EDITORS: Zwart, Jan-Wouter; de Vries, Mark
TITLE: Structure Preserved
SUBTITLE: Studies in Syntax for Jan Koster
SERIES TITLE: Linguistik Aktuell / Linguistics Today 164
PUBLISHER: John Benjamins
Atakan Ince, MLS, MD, USA
Jan-Wouter Zwart and Mark de Vries' ''Structure Preserved: Studies in Syntax for
Jan Koster'', edited in honor of Jan Koster, is a collection of original squibs
written by Jan Koster's students, colleagues and friends. There are 41 chapters,
mostly between 8 and 10 pages in length.
The book starts with a preface by the editors, where they state that this volume
is edited for Jan Koster and his contributions to linguistics (in Europe).
Then, it gives a list of Jan Koster's printed/to appear/ unpublished works, as
well as works edited by him, and interviews with him.
Chapter 1: 'Unaccusative verbs in Chinese: Derivation in the lexicon or in
syntax?', by Werner Abraham. (1-12)
Abraham argues that unaccusative (ergative) verbs in Chinese are derived in
syntax. For this, he applies tests from other languages such as non-occurrence
of (a) agentive nominalization with, (b) passivization of, and (c) imperative
form of unaccusative verbs -pointing out their inadequacy for a complete
conclusion-, and then shows that the past participle forms of unaccusative verbs
turn up as an adjectival modifier to a direct object and applies
Chinese-specific tests. The arguments in the squib are quite clear and
convincing. However, although the title has 'unaccusative', he uses the
equivalent term 'ergative' in the rest of the squib.
Chapter 2: 'Gapping is always forward', by Peter Ackema. (13-19)
In this squib, Ackema comes up with an apparent problem with Koster's (2009a)
Principal of Natural Word Order, which requires that in a dependency relation an
antecedent must precede the dependent, which is a component of his
'configurational matrix'. Ackema remarks that whereas both gapping and
determiner sharing are forward in (verb-object) (VO) languages, gapping is
backward but determiner sharing is forward in OV languages, such as Japanese and
Korean. Ackema puts forward an analysis of gapping for Japanese (J) and Korean
(K), where it is forward as in VO languages. In his analysis, in J/K verbs, VP
conjuncts undergo Across-the-Board movement (ATB) to Tense/Complementizer, after
which V in the second conjunct is elided. As in English, the 0 head licenses
determiner sharing in the second conjunct (a. la. Ackema & Szendroi, 2002).
Therefore, Koster's Principal of Natural Word Order is preserved.
Chapter 3: 'Focus particle doubling', by Sjef Barbiers. (21-29)
Barbiers argues that the structure of Dutch middle field (the domain between
Complementizer and verb (-cluster) in Germanic languages) is more complex than
assumed in Koster (2008), giving evidence from focus particle doubling. In this
domain, semantically-triggered internal movement occurs. For Barbiers, focus
particles move to a functional position in the middle field to find their second
argument because they are quantifiers which need two arguments. In the case of
doubling, both copies of the focus particle are pronounced.
Chapter 4: 'Wh-drop and recoverability', by Josef Bayer. (31-39)
Bayer analyzes wh-drop in Bavarian, after documenting some parallelisms between
wh-drop and pronoun-drop. He notices that the only wh-element that can be
dropped is ''wos'' ('what'), which is [-human]. He reasons that only this
wh-element can be dropped because it is the maximally underspecified wh-element
in Bavarian. The availability of interrogative interpretation of a sentence is
due to (a) its rising intonation pattern; and (b) the presence of the clitic -n
attached to the verb in wh-questions.
Chapter 5: 'Two futures in infinitives', by Janneke ter Beek. (41-48)
In this squib, ter Beek argues that Dutch infinitivals support both of the
competing theories of future tense/orientation in infinitivals: the first theory
assumes a covert modal component, and the second theory assumes both tense
component and a covert modal component. The future orientation of propositional
infinitivals in Dutch involves both the tense and the modal component, whereas
that of irrealis (i.e. non-assertive/future-oriented) infinitivals involves only
the modal component. Her main evidence for this difference is that the auxiliary
''zullen'' ('will') occurs only with propositional infinitivals.
