LINGUIST List 26.2785

Fri Jun 05 2015

Review: Lang Doc; Morphology; Syntax; Typology: Nedjalkov, Otaina, Geniušienė, Gruzdeva (2013)

Editor for this issue: Sara Couture <>

Date: 31-May-2014
From: Anna Alexandrova <>
Subject: A Syntax of the Nivkh Language
E-mail this message to a friend

Discuss this message

Book announced at

AUTHOR: Vladimir P. Nedjalkov
AUTHOR: Galina A. Otaina
EDITOR: Emma Š. Geniušienė
EDITOR: Ekaterina Gruzdeva
TITLE: A Syntax of the Nivkh Language
SUBTITLE: The Amur dialect
SERIES TITLE: Studies in Language Companion Series 139
PUBLISHER: John Benjamins
YEAR: 2013

REVIEWER: Anna Alexandrova, Scuola Normale Superiore di Pisa

Review's Editor: Helen Aristar-Dry


This book is a translation of the “Sintaksis nivxskogo jazyka (Amurskij dialekt)” (“Syntax of the Nivkh language: The Amur dialect”), originally published in Russian (Nedjalkov and Otaina 2012). This is a posthumous edition of a draft found in the archives of Vladimir P. Nedjalkov, a major representative of the Leningrad/St.Petersburg Typological School. The work was not finalized because of the complicated situation in Russia in the early 1990s and the untimely death of Galina A. Otaina (1995), a native speaker of Nivkh.

Nivkh is a Paleo-Siberian isolate spoken in the Far East of Russia. Considering that there is no complete grammatical description of Nivkh in English yet, the volume under review is supposed to fill a gap in the literature on this endangered but still largely under-investigated language, and to make the Nivkh data more accessible to the linguistic community. The title is deceptive in a good way, in that the book covers a wide range of topics in morphology and semantics, going far beyond the syntactic level.

The opening chapter provides a sketch of the main typologically relevant parameters of Nivkh, which is characterized as a pro-drop, SOV, case-marking, almost exclusively suffixing, “agglutinating synthetic language” (p. 1). At the same time, it is remarkable for a complex system of consonant alternations, triggered by morphonological factors. Nivkh lacks adjectives. Qualities are expressed by verbs. The tense system is two-term (non-future vs. future). Adnominal possession is zero-marked, with the possessed preposed to the possessee. There is a huge inventory of converbs, while finite subordinate clauses and conjunctions do not exist.

In Chapters 2-5 and 7-8, morphonological rules are discussed: the change of initial plosive and fricative consonants at morpheme junction and in the syntactic complexes of the form ‘direct object + predicate’ and ‘attribute + head word’, the morphonological restrictions on the initial consonant (initial plosives occur in free variants of nouns and intransitive verbs, whereas initial fricatives are characteristic of free variants of transitives). The syntactic structures undergoing such morphonological alternations and those that do not (e.g., ‘subject + predicate’) are labeled ‘bound complexes’ and ‘free complexes,’ respectively. The problem of incorporation in the aforementioned complexes is discussed in Chapter 9 and a more articulated account of different types of syntactic complexes is provided in chapters 10-11. It is cautiously claimed that there are arguments against as well as in favor of incorporation (the overall identity of junction phenomena between bound morphemes and words and the phonetic wordhood of such complexes). The answer ultimately depends on the definition of incorporation.

Chapters 14-15 are dedicated to converbs, defined as non-finite verbal forms, “syntactically dependent on another verbal form without being its syntactic actant” (p. 39). Nivkh exhibits a huge inventory of converbs (about 20) with an extremely wide range of functions. The relevant features for a classification are: ±marking for subject agreement, ±same-subject converbal and independent action, ±tense marking, ±expression of temporal relations (taxis), ±occurrence in analytical verbal forms. It is shown how converbs are used to express temporal (taxis) relations, including sequential actions, and non-temporal ones, e.g., concession and consequence.

