LINGUIST List 23.1792

Sat Apr 07 2012

Review: Ling Theories; Semantics; Syntax: Rooryck & vanden Wyngaerd (2011)

Editor for this issue: Monica Macaulay <monicalinguistlist.org>



Date: 07-Apr-2012
From: Jason Ginsburg <jginsburgmail.com>
Subject: Dissolving Binding Theory
E-mail this message to a friend

Discuss this message

Announced at http://linguistlist.org/issues/22/22-3047.html
AUTHORS: Rooryck, Johan and Guido J. vanden WyngaerdTITLE: Dissolving Binding TheorySERIES TITLE: Oxford Studies in Theoretical LinguisticsPUBLISHER: Oxford University PressYEAR: 2011

Jason Ginsburg, Center for Language Research, University of Aizu

SUMMARY

The purpose of the analyses presented in this book, as can be gleaned from thetitle, is to dissolve Binding Theory. The authors take the position that''mechanisms and principles that are independently needed in the grammar'' (p. 1)account for the complex facts regarding pronouns and anaphors. In the eightchapters of this book, the authors lay out analyses for a wide variety of data(much of which is problematic for Binding Theory) in Germanic and Romancelanguages.

Chapter 1: Introduction (pp. 1-5)

In this chapter, the authors explain their goals and summarize the followingseven chapters of this book.

Chapter 2: Binding, Agree, and the Elsewhere Principle (pp. 6-53)

In this chapter the authors explain their basic views about anaphors andpronouns. The authors argue that an anaphor enters a derivation with unvalued,but interpretable, features that become valued by forming an Agree relation withan antecedent. An important component of this proposal is that an anaphorfunctions as a probe that must c-command its antecedent, a goal, in order toobtain an interpretation. For example, in the German example (1), the anaphor'sich' initially has unvalued phi-features, whereas the antecedent 'Johannes'has valued phi-features, as shown in (1b).

(1) (a)Johannes_i liebt sich_i/_*j.Johannes loves himself

(b) sich_{P:_,N:_,G:_}, Johannes_{P:3,N:sg,G:m}

(c) [vP sich [vP Johannes liebt sich]]

(d) sich_{P:3*,N:sg*,G:m*}(p. 12)

The anaphor 'sich' raises and adjoins to the vP, from where it probes for andAgrees with 'Johannes' (1c). After Agree, the anaphor and antecedent sharefeatures, as shown in (1d); '*' signifies that the features are shared. Comparethis with the German (2), where the pronoun 'ihn' cannot co-refer with theR-expression.

(2) (a)Johannes_i liebt ihn_*i/_j.Johannes loves him

(b) Johannes_{P:3,N:sg,G:m}, ihn_{P:3,N:sg,G:m}(p. 12)

Both 'Johannes' and the pronoun 'ihn' initially have valued features (2b). Thus,the pronoun does not form an Agree relation with 'Johannes'. Relying on theDistributed Morphology view (Halle & Marantz 1993, Harley & Noyer 1999) thatlexical insertion of Vocabulary Items into morphemes occurs post-syntactically,the anaphor 'sich' is inserted in (1) because the features are shared, and thepronoun 'ihn' is inserted in (2), since the features are not shared. Thus, thisanalysis requires that the lexical insertion process be sensitive to whether ornot features are shared.

In constructions such as (3), a pronoun may or may not corefer with anR-expression. In this case, the pronoun and R-expression are both inserted intothe derivation with fully valued sets of phi-features.

(3) John thinks that he is smart. (p. 15)

The possibility of coreference, following Phase Theory (Chomsky 2001, 2004,2008), results from 'John' and 'he' appearing in different phases. Since ''phasesare transferred to the semantic component separately'', they ''may or may notcorefer'' (p. 16).

In (2), in order to discount the possibility of coreference, the authors assumethat the phi-features of the pronoun are present on v, thus essentially placingthe R-expression and the pronoun in the same phase. Since the R-expression andanaphor are in the same phase, they cannot corefer.

The authors go on to argue that exceptions to Principle B effects can occur whena language lacks an appropriate reflexive pronoun. This is accounted for via theAPBE:

(4) Absence of Principle B effects (APBE)Pronouns behave like anaphors when a dedicated class of reflexive pronouns islacking. (p. 19).