Chapter 6: 'A dynamic perspective on inflection', by Hans Bennis. (49-56)
Contra Koster (2009b), Bennis argues that biological conditions restrict the
acquisition of certain structural properties of language by showing that age is
a decisive factor for inflection strategies in language acquisition. His
evidence for this is that it is hard to acquire inflection after the critical
period, as in L2-learning and creoles. He also handles deflection, or loss of
inflection. For Bennis, deflection, closely correlated with L2-learning, is a
consequence of change in feature structure, which may be related to economy
conditions in language change, or, as he puts it, 'a consequence of biologically
determined principles' (p. 55)
Chapter 7: 'Is there ''preposition stranding in COMP'' in Afrikaans? No way!', by
Hans den Besten. (57-63)
Den Besten argues that Afrikaans is not a preposition stranding language, contra
du Plessis (1977), who claims that wh-movement in Afrikaans can strand
prepositions in intermediate cycles. Den Besten contends a paratactic analysis
in which a 'matrix clause'-like structure, such as 'I think', is inserted into
the existing structure between a preposition and its wh-complement.
Chapter 8: 'Restructuring verbs and the structure of Spanish clauses', by
Reineke Bok-Bennema. (65-72)
Bok-Bennema puts forward an analysis of the following properties of Spanish: (1)
A sequence of one or more restructuring verbs and a lexical verb can occur to
the left of the subject; and (2) this type of verbal cluster can be broken by
adverbs and complementizer-prepositions. She assumes that there is a Predicate
Phrase (PredP) located between the final landing site of the finite verb (i.e.,
T) and light verb phrase (vP). Under her analysis, a restructuring verb moves to
T, and a complementizer phrase (CP) (headed by a complementizer-preposition)
moves to the specifier position (Spec) of PredP. All the intervening adverbs are
generated between TP and PredP. These movement steps derive the properties of
Spanish provided above.
Chapter 9: 'Cantonese as a tense second language', by Lisa Lai-Shen Cheng. (73-79)
Cheng argues that Koster's (2003) Tense Second Constraint, which requires that
all languages mark tense/(clausal) type in the ''second'' position (C) of the main
clause, is also true for Cantonese, which is assumed by others to lack tense,
and therefore TP. Following Sybesma's (2007) proposal that Mandarin has a tense
node but lacks an overt tense agreement marker, Cheng proposes that
complementizer particles carry tense information. Since TP complement of a CP
moves to Spec, CP, it precedes C-elements, and these C-elements occur in
Chapter 10: 'On a selective ''violation'' of the complex NP constraint', by
Guglielmo Cinque. (81-89)
In this squib, Cinque handles why the Complex Noun Phrase Constraint (CNPC) does
not hold up in certain languages while it does in others. CNPC does not hold up
only in structures where the head of a relative clause is indefinite &
nonspecific, the verb of which the head is an argument is an existential verb,
etc., and the position relativized in the relative clause from which a
constituent is extracted is the subject, he notes. The crux of his proposal is
that the CNPC structures out of which a phrase can be extracted are not
''ordinary'' relative clauses. Since the head of these relative clauses is an
indefinite nonspecific, it is not a determiner phrase (DP). Therefore, nothing
blocks extraction independently. Also, when the relative pronoun is not an
''ordinary'' one (which, who, etc.), the higher Spec of CP is available as an
escape hatch. Noting that CNPC does not hold in English and Romance languages
when these conditions hold, he concludes that CNPC is not parameterizable.
Chapter 11: 'Dressed numerals and the structure of universal numeric
quantifiers', by Norbert Corver. (91-99)
Corver proposes that Universal Numeric Quantifiers in Dutch (e.g. ''alle drie''
('all three') in ''alle drie de studenten'' ('all three students')) are phrasal
constituents merging low in DP and then raising to Spec, DP as a result of
DP-internal displacement. His evidence for DP-internal movement is from ''dressed
(i.e. inflected) numerals'', or the heads of Number Phrases (NumP), which carry
the morpheme -e. For Corver, this morpheme is a reflex of Spec-Head agreement.
Chapter 12: 'Embedded inversion and successive cyclicity', by Marcel den Dikken.