In Chapter 16, the nominal system and nominalization morphology is described, as well as denominal and deverbal postpositions and numerals. The case system comprises the zero-marked nominative (encoding subjects and direct objects), agentive, dative/additive, instrumental, locative, destinative, and comparative. There are at least two typologically uncommon cases in Nivkh, the agentive, and the form in ‘-ʁan’, which is almost extinct in the Amur but well-attested in the East-Sakhalin dialect. The uniqueness of the so-called agentive (‘dative/accusative’ in Panfilov 1962 and Gruzdeva 1998) consists in the restrictions imposed on the syntactic level. It is used to mark the causee with verbs containing causative markers, with reported speech converbs and some other constructions (cf. Malchukov 2008). The ‘-ʁan’-form is a nominal evidential, marking the subject of a verb denoting a hearsay action (‘reported nominative’ in Krejnovich 1979). The destinative case (‘limitative’ in Gruzdeva 1998) denotes “a limit in space or in time” (p. 54). Two series of markers encode the comitativity of single and plural items (dual and plural comitative, respectively). The comitative is not treated as part of the case system, presumably because it can co-occur with other case markers and is used as a means of coordination for NPs. In Mattissen (2003), the same gram is interpreted as the ‘correlative-associative number,’ standing in the number slot rather than in the case slot (Mattissen 2003: 8). One salient feature that distinguishes Nivkh from the neighbouring languages is its numeral classifier system, claimed to yield 27 nominal classes (in the literature, 24-30 classes have been reported; see Mattissen 2003: 15-16).

Chapter 18 is dedicated to adverbs, which can be (1) underived (e.g., pət ‘tomorrow’, naf ‘now’, ara ‘almost’), (2) lexicalized forms with instrumental or dative/additive case markers (e.g., ork-toχ ‘back; backwards’, məks-kir ‘truthfully’), and (3) derived by means of converbal suffixes (e.g., ‘be different’ &gt; ena-gu-r, ena-gu-t ‘otherwise, differently’).

In Nivkh, like in Turkic and Manchu-Tungus, ideophones constitute a remarkably large class, over 100 items (see chapter 19). Nivkh ideophones are simplex, invariable words. Verbs can be formed from them by suffixing the auxiliary verb ‘ha-’ (‘be so’). It should be stressed that ideophones are not always onomatopoeic. They frequently refer to non-audial domains: visual (e.g., matχ-matχ ‘being shaggy’, kmə ‘swarming, teeming of insects, fish, etc.’), bodily sensations (q‘ma q‘ma ‘sensation of an insect crawling on bare body’), and mental (q‘orx q‘orx ‘being stupid’).

Tense and aspect are covered in various chapters throughout the book. In Chapter 17, the future (the suffix -nə-) and desiderative (the suffix -inə-/-jnə-/-ijnə-) marking are discussed in detail. Within the two-term tense system (non-future vs. future), the future tense is overtly marked, while the non-future is zero-marked. The future marking is obligatory under future reference. Apart from the prototypical futurate meaning, the suffix -nə- has also modal and relative-tense uses. Two aspectual forms, the continuative and the usitative, are discussed in Chapter 20 (“Analytical and grammaticalized verbal constructions”). The Nivkh continuative is a grammaticalized construction of the form ‘converb in -r, t + auxiliary verb ‘hum-’ (‘be/exist’)’, an aspect which “expresses an action in progress or a state, often at the moment of utterance or during another action” (p. 93). The continuative is unacceptable with punctual and qualitative verbs (i.e., verbs expressing “adjectival” stative and, under certain conditions, a change of state). The usitative construction, formed with the auxiliary ‘ha-’ (‘be so/be like that/happen’), denotes a repeated, regularly performed action. In the same chapter several patterns of negation are accounted for, based on ‘negative’ verbs such as ‘-molo-’ (‘not want’), ‘-iki-/-əki-’ (‘be physically unable’), ‘-ləγə-’, ‘-qavr-’ (‘not have’), as well as the constructions denoting motion of the type ‘converb in -r, -t + finite verb’. The latter are labeled as serial verbs and are meant to “display a particular semantic cohesion”, since they “describe the same situation” (p. 104). This understanding of serialization, however, is not conventional at present: the verb forms that make up a serial verb construction should be combined without any linking means (see, for instance, Aikhenvald and Dixon 2006, Haspelmath 2015). The Nivkh “serial” construction consists of a converb from a verb denoting the manner or path and reference point of motion (e.g., pəi- ‘fly’, j-ur-/-ur- ‘move along some spatially stretched object / follow sth/sb closely’) and a finite form of a deictic verb of motion (e.g., viḑ ‘go’, p‘rəḑ ‘come’, mərḑ ‘ascend’, p‘uḑ ‘go out’ etc).

