LINGUIST List 15.282

Mon Jan 26 2004

Review: Syntax: Karimi (2003)

Editor for this issue: Naomi Ogasawara <>

What follows is a review or discussion note contributed to our Book Discussion Forum. We expect discussions to be informal and interactive; and the author of the book discussed is cordially invited to join in.

If you are interested in leading a book discussion, look for books announced on LINGUIST as "available for review." Then contact Sheila Dooley Collberg at


  • Mohammad Rasekh, Word Order and Scrambling

    Message 1: Word Order and Scrambling

    Date: Sun, 25 Jan 2004 23:53:02 -0500 (EST)
    From: Mohammad Rasekh <>
    Subject: Word Order and Scrambling

    Karimi, Simin, ed. (2003) Word Order and Scrambling, Blackwell Publishers, Explaining Linguistics 4.

    Announced at

    Mohammad Rasekh Mahand, Linguistics Department, Bu-Ali Sina University, Hamadan, Iran.


    The book under review introduces readers to recent research into the linguistic phenomenon called scrambling, or free word order. Also it explores major issues including factors responsible for word order variations, how scrambled constructions are processed, and whether variations are available in early child language development and in second language acquisition. The book provides enlightening information on different aspects of word order variation and the consequences for our understanding of the nature of human language. Simin Karimi is Associate Professor of Linguistics at the Arizona University and on the editorial board of the journal Linguistic Analysis. Her main interests are syntax and syntax-semantics interface. Most of her writings are about Persian syntax.


    Scrambling, or free word order, is the feature of some of the languages. It was firstly believed to be a stylistic rule, and there are two approaches in studying scrambling: base-generated or syntactic movement. Different studies have indicated that the various sorts of scrambling are the instances of A(rgument)-movement or A'(non- argument)-movement. Optionality of scrambling, first advocated by many linguists, has become a challenge for the Minimalist Approach (Chomsky 1995), a theory that allows syntactic movement only if they are triggered by morphological factors. Some new solutions are proposed to solve the optionality problem. In recent years, the main trends of research in this area include the discourse effects of scrambling and its relation to Topic/Focus, the semantic effects of scrambling, its acquisition and the interaction between scrambling and prosody.

    The book under review is the result of International Conference on Word Order and Scrambling held on April 7-9, 2000, in Tucson, Arizona.

    The first paper in this volume written by Kenneth Hale, Eloise Jelinek, and Maryann Willie, entitled 'Topic and focus positions in Navajo' addresses some problems in Navajo sentence structure and argues that this language belongs to a parametric class of languages that exhibits two features at spell-out: nominals and quantifiers are in operator scope positions at spell-out; so, it is a Discourse Configurational language, and it is a Pronominal Argument language; in which these operators c-command and bind overt pronominal variables in argument positions at spell-out. The writers have indicated that the order of the nominal operators reflects Topic/Focus structure; while, the pronouns are ordered according to their grammatical relations. It is also argued that an Np in a left-peripheral position may be in the Contrast operator position.

    The second paper 'Argument scrambling, operator movement, and topic movement in Hungarian', written by Katalin �. Kiss demonstrates the different nature of free word order in three fields of Hungarian sentence. It is shown that the word order of Hungarian in postverbal position is affected specifying feature of arguments. The middle field of the sentence consists of a strict hierarchy of an aspectual operator, a focus, and distributive quantifiers. The topic field of the Hungarian sentence harbors constructions which are not operators but externalized arguments, functioning as the subjects of predication.

    The third is 'Grammatical relations in Tohono O'odham: an instrumental perspective' written by Mizaki Miyashita, Richard Demers and Delbert Oritz. Tohono O'odham is a pronominal argument language in which pronominals are discourse variables and their positions in a sentence is not the result of movement in the traditional sense. The primary mechanism for interpreting ambiguous sentences in this language is the features of discourse.

    The fourth article is 'Bare nominals, non-specific and contrastive readings under scrambling' written by Veneeta Dayal. It explores empirical validity of the generalization that scrambling of indefinites correlates with the loss of non-specific readings. It is argued that if contrastive readings are non-specific, the generalization has to be restated to prohibit non-specific indefinites from scrambling without the additional support of contrast. It is also suggested that leftward scrambling forms a Ground: link structure.

    The fifth article is 'On object positions, specificity and scrambling in Persian' by Simin Karimi, which examines the syntactic, semantic and morphological asymmetries between specific and non-specific direct objects in Persian. She proposes two object position hypothesis for specific and non-specific objects in this language and explains why scrambling applies freely to the specific objects but is restricted in the case of non-specific ones. A third object position, created by scrambling, is also proposed and it is argued that scrambling is triggered by discourse features such as topic and contrastive focus.