In this squib, den Dikken questions the validity of Belfast English and Romance
languages as evidence for successive-cyclic movement via Spec, CP. Since
embedded inversion does not occur in relative clauses, he remarks, it is
doubtful that it's the wh-trace in embedded Spec, CP that triggers embedded
inversion. The Belfast English data, he proposes, supports wh-scope marking or
wh-copying constructions, where an embedded clause is also a question. As for
the Romance data, where embedded inversion occurs in all the embedded sentences
but the clause in which the operator originates, he concludes that these
embedded inversion facts support movement of the operator to the intermediate
Spec, vP positions, since inversion is not possible with wh-adjuncts, which are
assumed to merge above vP. Also, he proposes that inversion does not occur in
the clause in which an operator originates because the operator originates in
the next higher clause and binds a (null) resumptive pronoun in the lower clause.
Chapter 13: 'Little words don't lie: X' have initial Xo', by Joseph Emonds.
In this squib, Emonds looks at certain 'short modifiers' that can or must
precede the head of the phrase they modify. He observes that these modifiers can
neither take a complement nor undergo displacement. He proposes that these
Adjective+ly (A-ly) adverbial modifiers are heads that merge with another
null/overt head in order to be recategorized rather than adjoin to or merge in
the Spec position of the phrase that they modify.
Chapter 14: 'Repairing head-to-head movement', by Arnold Ernest Evers. (125-130)
Evers looks at recursively stacked verb-argument constructions in Dutch,
focusing on why pied-piping (cases of a moving element dragging along a larger
phrase in which it occurs) does not occur when the verb raises to complementizer
(Co). He proposes that pied-piping is restricted to A-movement, but A'-movement
does not allow pied-piping. Therefore, since Vo-to-Co movement is A'-movement,
pied-piping cannot occur.
Chapter 15: 'On the duality of patterning', by Jordi Fortuny. (131-139)
Fortuny looks at the linguistically universal fact that linguistic expressions
can be decomposed into morphemes -minimal meaningful units- that can be
decomposed into phonemes -minimal distinctive but meaningless units, or what
Hockett (1958, 1960, 1963) refers to as duality (of patterning). He states that
the duality of patterning makes language efficient with respect to data
compression, data transmission and structural complexity.
Chapter 16: 'Reflexive cartography', by Elly van Gelderen. (141-148)
In this squib, van Gelderen looks at English reflexives from a language change
perspective and proposes that reflexives change as a consequence of the pronoun
reanalysing from noun (N) to determiner (D) due to economy principles. She gives
the steps of the change as follows: In Old English, both pronouns and reflexives
were moving to D and therefore DPs; in Middle English and Modern English,
reflexives ceased moving to D and stay in N position; and in Modern English, a
NumP also appears in DP, selecting NP as its complement and being selected by D
as a complement.
Chapter 17: 'What does eye-tracking reveal about children's knowledge of
linguistic structure?', by Petra Hendriks. (149-155)
Hendriks looks at Sekerina et al. (2004) and states that one cannot conclude
that eye-tracking is more sensitive to children's knowledge than off-line
methods based on the evidence that they provide. She states that the problem
with picture-selection tasks is that they are forced tasks, which does not mean
they are less sensitive to children's knowledge than eye-tracking.
Chapter 18: 'Depictives and the word orders of English and Dutch', by Herman
Heringa focuses on why the order of subject-oriented and object-oriented
depictives are the mirror image of each other in Dutch and English. He proposes
that this difference is due to the overt V-to-v movement and movement of vP over
the subject-oriented depictive phrase in English. This explains why depictives
are preceded by the verb in English but followed by the verb in Dutch, as well
as the mirror image order of subject-oriented and object-oriented depictives in
Dutch and English.
Chapter 19: 'Feature percolation in the Dutch possessive', by Jack Hoeksema.
Hoeksema tackles why complex phrases with determiners are possible in the
possessive construction only when their left member is also possessive in Dutch.
He contends that a phrase in Spec, DP agrees with D in the [possessive] feature,
and then this feature is passed onto the D head of the phrase in Spec, DP. Due
to the passing of this feature, this D head is also pronounced as a possessive
Chapter 20: 'On the interruption of verb-raising clusters by nonverbal
material', by Eric Hoekstra. (175-183)
Hoekstra proposes that the fact that head-initial clusters can be broken up by
nonverbal material but head-final clusters cannot be broken up in Old English
and Old Frisian, which demonstrates absence of a mirror effect, provides support
for Kayne's (1994) Linear Correspondence Axiom because head-final structures
obey stricter spell-out rules since they are derived by internal merge (i.e.