Verb morphemes are classified, in Chapter 21, according to the slots they occupy. Nivkh morphology is predominantly suffixing. An exception is constituted by the reciprocal and the reflexive, encoded by distinct prefixes. The reciprocal ‘v-/u-/o-’ can be added only to a limited set of verbs, covering a wide range of common reciprocal situations (about 30 verb stems according to the present book, about 45 stems including non-verbal ones according to Otaina and Nedjalkov 2007). The most productive reciprocal marker is a free pronoun (p‘-ŋafq-ŋafq [REFL-friend-friend] ‘each other’). Upon the whole, suffixes can be classified into three groups: (a) finiteness markers (-ḑ/-ţ and about 20 other suffixes, whose co-occurrence is not allowed); (b) about 10 particles and auxiliary words post-posed to the finiteness marker, most of which are transcategorial to a greater or lesser degree; (c) about 20 suffixes and bound verbal stems (e.g., negative verbs) preceding the A-slot (see below). Chapters 22-24 describe the distribution and functions of the morphemes pertaining to the three slots to the right of the verbal stem. Group A includes affirmative, interrogative, imperative, preventive, and optative markers. The ‘affirmative’ suffixes, i.e. those marking finiteness in affirmative sentences, encode epistemicity (e.g., certainty, probability, doubt) and focus. Group B mainly includes markers of modality, epistemicity, evidentiality, as well as emphasis and emotional attitude (annoyance, displeasure, irony). It contains also question markers, used in polar (the particle -la, synonymous to -lo/-l from group A) and, optionally, content questions (-ŋa). The evidentials include hearsay/imperceptive (-furu/-p‘uru/-vuru) and mirative (-hari). The former is a striking example of a transcategorial morpheme, compatible both with finite verb forms and nouns. When combined with nouns, it implies that the entity is not consistent with a standard. Unlike Groups A and B, which do not allow the co-occurrence of several morphemes on the same verb form, Group C does allow it. It includes valency-changing derivational morphology (causatives, anticausatives, resultative), tense (the future -nə-), aspect and Aktionsarten (desiderative/inchoative, distributive/intensive/completive, progressive, several usitatives, verbal diminutive and augmentative, etc.), and epistemicity (uncertainty marker -bañəvo-). Nivkh has both simplex and derived transitives. The valency-increasing causative suffix -ku-/-γu-/-gu-/-xu- expands the argument structure of the verb with an object in the nominative case. Valency-decreasing derivations are yielded by the anticausative suffix -r- (e.g., e-mq-/-moq- ‘break something in two’ ~ moq-r- ‘get broken in two’) and the resultative suffix -kəta-/-γəta-/-gəta-/-xəta-, introducing a result state of a prior event (Әmək ŋir+ḑosq-ţ mother cup+break-IND ‘Mother broke the cup’ ~ Ŋir ţosq-xəta-ḑ cup break-RES-IND ‘The cup is broken’). The progressive, marked by the suffix -ivu-/-jvu- (‘imperfective’/‘ingressive’ in Gruzdeva 1998, ‘progressive’/‘inchoative’ in Mattissen 2003), in combination with qualitative verbs, “expresses inchoativity, formation of a quality”, i.e., ultimately, a change of state in progress (p. 137; cf. Gruzdeva 1998: 29). The ‘positive usitative’ suffix -řa-/-t‘a- (‘iterative’ in Gruzdeva 1998: 31) denotes “regular resumption of an action or a continuous uninterrupted action” (p. 138). The suffix -su-/-ksu-/-γsu-/-jsu-/-ijsu is its negative counterpart (‘negative usitative’; cf. Gruzdeva 1998, 1997, where it is analyzed rather as the negative counterpart of the suffix -xə-, see below). It is meant to ‘denote actions that have not been taking place for a long time or not at all’ (p. 139). The ‘qualitative habitual’ -xə- (‘usitative’ in Gruzdeva (1998: 31), ‘habitual’ in Gruzdeva (1997: 183) and Mattissen 2003) characterizes the human subject as disposed to perform a certain action habitually. The verbal diminutive (-jo-) and augmentative (-kar-) pertain to Group C. The diminutive suffix denotes incomplete and/or attenuated manifestation of actions or qualities and, hence, it can be added both to event and qualitative verbs. The augmentative suffix yields an intensive meaning on a group of highly frequent qualitative verbs (see also Gruzdeva 1997, showing that both the diminutive and the augmentative suffixes mark semelfactivity).

