Book announced at https://linguistlist.org/issues/34.217

AUTHOR: Barbara Citko
TITLE: Phase Theory
SUBTITLE: An Introduction
PUBLISHER: Cambridge University Press
YEAR: 2022

REVIEWER: Hassan Makhad

SUMMARY

Citko (2014) offers an overview of Phase Theory (PT). The book consists of eight chapters. The introduction examines literature on PT. It postulates that the language faculty (FL) intermingles with two cognitive interface systems (Conceptual-Intentional (CI) and Articulatory-Perceptual (AP)) at particular stages during derivations. These junctures are phases. When a derivation reaches a phase, it becomes inaccessible for any supplementary computations of narrow syntax. The author specifies that some PT attributes face controversial challenges. The section ends with a brief summary of the book chapters.
Chapter one introduces the Minimalist Program (MP). It addresses the overall architecture of MP, specifying that Language encompasses narrow syntax, phonological, and semantic mechanisms. The process involves applying Merge and Agree on a given Lexical Array (LA). The output results in a <PHON, SEM> representation, which is accessible to the two external language systems. It shows that LAs are managed by Merge operations. External Merge combines LA items. It is regulated by the Inclusiveness Condition, which excludes adding any material which is not included in the input. Internal Merge involves remerging items into new positions. The chapter also discusses Features. A feature can be interpretable (iF) or uninterpretable (uF). A given feature is (un)interpretable with reference to its relevance to an interface or its location. The chapter then examines Agree, a syntactic operation which values unvalued features. Agree needs to fulfill four conditions: activity, matching, domain, and locality.

Chapter two offers an overview of PT. It begins by tackling the Merge over Move (MOM) preference. MOM claims that merge is favored over move. The subsection demonstrates that if MOM is not respected, ungrammaticality results. The second part deals with the incentives behind PT. It shows that the concept of phase was introduced as a resolution for MOM. The latter applies within subsets of lexical items rather than across subarrays. The third part addresses properties of PT. A phase is a stratified syntactic object with semantic content. Hence, CPs as well as transitive and unergative vPs are phases. The author indicates that phase heads are positions of uninterpretable features, which foster syntactic processes. The fourth part tackles the Phase Impenetrability Condition (PIC). It states that once phases are completed, they become inaccessible to syntactic operations. The domain of a phase is opaque, while its edge is not. The author notices that there is disagreement in the literature as to whether PIC applies across the board to the procedures of Move and Agree. Long Distance Agree has been argued to disregard PIC, in some languages. This issue raises questions about the timing of transfer operations. The reason is that after a phase complement is transferred to the interfaces, it is no longer available to the syntax. The fifth part examines the notion of Multiple Spell-Out. Since a clause contains more than one phase, its computation needs to include multiple applications of Spell-Out. The transfer to interfaces may be asynchronous, i.e., not necessarily applying to both CI and AP at the same time. The sixth part explores Feature Inheritance (FI). It states that phase heads have the privilege of hosting uninterpretable features. This makes them trigger derivations. The uninterpretable features of phase heads may be passed on to non-phase heads in the course of derivations. For example, Tense (T) inherits uninterpretable features from the complementizer (C) subsequent to C merger. Likewise, FI concretizes the likelihood of distinct syntactic procedures to take place simultaneously at phases. One such synchronized process is feature valuation and deletion. Uninterpretable features are valued and concurrently deleted to maintain the dissimilarity between patterns of feature types.

Chapter three provides evidence for phases. It is made up of four sections. The first acknowledges the relevance of phasehood. It shows that a phase is complete at the interfaces. It is differentiated in accordance with interface sensitivity, as well as syntactic involvement. The second section offers an examination of PF properties of phases. One assumption is that a phasal category needs to be autonomous and able to undergo movement. The other hypothesis requires phase heads to trigger Spell-Out. In this regard, sluicing allows the complement of C (TP) to be deleted. Verb phrase ellipsis eliminates the complement of little v. Nominal ellipsis omits the complement (NP) of determiners (D). The third section surveys LF peculiarities of phases. It establishes that phases must be both complete and interpretable. The author offers reconstruction and quantifier raising processes as concrete identifiers which entail semantic relevance in PT. Additionally, supplementary LF qualifiers relate to existential closure and clause quantificational partition. The fourth section deals with syntactic idiosyncrasies of phases. They encompass syntactic movements relative to the PIC. The issue concerns the interpretation, articulation, and stranding of material at the edge of an XP after a movement operation. Additional queries concern the connection of uninterpretable features with phase heads and feature valuation within XP domains. Another theme addressed is the magnitude of island effects in PT. For a category to be an island depends on the Condition on Extraction Domains (CED), which is regulated by the PIC.