For example, Dutch contains a 3rd person reflexive pronoun 'zich', but it lacksa reflexive pronoun that can be used in the 1st and 2nd person. Thus, in accordwith the APBE, when a 1st and 2nd person anaphor is required, a pronoun must beused.

Chapter 3: The Syntax of Simplex Reflexives (pp. 54-115)

In this chapter, the authors develop an analysis that explains the differingsyntactic and semantic behaviors of the Dutch simplex and complex reflexives'zich' and 'zichzelf'.

The authors propose (following work by den Dikken 2006) that a simplex reflexiveis Merged in the specifier position of a possessive phrase that functions as the''internal argument of an unaccusative verb'' (p. 54). In (5a), 'zich' is Mergedin specifier position of a possessive phrase referred to as a RP (RelatorP) andthe possessor (the antecedent) 'Milo' originates in a Predicate Phrase (PP)complement to the relator R (5b).

(5) (a)Milo heeft zich bezeerd.Milo has REFL hurt'Milo hurt himself.' (p. 55)

(b) [RP [DP zich] R+P [PP P [DP Milo]]]

The P (Predicate) head then incorporates into the relator R, and P+R incorporateinto the unaccusative verb, thus enabling it to assign case. The reflexive,which is merged with unvalued features, Agrees with the possessor antecedent,thus obtaining a set of valued (and shared) phi-features. The antecedent thenmust move out of the RP.

The authors propose that a complex (unlike a simplex) reflexive does notoriginate in a possessive DP but is ''merged as the internal argument of atransitive verb'' (p. 54) as in (6). The reflexive 'zichzelf' moves and adjoinsto the vP (6b), from where it c-commands its antecedent, thus enabling it toobtain a shared and valued set of phi-features.

(6) (a)Milo heeft zichzelf bezeerd.Milo has REFL hurt'Milo hurt himself' (p. 54)

(b) [vP zichzelf [vP Mil bezeer zichzelf]]

The authors go on to provide a variety of arguments to demonstrate that thesimplex Dutch reflexive is the internal argument of an unaccusative verb,whereas the complex reflexive is the internal argument of a transitive verb.

Chapter 4: Self-reflexives as Floating Quantifiers (pp. 116-152)

The authors begin by demonstrating that floating quantifiers, self-reflexives,and intensifiers share many properties, such as the need to be c-commanded by anantecedent, the requirement that an antecedent be local, and a ban on splitantecedents. They also can be used as arguments (7a) or adjuncts (7b).

(7)a. John shaved himself.b. John ate the pizza himself. (p. 125)

Floating quantifiers and intensifiers are argued to be base-generated in aposition adjoined to a vP, whereas a true self-reflexive is base-generated asthe object of a transitive verb (as proposed in chapter 2), from where it movesto adjoin to the vP edge. This proposal explains why ''in many languagesintensifiers and anaphors are identical in form (p. 139)''; they are bothelements that have unvalued phi-features and that must adjoin to a vP.Intensifiers and reflexives, however, do not always behave in the same mannerbecause reflexives are initially merged as arguments and intensifiers are mergedas adjuncts.

The authors then examine logophors, as in (8), which demonstrates the logophoricuse of 'himself'.

(8) Max_i boasted that the Queen invited Mary and himself_i for a drink.(p. 145, per Zribi-Hertz 1989)

The authors take the position that logophors are pronouns, merged into aderivation with a full set of phi-features, but which, in languages such asEnglish and Dutch, have the same pronunciations as anaphors due to ''meremorphological syncretism'' (p. 145).

Chapter 5: Extending the Analysis (pp. 153-187)

In this chapter, the authors develop an analysis of the behavior of simplex andcomplex Dutch reflexives in PPs. The simplex 'zich', unlike the complex'zichzelf', cannot occur as the complement of a functional P, as shown in (9).

(9)Fred luisterde naar zich*(zelf) op de radio.Fred listened to REFL on the radio'Fred listened to himself on the radio.' (p. 157)

In this case, 'zich' is the complement of P and does not originate inside of apossessive RP (see the discussion of chapter 2). Assuming that a functional PPis the complement of V, 'zich' is never able to establish a c-commandrelationship with its antecedent and thus its phi-features can never be valued.The complex reflexive 'zichzelf' is able to occur in this construction becauseit undergoes movement to adjoin to the vP, from where it is able to establish anAgree relation with its antecedent. Unlike in a functional PP, the simplex'zich' can occur in a spatial/temporal PP, as in (10a). As shown in (10b), thespatial/temporal PP is Merged in a position adjoined to the vP, from where'zich' is able to c-command and Agree with its antecedent.