The Nivkh valency classes (Chapter 25) are singled out according to three features: (1) ±ability of a verb to form an object complex; (2) the number of arguments; (3) argument marking (case, postposition, coordinating particle). In Nivkh, transitive verbs are such verbs that form an ‘object+verb’ complex (i.e., synthesize with) the second participant in the nominative case. Intransitive verbs range in valency from zero (impersonal verbs denoting atmospheric phenomena) to four, whereas transitives can have, at least hypothetically, up to five actants. Some observations concerning the semantics underlying the valency classes are sketched. For instance, the first participant (experiencer) of two-place emotion verbs (t‘axta ‘get angry’, q‘аlа- ‘hate’, k‘inŋu- ‘feel friendly / regard kindly’) is in the nominative and the second participant, the stimulus, is marked for the dative/additive. In three-place verbs implying the manipulation of a theme, argument marking may depend on a distinction such as between a stationary and a movable object. With the verbs of motion (e.g., vi- ‘go’, laγ- ‘go on a visit to another village’, ravi- ‘move to a new place’), the case marking on the second argument disambiguates between the direction of movement and the achievement of a spatial goal (dative/additive vs. destinative, respectively).

In Chapter 26 (“Semantic classes of verbs”), the lexical aspect system is sketched. The [±durativity] feature splits the verbs into ‘momentatives’ and ‘non-momentatives’. Momentatives and verbs of emotion such as ‘love’, ‘hate’, ‘despise’ (states in Vendler’s (1967) terms) cannot be used in the continuative. However, it is grammatical with the resultative forms of the momentatives and, quite predictably, with their iterative derivations. The verbs of motion, such as ‘p‘rə-’ (‘come’), can occur with the inchoative/progressive suffix -ivu-/-jvu-. The completive aspect (the suffix -kət-/-γət-/-gət-/-xət-) yields the meaning of completion of an action with homogeneous/multiplicative events. For it to be compatible with momentative verbs that do not presuppose any perceptible completion (e.g., ‘enter’, ‘come’) the subject has to be plural, yielding therefore a distributive meaning (‘each’, ‘all’). With stative and change-of-state verbs, the completive receives an intensifying reading. The ‘neutral’ verbs denote “both an action and the result of an action” (p. 156). It concerns mainly the state/change-of-state alternation (e.g., i-ndə-/-ñřə- ‘see’ vs. ‘find, get sight of’; kəpr- ‘stand’ vs. ‘stop’). One more class is constituted by the so-called ‘resultative-terminative’ verbs, “denoting actions which result in a visually observable state” (p. 157), yielding a result-state meaning with the resultative suffix -kəta-/-γəta-/-gəta-/-xəta-.

Chapter 27, “Sentence word order”,) mainly revolves around the position of the direct object, the subject, and the predicate. Word order is largely not involved in the expression of emphasis or focus, which are expressed morphologically and prosodically. A constituent with the focus particle -ra/-ta/-da can be optionally moved to the initial position in the sentence. What is more, a direct object can be stressed by using a free form of the verb instead of the bound one, i.e. without synthesizing with it. An intransitive verb can precede the subject if strong emphasis is put on it.

In Chapters 28-29, the main strategies of introducing direct and indirect speech are tackled. Most frequently verbs of speech, such as it- ‘say’, k‘əmlə- ‘think’, k‘esp‘ur- ‘tell, narrate, talk’, precede the direct speech, but it may also follow or be framed by two sentences containing introductory verbs. One more construction is presented by an auxiliary word (ha-r, ha-t &lt; converb in -r, -t from ha- ‘be so’), placed between the direct speech and the verb of speech in the final position. For indirect speech, Nivkh has a reportative converb (-vu-r, -vu-t), occurring only with the verb -it ‘say’.