Chapter four explores typical phases: CPs, vPs, and DPs. The first part considers evidence for the phasehood of CPs. It builds on the assumption of the edge property of phases. The latter assumes that successive-cyclic wh-movements target every intermediate CP position, because they are phases. The evidence for CP phasehood comes from wh-islands. Languages which do not obey the Wh-Island Condition allow multiple specifiers. To identify the edge property of a CP, the author examines whether a moved element may be pronounced, interpreted, and able to leave material at the CP edge. Likewise, she addresses the properties of C, in terms of features and triggering Spell-Out. Additional evidence for CP-phasehood comes from reconstruction effects, where interpretation of a displaced category occurs lower down in the chain rather than where it is spelled out. This property shows that successive-cyclic movement operates via the edge of CP. The author shows that C has uninterpretable features either in matrix or embedded contexts. This provides suggestive evidence that C is a phase head. The final argument in favor of CP phasehood derives from the fact that the complement of C can undergo ellipsis. This is what happens in sluicing constructions, where wh-phrase movement is followed by the deletion of the complement of C; i.e., TP. The second subdivision reviews evidence which suggests the phasehood of little vPs. The author distinguishes four types of vPs: transitive, unergative, passive, and unaccusative. The last two do not subcategorize for an external argument and do not discharge accusative Case. In transitive and unergative clauses, v values accusative Case. Likewise, VP-ellipsis involves the omission of the complement of v. Reconstruction, i.e., wh-interpretation at the vP edge, provides confirmation for vPs being phases. Another argument comes from Antecedent-Contained Deletion (ACD). To avoid infinite regress in ACD constructions, the system resorts to Quantifier Raising (QR), targeting adjunctions to vP. The author shows that there is a controversy over whether vPs containing passive and unaccusative verbs are phases or not. She shows that raising out of the complement of v to the specifier of vP does not violate antilocality. Therefore, passives and unaccusatives operate indistinguishably from transitives and unergatives with regards to movement out of vP. This is supported by reconstruction at the vP edge in passives. The same position is proven to be appropriate for hosting QR. However, passives and unaccusatives fail to respect the other diagnostics. They are referred to as weak phases, while transitives and unergatives are strong phases. The third section surveys DPs as phases. DPs have uninterpretable case features, which is a property of phases. DPs comprise domains of feature valuation. This fact is exemplified by the adnominal Genitive Case in Polish and Slavic numerals. DP phasehood also comes from the edge property. Polish shows that internal elements of DPs agree in Case and φ-features. When an element moves out of a DP, this agreement is preserved, which shows DPs to be phases. Likewise, the DP edge offers an escape hatch for possessor extraction in Hungarian. This illustrates that movement out of a DP passes through its edge. Such displacement is blocked when Spec DP is occupied. Moreover, the complement of D may undergo deletion.

Chapter five explores extending phasehood to other categories. It consists of three sections. The first deals with Predication Phrases (PrP). The predicate head is a phase. Movement out of PrP proceeds through the edge of PrP. Predicative Case (instrumental) in Polish and Russian predicate nominals is determined by the Pr head valuing the uninterpretable Case feature of the predicate. In addition, extraction of the complement of Pr reveals that only instrumental complements can undergo wh-movement. QR, negative polarity licensing, and ACD, in copular constructions, provide additional support for adjunction to PrP as a phase. Lastly, PrP fails to determine Spell-Out, due to the absence of the complement of Pr head deletion. The second section concerns broadening PT to incorporate Prepositional Phrases (PP). The author shows that Polish locative and directional prepositions value Case. Moreover, raising of the P-complement in Dutch proceeds through the edge of PP. When the latter is filled, extraction is blocked. The third section extends phasehood to Applicative Phrases (ApplP). The assumption is that, in double object constructions, applicative heads project with direct objects in the complement position and indirect objects as specifiers. This high applicative permits the direct object to move above the indirect one. Similarly, passivization in high applicatives applies within the edge of ApplP. The same procedure is involved in wh-movement of the direct object. Besides, complex wh-phrases may strand a nominal when moving through the edge of ApplP. Finally, the complement of the high applicative head can be deleted. Accordingly, high ApplPs are phases.
Chapter six explores further differences in phasehood. It consists of three sections. The first examines the possibility of altering the phasal status of categories. Head movement is the process responsible for phase changeability. Movement of a phase head may change its phase properties.The second addresses crosslinguistic variation over phasehood. The author states that LF criteria on phases are not open to crosslinguistic variation. Yet, PF principles for determining Spell-Out are. She points out that phasehood discrepancy may result from lexical gaps, by-product effects, and parametric setting variations. The first distinction occurs when a language lacks a category which is universally a phase head. This possibility has been demonstrated by passivization in high applicative structures, as well as predicative Phases in Polish and Russian. Another instance relates to the status of DP crosslinguistically. While some languages have articles, others lack them. The parametric setting distinction assumes the existence of variation, the outcome of which is regulated by parameter settings. Nonetheless, the author observes that Phase may not be a feature similar to other syntactic features. Therefore, parametric variation over phasehood may not be a solvable issue. The last topic relates to non-simultaneous phasehood. It addresses potential incompatibilities between phonology and semantics with respect to what is valid as a phase. Given the possibility of multiple Spell-Out, Transfer to PF and LF can be asynchronous. A case in point is scrambling in Japanese, which requires Transfer to LF prior to PF.

