit properties typical of both clause-internal and of clause-external constituents. For this, the author argues that in Romance languages, Right Dislocations (RD), as in (1), are better accounted for with a biclausal interpretation.

(1) [Yo lo suelo hacer con menta]HC, [el cordero]D (Spanish)
I it often cook with mint the lamb
‘I usually cook it with mint, the lamb.’
(Fernández-Sánchez, 2020:6)

In Chapter 1, Fernández-Sánchez explains the general ‘paradox’ of the structure of dislocation, by reporting several properties of the Dislocated constituent (D) that would support the hypothesis that D is clause-internal. For example, D acts as if it was within the Host Clause (HC) for binding purposes. The author reports other properties, in favor of a clause-external interpretation, such as the fact that a HC can function as an independent clause, as D does not leave a gap. The author then claims that the best approach to account for these seemingly contradictory properties is to assume that dislocations are biclausal structure, D and HC belonging to two different clauses. D would therefore be external to HC, but internal to another clause. This approach was originally put forward by de Vries (2009) and Ott (2015) for left- and right-dislocated constituents. The originality of the contribution of Fernández-Sánchez is that this author provides a unified account of RD for Romance languages.

In Chapter 2, Fernández-Sánchez provides a detailed account of the biclausal interpretation of RD. In this analysis, D is assumed to belong to a separate clause that has a similar syntactic and semantic structure as the HC. The parts of the clause of D which are similar to the HC are then elided, as shown in (2).

(2) [L’at mandata Juanne] [At mandata cudda líttera Juanne] (Sardinian)
It has sent John has sent this letter John.
‘John sent it, this letter’
(based on Jones, 1993: 320)

The author explains that this interpretation accounts for the clausal-internal properties of D. Both the clitic pronoun within the HC and D have a theta role, but this does not violate the Theta Criterion of Chomsky (1981) that only one argument can carry a theta role, as the clitics in the HC and D have a theta role in different clauses. Besides, D carries case-marking, not because it is belongs to the HC, but because it carries the case of its own separate clause. The biclausal analysis can also account for the clause-external properties of D: D does not leave a gap in the HC because D did not move from the HC.

In that chapter, Fernández-Sánchez also provides evidence against other syntactic interpretations of RD. It is sometimes claimed in the literature that clitic pronouns are agreement markers. Ds would therefore correspond to the actual argument within the sentence, which could also explain some of the clause-internal properties of D mentioned above. The author shows however that RD cannot be analyzed in terms of agreement-marking, because in French, the resumptive expression may be a strong pronoun, or even a lexical expression, instead of a clitic pronoun (see also De Cat 2007, Horváth, 2018). He also argues against a clitic doubling interpretation of RD, based on the fact that in RD, D does not belong to the same intonation phrase as the HC. Finally, Fernández-Sánchez demonstrates that Ds are opaque domains for extractions, which is expected under a biclausal interpretation of RD.

In Chapter 3, Fernández-Sánchez reviews the other syntactic analyses of RD, which all assume that RD is a monoclausal phenomenon, and highlights the issues they have to accommodate RDs in Romance languages. The author first reports the study of Kayne (1994), which states that D are produced in situ instead of a derived position. The criticisms against the interpretation of Kayne by Fernández-Sánchez include the facts that (i) the in situ interpretation cannot account for the less strict ordering of right-dislocated constituents than of their non-dislocated constituent counterparts in Catalan, (ii) D does not allow extraction, (iii) the in situ interpretation cannot account for the fact that RD is allowed in French and Italian but not clitic doubling.

The next approach reviewed by Fernández-Sánchez is the peripheral approach, according to which in RD, D is located in the C-domain of the clause. Some authors argue that D is directly generated within the C-domain (De Cat, 2007) while others claim that D is located in the C-domain after movement (Fernández-Sánchez, 2012). The author demonstrates that this approach, both for the base-generated view and the movement interpretation, leads to the assumption of more complex processes to account for the clause-internal and clause-external properties of D than the biclausal interpretation.