Chapter 30 is dedicated to the hearsay/imperceptive marker -furu/-p‘uru/-vuru. It seems that it has grammaticalized from the verb fur-/-p‘ur- ‘tell about (sth/sb)’. The form ha-ḑ-furu [] ‘that’s how it was’, common in folklore, indicates that the event was not witnessed by the narrator.

Chapter 31, “Two-predicate constructions”, contains an account of constructions ‘deverbal noun in -ḑ + finite verb’, involving such meanings as ‘be late doing sth’, ‘not let sb do sth’, ‘like doing sth’, ‘be able to do sth’, ‘venture to do sth’, ‘want to do sth’. The nominalizing suffix -ḑ is homonymous with the marker of finiteness. Most of the dependent elements require a nominalization in the future tense. However, the choice of the tense form depends on the lexical class of the main verb. Three subclasses can be therefore singled out: (1) finite verbs taking a dependent ‘-ḑ’-suffixed form in the future tense only (j-ān-/-ān- ‘want’, j-azra-/-azra- ‘fear; apprehend’, t‘xərp-/-řxərp- ‘forget’, əskəsk- ‘be shy; be timid’, metra- ‘doubt’ etc.); (2) finite verbs accepting dependent forms in both tenses (e-zmu-/-smo-/-ţ‘mo- ‘rejoice; love’, j-ali-/-ali- ‘be unable to finish’, vār ‘be shy’, əγi- ‘not want’ etc.); (3) those requiring a dependent form exclusively in the non-future tense (j-im-/-him- ‘be able’; tvi- ‘finish’, muli- ‘often do sth’). The finite verb typically has some modal or phasal meaning or conveys the subject’s attitudes towards the action, expressed by the deverbal noun. The two forms can be same-subject or have non-corefential subjects. Another construction analyzed here is a different-subject construction ‘converb in -gu-r/-gu-t + finite verb’. The main verbs belong to several semantic domains: (1) sensory perception (mə- ‘hear; listen to’, i-ndə-/-ñřə- ‘see; find’, j-ama-/-ama- ‘look’); (2) emotional reaction and attitude (e.g., j-aŋr-/-aŋr- ‘be surprised; wonder’, j-itru-/-hitru- ‘ridicule; make fun of’; e-γəjiv-/-qrev-/-rev- ‘laugh at sb’, j-īz-/-hīz- ‘imitate’); (3) hindering and help (j-aγaγ-/-aγaγ- ‘disturb’, ro-/-to-/-do- ‘help’); (4) the speech verb it- ‘say’. The dependent converbal form presents an action as a process in its development, giving place to such sentences as ‘X heard Y saying sth’, ‘X saw Y doing sth’, etc. In an analogous construction the converb marked for the resultative (-γəta-gu-r, -γəta-gu-t) denotes a result state: If hu-in hə+ʁan e-sp-r i-x-kəta-gu-r i-də-ḑ [s/he there-LOC that+dog 3SG-stick-CONV:NAR:3SG 3SG-kill-RES-CAUS-CONV:NAR:3SG 3SG-see-IND] ‘There he found a dog stabbed to death’ (p. 227).

In Chapter 32, titled “Causative constructions formed by verbs with the suffix -ku-/-γu-/-gu-/-xu-”, the authors, in line with Xolodovič 1969, distinguish between factitive vs. permissive causation, concerned, respectively, with the modal meanings of necessity (causing events or result states) and possibility (e.g., ‘let X do sth’, ‘prevent X from doing sth’). In the former case, the Causer is the trigger or the only source of the events. In the latter case, the Causer only permits or prevents thems. Next, contact and distant causation are distinguished. Distant causation implies “a mediated connection between the causer subject and the state caused, when the causee is to a greater or lesser degree independent in accepting or rejecting the new state” (p. 233). Hence, permissive causation is always distant. Two-place transitive verbs, derived from one-place intransitives with the suffix -u-, are contact-factitive causatives, whereas the causative suffix -ku-/-γu-/-gu-/-xu- normally yields the distant-factitive or permissive meaning.