Chapter 21: 'Referring to yourself in self-talk', by Anders Holmberg. (185-192)
In this squib, Holmberg characterizes and explains the constraints on when 'I'
and 'you' can be used in self-talk. He states that there are two aspects of the
'self' involved: the mindless self, which can be expressed by both 'I' and
'you'; and the 'thinking & feeling' self, which can be expressed only by 'I'. He
puts forward a feature-based expression of the use of these pronouns in
self-talk. 'I' bears the feature (bundle) [+Self's Mind], which occurs in the
highest subdomain of the C-domain of finite clauses and denotes the mind of the
speaker, while 'you' cannot bear this feature (bundle)
Chapter 22: 'Case-agreement: diptotic evidence in construct state noun phrases',
by Riny Huybregts. (193-205)
Huybregts, looking at the diptotic nominals in Modern Standard Arabic, concludes
that AGREE is a recursive operation in the computational core and that case
valuation is linked to AGREE. Therefore, case and agreement are different
reflexes of the same mechanism: Case-Agreement.
Chapter 23: 'Syntactic predictions in second-language sentence processing', by
Edith Kaan, Andrea Dallas and Frank Wijnen. (207-213)
Summarizing previous studies, Kaan et al. report that predictive parsing in
second-language processing is proportional to proficiency. Although beginning
second-language learners cannot use information to make predictions about
upcoming syntactic structure, more proficient learners can use information for
Chapter 24: 'Notes on French and English demonstratives', by Richard S. Kayne
and Jean-Yves Pollock. (215-228)
Kayne & Pollock show the parallelisms between demonstratives in English and
French: demonstratives in both languages include an abstract THING in their
composition (this/that: th- (definite article) + -at/-is (deictic element) +
THING; celà/ceci: ce + THING + là/ci). They also state that in both languages
demonstratives require the presence of a noun because they merge as part of a
larger DP out of which they are extracted to the subject position in
Chapter 25: 'The accusative infinitive in Latin, English and Dutch', by Maarten
Klein states that it is not necessary to look for explanations for the absence
of certain structures and proposes that this could also be a matter of
habituation and/or culture. In other words, the absence of certain structures is
coincidental, and they can be learned. One such case is the lack of the
Accusativus Cum Infinitivo in Modern Dutch, which can, however, be optionally
used by native speakers. Klein concludes that language changes always occur at
the level of language culture, not at the level of language nature.
Chapter 26: 'Identifying in Dutch', by Wim Klooster. (239-248)
In this squib, Klooster gives differences between equative constructions and
identifying sentences on the one hand, and tackles the puzzle of how the Dutch
neuter demonstrative, which carries the feature [singular], agrees with a finite
verb in the plural on the other hand.
Chapter 27: 'What you (and God) only know', by Marlies Kluck. (249-255)
Kluck proposes that the 'you-know-wh'-amalgams (interrupting sentences) are
idiomatic, frozen forms that are different from other wh-amalgams. She documents
many differences between 'you-know-wh'-amalgams and other wh-amalgams, showing
that 'God-knows-wh' amalgams pattern with other wh-amalgams.
Chapter 28: 'Is agreement resolution part of core grammar?', by Olaf Koeneman.
Koeneman's basic thesis in this squib is that agreement resolution is not part
of the core grammar, based on the fact that (a) a sentence can be grammatical
even if an apparent syntactic feature mismatch occurs; and (b) neither
morphology nor semantics proper seem to repair feature mismatches. To show this,
he looks at subject-verb agreement with disjunctive subjects in Dutch.
Chapter 29: 'On Dutch allemaal and West Ulster English all', by Hilda Koopman.
Koopman looks at the quantifier float (the case of a quantifier being
non-adjacent to a noun phrase that it quantifies over) in Dutch and West Ulster
English and proposes that in these languages, 'all'/''allemaal'' select vP as
their complements and attract a DP to their complements.
Chapter 30: 'The universality of binding principles: how appearances are
deceptive', by Eric Reuland. (277-287)
In this squib, Reuland argues that since language is a complex system, no
property of language can be interpreted independently of the system as a whole.