(10) (a)Peter keek achter zich.'Peter looked behind himself'

(b) [vP [PP achter zich] [vP Peter keek]](p. 159)

The authors then discuss nonlocal uses of the simplex reflexive 'zich' (whichthe authors admit that they are unable to account for) and cross-linguisticvariation in reflexives.

Chapter 6: The Semantics of Simplex and Complex Reflexives: the Case of zich andzichzelf (pp. 188-230)

In this chapter, the authors develop a semantic analysis of Dutch simplex andcomplex reflexives. Following work by Coppieters (1982) and Bouchard (1995), theauthors separate DPs into I-Subjects and Concepts. I-Subjects have ''internaltemporal structure'' (p. 189) and demonstrate an individual and stage leveldistinction, whereas Concepts do not show this individual-stage leveldistinction. The simplex 'zich' can only occur with an I-Subject; 'zich' refersto ''a temporal interval'' (p. 196) and thus it must have an antecedent that hasinternal temporal structure. For example, (11a) favors the complex 'zichzelf',since the image of Freddy in the video recording is dissociated from Freddy'sactual self and functions as a Concept; the verb is also transitive. In (11b),'zich' is preferred since the image in the mirror is much closer to Freddy'sactual self, and functions as an I-Subject; the verb is unaccusative.

(11) (a)Freddy zag zichzelf/?*zich op de video-opname.Freddy saw REFL.self/REFL in the video recording.'Freddy saw himself in the video recording.'

(b)Freddy zag ??zichzelf/zich in de spiegel.Freddy saw REFL.self/REFL in the mirror.'Freddy saw himself in the mirror.' (p. 226)

Chapter 7: The Syntax of Spatial Anaphora (pp. 231-288)

This chapter focuses on ''snake sentences'' such as (12a-b), which containlocative PPs in which anaphors and pronouns do not show the usual complementarydistribution.

(12) (a) John_i saw a snake behind him_i.(b) John_i saw a snake behind himself_i. (p. 248)

To account for these types of constructions, the authors rely on an AxPart(Svenonius 2006) projection of a PP, which is ''a category like aspect ormodality'' (p. 240) that has axial features (features that refer to directions).The authors propose that ''pronouns lack grammatical axial dimensions'' (p. 247),but that 'self' contains axial dimensions. In (12a), 'him' does not have axialdimensions. Thus, the AxPart head must contain a valued set of axial features.In addition, the authors assume that AxPart is a variable that is bound by thespeaker, which gives this construction an ''observer-centered interpretation'' (p.248). This additional variable binding by the speaker essentially turns the PPinto a phase (if the R-expression and the pronoun were in the same phase,coreference would not be allowed). In (12b), AxPart has unvalued axial featuresthat become valued via an Agree relation with valued axial features of the'self' part of the anaphor. In this case, AxPart does not function as a variablethat is bound by the speaker, and this PP does not function as a phase. Theauthors go on to explain how these proposals account for the sensitivity ofsnake sentences to ''perspective, the nature of the location, andquantifier-binding'' (p. 246). The authors extend their analysis to account forsnake sentences in Dutch, which show differences from English, and to accountfor the behavior of anaphors in possessive constructions with 'have'.

Chapter 8: Conclusion (pp. 289-293)

This chapter summarizes the main arguments of each of the previous chapters ofthis book.

EVALUATION

The authors have succeeded in developing detailed analyses of a wide variety ofdata, some of which is very complex, without recourse to Binding Theory. I thinkthat in many ways (especially with respect to the extent of the data that theyare able to cover), their analyses achieve a greater level of explanatoryadequacy than does Binding Theory. The analyses presented in this book alsoraise a number of issues, some of which I touch on below.