In Chapters 33–34, relativization strategies are assessed. Both subjects and objects can become heads of an embedded subordinate relative clause. The RC precedes the noun. However, the subject of the direct-object RC does not synthesize with the adjacent verb of the main clause. When the bound form of the noun in the ‘direct object + verb’ complex formally coincides with the respective free form (i.e., no contact morphonological changes occur), ambiguity may arise, for instance: Әtək za+umgu ţ‘oχt-ţ [father strike+woman be.drunk-IND] ‘The woman beaten by father is drunk’ vs. Әtək+za+umgu ţ‘oχt-ţ [father+strike+woman be.drunk-IND] ‘The woman who beat father is drunk’ (p. 262). To disambiguate an object RC from the other interpretation the object can be doubled by a 3rd person pronoun (the 3SG reduced form i- ‘s/he’ and the 3PL pronoun imŋ ‘they’). World knowledge is also important, because one of the syntactically possible readings may be blocked for pragmatic reasons. Coreferentiality between the subjects of the RC and the main clause is marked by the reflexive (p‘i ‘self’). In order to relativize an instrumental NP (e.g., ‘by boat’), a construction with a converb of the verb i-γr-/-kir-/-xir- ‘use’ is required.

In Chapter 35, “The structure of a narrative text”, a detailed account of the parameters of the Nivkh discourse is provided. It is shown how different types of converbs promote narration, denoting sequences of events. A succession of completed actions can be marked by finite forms in -ḑ/-ţ, but in this case each of them is perceived as a completed statement and is stressed by the speaker “thus building up tension for some culminating event” (p. 295). Nivkh offers different strategies of text cohesion, among them lexical repetition of the last verb of the preceding sentence in the form of a converb; conjunctive adverbs (‘taxis localizers’), placed as a rule in the initial position and expressing temporal (sequential) and causal relations. The latter are derived from demonstrative verbs ha- ‘be so’ and hoʁo- ‘be like that / do so’ with converbal suffixes and include meanings such as ‘then’, ‘therefore’, ‘after that’, ‘at this time’ etc. The authors provide plenty of quantitative data, for instance, the frequencies of conjunctive adverbs in dialogue and narration. Constructional variation in the use of converbs is illustrated by tasks where the informants restored the converbs in a text. In dialogue, as compared with narration, there is a significantly higher rate of the finite forms in -ḑ/-ţ and modal converbs.

In Chapter 36, titled “Aspectual and taxis characteristics of converbs”, special focus is made on the interaction of viewpoint and lexical aspect. The viewpoint aspect is intertwined with taxis (=relative tense): imperfectivity correlates with simultaneity, perfectivity is associated with anteriority. For instance, the durative converb in -ke cannot be formed for verbs denoting terminative non-durative actions, i.e. verbs such as oz- ‘get up’, p‘u- ‘go out’, i-γ-/-k‘u-/-xu- ‘kill’ etc. If the marker occurs on an inherently non-durative verb, the meaning of the latter is shifted to iterative. The progressive-inchoative converbs in -ivo/-jvo/-vo, whose only temporal meaning is simultaneity, descend from aspectual forms in -ivi-/-jvi-/-ivu-/-jvu- (subdialectal variants), also used as finite predicates, but compatible solely with verbs of motion and qualitative verbs. The continuative-stative converbs in -data-r, -data-t and -durŋu-r, -durŋu-t denote “a continuing action or state during which at one moment another action takes place” (p. 331); as a consequence, the only taxis meaning they can express is simultaneity. There are four perfective converbs, e.g. those in -ror, -tot, denoting a completed action and anteriority. However, if the head verb denotes an imperfective event, the event of the perfective converb partly overlaps with the latter. The so-called instantaneous converb in -ba/-pa refers to a completed action immediately followed by another. The completive converb in -fke expresses a completed event preceding the action of the main verb, which takes place regularly, continuously or after a certain time span. The authors also single out a group of perfective-imperfective converbs, whose aspectual interpretation (perfective vs. imperfective) relies on the lexical aspect of the verb (terminative vs. non-terminative). For instance, the converb in -ŋan marks contact anteriority or simultaneity, respectively. Neutral converbs are underspecified for aspect. A substantial portion of the chapter is dedicated to the assessment of the interaction between the aspectual meaning of specific converbs and aspectual grams (resultative, continuative, etc).