Chapter seven explores syntax-phonology and syntax-semantics intersections. It consists of two subsections. The first concerns phases and PF. PF operations are sensitive to phases. For example, linear order of syntactic constituents is spelled out cyclically phase by phase. This procedure imposes the preservation of the established linearization of components. The produced orders are not altered by Spell-Outs. The author shows that nuclear stress assignment is susceptible to syntactic domains. This susceptibility is reflected in the mapping between phase domains and prosodic domains. She calls attention to joining nuclear stress assignment to Spell-Out domains to eliminate extrametricality. The cycle involves spelling out the complement of a phase head when the next phase is merged. The process applies stress rules and projects the uppermost stress to the next level. Other phonological processes make use of phases. These include High Tone Anticipation analysis in Luganda, as well as prosodic structure of high and low applicatives in Bantu languages. The second subsection addresses the mapping between syntax and semantics. The author shows that there is a semantic difference between vP and CP phases. The distinction relates to nuclear scope and restrictive clause in quantificational structure. The idea is that vP phase maps things into nuclear scope, while CP maps to restrictive clause. In this sense, existential bare plural subjects are interpreted in Spec vP and generic ones in Spec TP. Objects seem to maintain the same logic of reasoning. Indefinite objects exhibit confusion between specific and existential readings. This ambiguity relates to their syntactic positions. Within the vP phase, such objects have an existential meaning, but within the CP phase the existential interpretation becomes unavailable. These facts demonstrate that vP and CP boundaries are considerable in semantic interpretations.

Chapter eight is a concise summary of the previous chapters. It reviews the central ideas presented in the work and recapitulates the content of each chapter by showing its key points, together with the justifications offered for the addressed issues.

EVALUATION

The book is an important piece of work. It is valuable for anyone who is interested in PT. It stands out for its comprehensive approach. The style is lucid and the concepts are discussed in a simplified manner. The discussion offers a wealth of facts about the addressed issues. The reviewed literature comprises both contemporary and dated works. In its entirety, the book is highly interesting, as it deepens our understanding of PT. The relevance of the book lies in the fact that it expands PT comprehension by its thought-provoking considerations, its simplification of the complexity of the discussed topic, and its ground-breaking examination. For instance, the discussion of two PIC versions is a great contribution to the MP model. The traditional PIC, Citko’s PIC1, forbids any movement of a phrase (XP), or out of it, once the dominating phrase (ZP) is complete. The former phrase is Spelled-out as soon as ZP merges with another head. Nonetheless, the second version of PIC, her PIC2, is only sent to Spell-Out when the next phase head is merged. This proposal of PIC2 has indeed provided an adequate account for a number of unexplained phenomena. A case in point is languages which tolerate P-stranding with A-movement, while prohibiting them with A-bar-movement. A more concrete example involving the A/A-bar difference relates to wh-movement. When the specifier of little vP hosts the subject of the sentence, it does not prevent wh-elements from moving out of vP. However, when the specifier of a CP is filled with a wh-phrase, it impedes other wh-elements from targeting that CP position.

Overall, this is an incredibly comprehensive book on PT. It is relevant for anyone seeking an exhaustive familiarization with the topic. It examines the fundamental motivating mechanisms of phasehood. It is a valuable resource for students, researchers, and scholars interested in understanding PT. Given its accurate research methodology and insightful analysis, the book is an essential reading in the field of MP syntax.

ABOUT THE REVIEWER

I am a Professor at Cadi Ayyad University, the Poly-disciplinary Faculty of Safi. My main interests are the syntax and morpho-syntax of Tashelhiyt Berber, Moroccan Arabic, and Literary Arabic. I have great interest in the areas of functional categories and language contact.