Finally, Fernández-Sánchez reviews the middle field approaches, according to which D undergoes Ā-movement to the left periphery of the clause (see for example Cecchetto, 1999). Under these approaches, RD and Left Dislocation (LD) are assumed to be structurally different, which leads to several asymmetries between RD and LD, including the possible extraction of a LD out of a RD but not the other way around or different interaction with the c-command, intra alia. Fernández-Sánchez reinterprets the original claims of asymmetry, and shows that they do not hold. He even demonstrates that LD and RD behave completely similarly in Italian and in Catalan. Therefore, it seems that the biclausal approach described in Chapter 2 provides a better fit for RD in Romance languages than the monoclausal approaches reviewed in Chapter 3.

In Chapter 4, now assuming that RDs are indeed biclausal, Fernández-Sánchez determines whether D undergoes movement within its clause before elision, as it is sometimes assumed in the literature that remnants from elided clauses escape the domain of ellipsis via movement to the left periphery of the clause (see for example Merchant, 2001). To test this claim, Fernández-Sánchez uses islands. For this test, the author generates RDs with the resumptive clitic in an island position. If D moves to the left periphery of its elided clause, there should be island-violation, and the RD should not be grammatical. If the RD is grammatical, then this would entail that D does not undergo movement. Fernández-Sánchez shows that if D is not adjacent to the island containing the resumptive clitic, the RD is not grammatical (3) but it becomes grammatical if D is adjacent to the island containing the clitic (4).

(3) * [Qu’ elle sera fâchée contre son frère] c’est clair, Marie. (French)
That she will-be angry against her brother it-is clear Marie
*’That she’ll be angry with her brother is clear, Marie.’
(Delais-Roussarie et al., 2004: 521)

(4) [Qu’ elle sera fâchée contre son frère], Marie, c’est clair. (French)
That she will-be angry against her brother Marie it-is clear
(Delais-Roussarie et al., 2004: 521)

These mixed results mean that D cannot freely move, and that there are at least restrictions if there is movement. Fernández-Sánchez argues that the simpler explanation is that D does not move, for the following reasons. First, there is no motivation that makes this movement necessary. Second, the position of D in the left periphery of the elided clause would mean that D would always take a wide scope, which is not supported by the data. Finally, right-dislocated PPs must be produced with their preposition, as Romance languages do not allow P-stranding. If D moved to the left periphery, it could be produced without its preposition.

In Chapter 5, Fernández-Sánchez offers a unified syntactic analysis for different types of right-peripheral fragments, namely Split Questions (SQ), as in (5a) and Afterthoughts (AT) as in (5b). The author argues that SQs and ATs exhibit similar clause-internal and clause-external properties as D in RD.

(5) a. Where did they stay, at the ritz?
b. He stayed at a nice hotel when he was in London - at the ritz.
(Fernández-Sánchez, 2020: 146)

The author shows that, similar to D, SQs and ATs do not undergo movement to the left periphery of the elided clause, except for predicative ATs (6). Fernández-Sánchez explains that this is due to the emphatic nature of predicative ATs, that requires them to move to the left periphery.

(6) He leído Rayuela durante las vacaciones de verano – Una novella maravillosa. (Spanish)
Have read Rayuela during the holiday of summer a novel wonderful
‘I’ve read Rayuela during the summer holiday – a wonderful novel.’
(Fernández-Sánchez, 2020: 172)

One aspect that differentiates RD from SQ and AT according to the analysis of Fernández-Sánchez is that, since the elided clause in SQ and AT have their own illocutionary force, but not in RD, the coordination of the two clauses in RD occurs below the level of Force0. By contrast, the author concludes that the coordination of the two clauses in SQ and AT involves the coordination of two ForcePs.

In Chapter 6, Fernández-Sánchez concludes his analysis of RD in Romance languages as a biclausal construction. The author explains that this interpretation of RD does not require any language-specific mechanism, which makes it generalizable across languages. He also mentions that this work has implications for the theory on ellipsis, as he showed in Chapter 4 that movement of D was not required to account for the structure of RD.