Undoubtedly, the audience of the book is primarily composed of typologists and specialists in Paleo-Siberian languages. In particular, the readers will find a wealth of data on verbal categories (TAM features, pluractionality, converbs), argument structure, valency classes and valency changing operations, numeral classifiers, reflexives and reciprocals, polysynthesis and the typology of word, and discourse structure. The range of topics covered in more detail goes in line with the major interests of the Leningrad/St.Petersburg Typological School. Sometimes full-length and more up-to-date accounts of certain phenomena can be found in later papers of the same authors, e.g. Otaina and Nedjalkov 2007 for the reciprocal constructions. This edition will be found handy even by those who can read the original Russian text, because only in the English version all the examples are glossed. A major advantage of this description of Nivkh morphosyntax is the abundance of extensively commented examples of each phenomenon discussed.

As far as the translation is concerned, the reader might have difficulty interpreting some linguistic terms pertaining to the Russian grammatical tradition, for instance, ‘homogeneous’, a literal translation of the Russian ‘odnorodlyj’, in such expressions as ‘coordination of homogeneous predicates’ (p. 47; Rus. ‘Sočinenie odnorodnyx skazuemyx’) or ‘conjoining of homogeneous nouns’ (p. 56; Rus. ‘soedinenie odnorodnyx suščestvitel’nyx’), which could be probably rendered better, for the sake of clarity, as ‘coordination’.

Needless to say, the publication of an unfinished draft of a grammatical description is a particularly challenging task for the editors who did an enormous work to make it accessible for the international linguistic community.


Aikhenvald and Dixon 2006 – Alexandra Y. Aikhenvald and R.M.W. Dixon. Serial Verb Constructions: a cross-linguistic typology. (Explorations in Linguistic Typology, 2.) Oxford: Oxford University Press, 2006.

Gruzdeva 1998 – Ekaterina Gruzdeva. Nivkh. München; Newcastle: LINCOM Europa, 1998.

Gruzdeva 1997 – Ekaterina Gruzdeva. Plurality of situations in Nivkh. In Viktor S. Xrakovskij (ed.), Typology of Iterative Constructions, 164–185. München; Newcastle: LINCOM Europa, 1997.

Haspelmath 2015 – Martin Haspelmath. The serial verb construction: Comparative concept and cross-linguistic generalizations (draft). 2015.

Krejnovich 1979 – Eruxim A. Krejnovich. Nivxskij jazyk. Jazyki Azii i Afriki, Vol. III, G.D. Sanzheev et al. (eds.). Moskva: Nauka, 1979.

Malchukov 2008 – Andrej Malchukov. Rare and ‘exotic’ cases. The Oxford Handbook of Case. / A. Malchukov and A. Spencer (eds.). Oxford University Press, 2008.

Mattissen 2003 – Johanna Mattissen. Dependent-Head Synthesis in Nivkh: A contribution to a typology of polysynthesis. Amsterdam: John Benjamins Publishing Company, 2003.

Nedjalkov and Otaina 2012 – Vladimir P. Nedjalkov and Galina A. Otaina. Očerki po sintaksisu nivxskogo jazyka. Edited by E. Geniušienė. Moskva: Znak, 2012.

Otaina and Nedjalkov 2007 – Galina A. Otaina and Vladimir P. Nedjalkov. Reciprocal constructions in Nivkh (Gilyak). In Vladimir P. Nedjalkov (with the assistance of Emma Geniusiene and Zlatka Guentchéva) (eds.), Typology of reciprocal constructions, 1715-1750. Amsterdam: Benjamins, 2007.

Panfilov 1962 – Vladimir Zinovjevich Panfilov. Grammatika nivxskogo jazyka. Moskva: Nauka, 1962.

Smith 1991 – Carlota S. Smith. The Parameter of Aspect. Kluwer Academic Press, 1991.

Vendler 1967 – Zeno Vendler. Linguistics in Philosophy. Ithaca, NY: Cornell University Press, 1967.

Xolodovič 1969 – Aleksandr Alekseevič Xolodovič (ed.). Tipologija kauzativnyx konstrukcij: Morfologičeskij kauzativ. Leningrad: Nauka, 1969.


Anna Alexandrova holds a degree in Russian and English philology. She is now a PhD student in linguistics at the Scuola Normale Superiore di Pisa (Italy). Her research interests include linguistic typology, Aktionsart, aspectual systems and verbal morphology, both in synchrony and diachrony.

Page Updated: 05-Jun-2015