His evidence comes from binding theory. He points out that Reinhart & Reuland's
(1993) generalization that a reflexive predicate is reflexive-marked is part of
the more general principle of Inability to Distinguish Indistinguishables
(Reuland 2005, 2008), which means that the computational system cannot
distinguish between occurrences of identical variables in a domain that lacks
order and structure.
Chapter 31: 'Grappling with graft', by Henk van Riemsdijk. (289-298)
Van Riemsdijk's squib is on constraining graft, which is a combination of
internal and external merge. To avoid overgeneration, he restricts grafting of
DPs to cases where a DP gets only one theta-role (i.e. it is argument of only
one verb), as in Transparent Free Relatives. As for Head Internal Relative
Clauses, he proposes that theta roles are case-bound, and since graft is applied
to NPs below kase phrase (KP), no violation occurs in Japanese and Korean, two
languages with rich case morphology.
Chapter 32: 'Game theory and the control of empty categories in grammar', Tom
Roeper & Tim Roeper. (299-312)
This squib proposes a 'game theoretical' account of control structures. In this
model, control is a consequence of cooperation, which is a product of a complex
negotiation which can be encapsulated in a single word.
Chapter 33: 'Copy what?', by Ankelien Schippers. (313-320)
Schippers, showing that wh-copy constructions in Dutch are not accountable under
the Copy Theory of Movement (where a moving element leaves behind its copy
rather than a trace), proposes that wh-phrases in these constructions cannot be
copies of each other but rather distinct lexical items, following Koster (2009a).
Chapter 34: 'Free relatives at the interface', by Radek Simík. (321-328)
Šimík proposes that the definiteness property of free relatives is not due to a
D head but rather the finiteness/phase-character of the structure in conjunction
with type lowering.
Chapter 35: 'Menace under the microscope: the two verbs menacer and the theory
of control', by Dominique Sportiche. (329-339)
In this squib, Sportiche looks at the properties of ''menacer'' ('to threaten')
in French and shows that it is between a raising verb and a control verb,
concluding that the subject has to control PRO but does not have to bear a
certain theta-role. Therefore, he states, control options cannot be reduced only
to thematic structures but are also sensitive syntactic realizations.
Chapter 36: 'Case-alignment and verb placement', Knut Tarald Taraldsen. (341-346)
In this squib, Taraldsen tackles why there is no ergative SVO language. His
analysis is based on the assumption that higher functional heads change the case
assigned by a lower functional/lexical head. So, an object bears accusative
(acc) case only if the subject bears nominative (nom) case, and a subject bears
ergative case only if the object bears absolutive case. Even if SOV order is
achieved by remnant movement, the case-alignment will remain nom/acc.
Chapter 37: 'Diminutive Ks?', by Jindrich Toman. (347-351)
Toman proposes that the diminutive marker in Russian is not a separate
projection but rather the sister of the case morpheme, lacking a categorical label.
Chapter 38: 'Don't forget the determiners, Jan', by Henk J. Verkuyl. (353-358)
In this squib, Verkuyl argues against Jan Koster's theses that (a) a word lacks
meaning; and (b) concepts are not meanings but rather terms in our theories
Chapter 39: 'Empty subjects and empty objects', by Mark de Vries. (359-366)
De Vries shows that Koster's (1978, 2000, 2001) analyses of empty subjects and
empty objects are partly correct by proposing that subject clauses are
right-hand satellites semantically related to some anchor in the matrix clause.
Chapter 40: 'Island fever', by Edwin Williams. (367-373)
In this squib, Williams contends that island violations (i.e. a phrase moves out
of an island structure (islands do not allow extraction of any element from
within)) are instances of a generalized ''Improper Movement'' condition, proposing
the requirement that an element extracted out of a structure must target a
bigger structure. Otherwise, the derived structure ends as an island violation.
Chapter 41: 'Something else on variables in syntax', by Jan-Wouter Zwart. (375-383)
Zwart proposes that 'else', in expressions like 'something else', creates a
division of a set of focus alternatives different from a given focus
alternative, and that binding is not involved in the interpretation of
expressions like 'something else', and concludes that, therefore, the
parallelism between the interpretation of expressions like 'something else' and
the interpretation of pronouns is illusory.
This is a very good collection of squibs, most of which are on state-of-the-art
issues in generative syntax. They are written in a clear and easy-to-follow way,
which is a great achievement by the individual authors as well as the editors.