The aim of this monograph is to provide a unified account of RD, and possibly of other types of constituents produced in the right periphery, in Romance. Fernández-Sánchez manages to provide a cohesive analysis, that seems to solve many issues of the other approaches to RD. The fact that this author’s analysis can be extended to any language is also extremely appealing. Another strength of this work is that the author strives to take into account a very large range of different syntactic properties that right-dislocated constituents may have. It is also highly commendable of the author to attempt to include such a wide variety of Romance languages, even those studied less frequently, such as Sardinian.

This monograph will be of high interest to researchers investigating generative syntax, especially those interested in the phenomena related to the peripheries, as this book provides a particularly interesting angle of analysis. The overview of the monoclausal approaches to RD seems quite complete, and exhaustive enough for readers who are less acquainted with dislocation to be able to follow the explanations and arguments for or against each approach. The number of examples throughout the book is also quite impressive, and really allows the reader to understand each point of the argumentation, even though in some instances, the author could have included more explanation on how the example illustrates the argument put forward. Nevertheless, the content of this book is still accessible and easy to read.

The only small issue with this monograph could be that by aiming at providing an analysis that could fit any language, and by only using created examples, the author might have missed some potential counter-examples for his interpretation. More precisely, I am not sure how the biclausal interpretation with no movement of D in the left periphery of the elided clause can account for dislocated strong pronouns, which are very common in French (De Cat, 2007). Such strong pronouns cannot function as the subject of the sentence, unless they belong to D. In the example in (7), the D ‘moi’ cannot belong to a clause with a similar structure as the HC, with elision without movement, because the clause hosting D is not grammatical in French.

(7) [Je ne l’ai pas vu] [moi ne l’ai pas vu]
I NEG him-have not seen me NEG him-have not seen
‘I haven’t seen him.’

Fernández-Sánchez does not consider RD with strong pronouns as Ds in his analyses, as he provides almost exclusively examples with lexical phrases. Maybe an extension of the current biclausal approach of the author could account for pronominal Ds. Overall, another potentially interesting way to carry this very interesting biclausal interpretation of dislocation forward could be to consider attested examples from corpora. This might allow testing of whether the different types of dislocation produced by speakers all follow the structure described by the author.


Cecchetto, Carlo. 1999. A comparative analysis of left and right dislocation in Romance. Studia Linguistica 53. 40-67.

De Cat, Cécile. 2007. French dislocation: Interpretation, Syntax, Acquisition. Oxford: Oxford University Press.

De Vries, Mark. 2009. Specifying coordination: An investigation into the syntax of dislocation, extraposition and parenthesis. In Language and Linguistics: Emerging Trends, Dreyer, Cynthia (ed.). New York: Nova. 37-98.

Delais-Roussarie, Elizabeth, Doetjes, Jenny & Sleeman, Petra. 2014. Dislocation. In Handbook of French Semantics, Corblin, Francis & de Swart, Henriëtte (eds). Stanford CA: CSLI. 505-530.

Fernández-Sánchez, Javier. 2012. The Syntax od PredNPs. MA thesis, University College London.

Horváth, Marton Gergeley. 2018. Le français parlé informel: Stratégies de topicalisation. Berlin/Boston: De Gruyter.

Jones, Michael Allan. 1993. Sardinian Syntax. London: Routledge.

Kayne, Richard. 1994. The Antisymmetry of Syntax. Cambridge MA: The MIT Press.

Ott, Dennis. 2015. Connectivity in left-dislocation and the composition of the left periphery. Linguistic Variation 15(2). 225-290.


Morgane Jourdain is a PhD candidate in linguistics at the university KU Leuven, Belgium, and the University of Lille, France.
Her research is centered around the L1 acquisition of information structure in French. She focuses on the development of the syntax/information structure interface in children’s production of clefts and dislocation, using both qualitative and quantitative methods. She is also interested in the L1 acquisition of register.

Page Updated: 03-Mar-2021