The analyses require that lexical items appear in various positions in whichthey are not pronounced. For example, the authors argue that self-reflexivesmove to adjoin to a vP; this movement is necessary so that the anaphor canestablish an Agree relation with its antecedent (13a-b). However, it is notclear what motivates this movement, a point that the authors are aware of; theywrite ''[i]t is not clear to us at this point what drives the movement ofself-reflexives to the edge of vP'' (p. 138). In addition, assuming that aself-reflexive moves to the vP edge also requires that the verb undergo movementlater, as can be seen in (13c), which shows the underlying structure of (13a).

(13) (a) Pete invited himself.(b) [vP himself [vP Pete invited himself]](c) [TP Pete invited [vP himself [vP Pete invited himself]]](p. 138)

The authors' proposal that the Dutch simplex anaphor 'zich' and its antecedentoriginate within a possessive RP in which the anaphor c-commands the antecedent(see (5a-b)) requires the R-expression to undergo movement out of the RP, sincethe antecedent ends up c-commanding its antecedent at surface structure. Whetheror not all of this movement is required and the details of how these movementoperations are carried out are issues worthy of further investigation.

The analyses presented in this book require that a number of verbs alternatebetween unaccusative and transitive syntax. In constructions with the simplexDutch 'zich', the authors propose that the verb is underlyingly unaccusative(e.g., see (5a)). Although a variety of interesting evidence is given for thisproposal, further investigation of this issue may be warranted, especiallybecause constructions of this sort appear, on the surface, to be transitive,with an apparent subject (the antecedent R-expression) and object (anaphor).

The analysis of spatial anaphors presented in chapter 7 relies on the idea thatprepositions have axial features (features associated with directions) whichplay an important role in determining the behaviors of anaphors in snakesentences. While the snake sentence data are complex, whether or not theproposals regarding axial features of this sort, and the complexity that theybring with them, are on the right track, could be worthy of further research.

This work should be of interest to linguists with interests in syntax(especially those working in the Minimalist Program), morphology (especiallythose interested in Distributed Morphology), and semantics. This work shouldalso be especially of interest to those who are interested in coreferencephenomena and the behaviors of pronouns and anaphors. Although this work focuseson Germanic and Romance languages, it should be of interest to those who work onother languages; one may want to see if the analyses presented in this book canaccount for similar constructions in other languages. Overall, I recommend thisbook, since it examines a wide variety of interesting data and the analysesraise a number of interesting issues which could lead to much fruitful work inthe future.

REFERENCES

Bouchard, Denis. 1995. The Semantics of Syntax. Chicago: University of ChicagoPress.

Chomsky, Noam. 2001. Derivation by phase. In Ken Hale. A Life in Language, ed.Michael Kenstowicz, 1-52. Cambridge, MA; MIT Press.

Chomsky, Noam. 2004. Beyond explanatory adequacy. In Structures and Beyond, ed.Adriana Belletti, 104-131. Oxford: Oxford University Press.

Chomsky, Noam. 2008. On phases. In Foundational Issues in Linguistic Theory, ed.Robert Freidin, Carlos Otero, and Maria Luisa Zubizaretta, 134-166. Cambridge,MA: MIT Press.

Coppieters, Rene. 1982. Descriptions and attitudes: the problem of reference toindividuals. Studies in Language 6:1-22.

Dikken, Marcel den. 2006. Relators and Linkers. Cambridge, MA: MIT Press.

Halle, Morris and Alec Marantz. 1993. Distributed morphology and the pieces ofinflection. In The View from Building 20, ed. Ken Hale and Samuel Jay Keyser,111-176. Cambridge, MA: MIT Press.

Harley, Heidi and Rolf Noyer. 1999. State-of-the-article: distributedmorphology. Glot International 4:3-9.

Svenonius, Peter. 2006. The emergence of axial parts. Nordlyd: Tromso WorkingPapers in Linguistics 33:50-71.

Zribi-Hertz, Anne. 1989. Anaphor binding and narrative point of view: Englishreflexive pronouns in sentence and discourse. Language 65:659-727.

ABOUT THE REVIEWER

Jason Ginsburg is an Assistant Professor in the Center for Language Research at the University of Aizu in Japan. He has a PhD in linguistics from the University of Arizona. His research interests are in syntactic theory (in the framework of Generative Grammar), computational modeling of syntactic theory, and applications of syntactic theory and natural language processing for teaching languages.


Page Updated: 07-Apr-2012