The diversity of the topics in the squibs, which take Koster's work as a basis,
clearly shows how influential Koster has been in the generative grammar framework.
However, with such a large collection, it would be better to arrange the squibs
in a topic-based way (such as 'wh-movement', 'DP-internal syntax', 'agreement',
etc.) rather than alphabetically according to the authors' names. This would
have made the squibs more accessible to the reader.
The following are a few closing comments on some squibs in the book.
In her squib, 'Cantonese as a tense second language', Lisa Lai-Shen Cheng
assumes that TP, complement to C moves to Spec, CP: Comp-to-Spec movement. She
refers the reader to others in order to see the motivation for this movement.
For such a controversial movement (for being 'too local'), it would be better to
summarize the motivations in others' works.
In his squib, 'Little words don't lie: X' have initial Xo', Joseph Emonds
proposes that the morpheme -ly is late-inserted in phonological form (PF)
between I(nfl) and a lexical verb. What is unclear about this proposal is how
phrase - non-phrase/head distinctions remain in PF. Assuming that information
regarding phrasal status remains in PF requires existence of Narrow Syntax in
PF, which makes the computational system more complicated. Also, the part of the
squib on the history of English includes unnecessary details.
Finally, it is unusual that Jindřich Toman's squib, 'Diminutive Ks?', has no
references. He could have referred the reader to at least descriptive studies on
diminutives in Russian and previous studies on diminutives in the framework of
Ackema, Peter & Szendröi, Kriszta. 2002. Determiner sharing as an instance of
dependent ellipsis. Journal of Comparative Germanic Linguistics 5: 3-34.
du Plessis, Hans. 1977. Wh movement in Afrikaans. Linguistic Inquiry 8: 723-726.
Hockett, Charles F. 1958. A course in modern linguistics. New York NY: MacMillan.
Hockett, Charles F. 1960. The origin of speech. Scientific American 203: 88-101.
Hockett, Charles F. 1963. The problem of universals in language. In Joseph H.
Greenberg (ed.), Universals of language, 1-22. Cambridge MA: The MIT Press.
(Also in a revised version, in The view from language: Selected Essays
1948-1974, 163-186. Athens GA: The University of Georgia Press, 1977)
Kayne, Richard S. 1994. The antisymmetry of syntax. Cambridge MA: The MIT Press.
Koster, Jan. 1978. Why subject sentences don't exist. In S. Jay Keyser (ed.),
Recent transformational studies in European languages [Linguistic Inquiry
Monograph 3], 53-64. Cambridge MA: The MIT Press.
Koster, Jan. 2000. Extraposition as parallel construal. Ms, University of
Koster, Jan. 2001. Mirror symmetry in Dutch. In Marc van Oostendorp & Elena
Anagnostopoulou (eds.), Grammar in Progress: Articles Presented at the 20th
Anniversary of the Comparison of Grammatical Models Group in Tilburg. Roquade,
Koster, Jan. 2003. All languages are tense second. In Jan Koster & Henk van
Riemsdijk (eds.), Germania et alia: A linguistic Webschrift for Hans den Besten.
Koster, Jan. 2008. A reappraisal of classical V2 and scrambling in Dutch.
Groninger Arbeiten zur germanistischen Linguistik 46: 27-54.
Koster, Jan. 2009a. IM not perfect: The case against copying. Ms, University of
Koster, Jan. 2009b. Ceaseless, unpredictable creativity; language as technology.
Biolinguistics 3: 61-92.
Reinhart, Tanya & Reuland, Eric. 1993. Reflexivity. Linguistic Inquiry 24: 657-720.
Reuland, Eric. 2005. Binding conditions: How are they derived? In Stefan Müller
(ed.), Proceedings of the HPSG05 Conference, 578-593. Stanford CA: CSLI.
Reuland, Eric. 2008. Anaphoric dependencies: How are they encoded? Towards a
derivation-based typology. In Ekkehard, König & Volker Gast (eds.), Reciprocals
and Reflexives: Theoretical and typological explorations, 499-556. Berlin:
Mouton de Gruyter.
Sekerina, Irina, Stromswold, Karin & Hestvik, Arild. 2004. How do children and
adults process referentially ambiguous pronouns? Journal of Child Language 31:
Sybesma, Rint. 2007. Whether we Tense-agree overtly or not. Linguistic Inquiry
ABOUT THE REVIEWER