    'Scrambling, subscrambling and case in Turkish' is the sixth article, written by Jakling Konflit. The paper tries to look at some new facts about scrambling out of larger DP's, or subscrambling, and to use Turkish facts to show that there is no specificity effect per se, as an independent principle of grammar. The writer proposes that incorporation can have different dimensions in different languages, but in ways that should be limited. He believes that while certain properties of incorporation could be observed in Turkish, but there are some properties which are not found in this language, like changes of thematic structure of the verb due to incorporation, and complete morphological merge of the incorporated element into its host.

    John Frederick Bailyn's article is 'Does Russian scrambling exist?' It tries to argue against the necessity of positing a process of scrambling in accounting for Russian free word order. Instead, he argues that a subset of the relevant phenomena are related to a purely syntactic process of inversion, a kind of raising to subject, and the rest related to focus which is represented in a unique sub-component of the interpretive interface. He concludes that A-scrambling is in fact generalized inversion and A'-scrambling is dislocation, a prosodic movement related to information structure.

    'A-scrambling and options without optionality' is the next paper by Shigeru Miyagawa. He tries to argue that A-scrambling of the object is Extended Projection Principle (EPP)-driven, and is not an optional movement. He argues that the situation in which a language allows an option is optional. The independent properties essential for EPP- scrambling are verb raising and the occurrence of morphological case marking.

    Helen de Hoop's 'Scrambling in Dutch: optionality and optimality' is the next article. It gives an optimality theory (OT) analysis for truly optional scrambling of definite objects in Dutch. He introduces two constraints (1) stay: no scrambling and (2) surface correspondence 1 (SC1): definite NPs scramble. The data show that anaphoric definites scramble in two-thirds of the cases, and anaphoric definites do not scramble in one-third. Also, non-anaphoric definites scramble in 50 per cent of the cases, and non-anaphoric defenites do not scramble in 50 per cent of the cases. The analysis captures the general tendency of anaphoric definites to scramble.

    The tenth is 'Word order and (remnant) VP movement' by Anoop Mahajan. He uses the antisymmetry approach to word order variations suggested by Kayne and suggests a different way of implementing the SOV versus SVO difference. He suggests that V-to-I movement should be eliminated from the syntactic component. He argues that the OV versus VO distinction may be triggered by the presence of a Determiner feature (for objects) that forces the formation of a VP remnant in OV languages leading to stranding of the object prior to VP-movement to [SPEC,IP].

    Vaijayanthi Sarma has written 'Non-canonical word order: topic and focus in adult and child Tamil'. Tamil is also a scrambling language. The writer has identified two extraction procedures, leftward and rightward, which have the properties of movements to non-argument or non-lexical positions. It is argued that scrambling in Tamil mimics topicalization or clefting. It is shown that focusing strategies are independent of the topicalization strategies. Children are aware of the case restrictions on scrambling and use order shifts to signal interpretive changes like adults.

    The next paper is 'L2 acquisition of Japanese: knowledge and use of case particles in SOV and OSV sentences' written by Noriko Iwasaki. It is observed that, even if L2 learners had knowledge of scrambling, they did not always accurately produce scrambled sentences. The gap between L2 learner's knowledge and performance was not random performance failure, instead it revealed that L2 Japanese speakers use a canonical sentence template as a processing strategy.

    'Scrambling and processing: dependencies, complexity and constraints' is written by Irina Sekerina. The paper tries to study how scrambled sentences are processed. The chapter tries to compare the sentence processing experiments conducted in German, Japanese, Finnish, Serbo- Croatian and Russian, and their implications for the linguistic and psycholinguistic theories of scrambling are discussed.

    The last article in this volume is 'WH-movement versus scrambling: the brain makes a difference' by Angela D. Friederici, Matthias Schlesewsky, and Christina J. Fiebach. In this paper two constructions, WH-movement and scrambling in German, are compared. They present a number of studies using event-related brain potentials which show that WH-movement and scrambling elicit different brain responses. Finally, it is concluded that scrambling in German induces a local syntactic violation while WH-movement does not.


    The volume under review consists of very well written papers on scrambling in a variety of languages. The number of languages covered in this study is considerable. Discussing scrambled structure from semantic, syntactic, pragmatic and psycholinguistic views is something novel and the book is a milestone from this perspective. The book clearly shows the current status of scrambling studies and it also highlights the main problems confronting linguists in studying scrambling.


    Mohammad Rasekh Mahand is Assistant Professor of Linguistics at Bu-Ali Sina University, Hamadan, Iran. His research interests include syntax, syntax-pragmatics interface and typology.